perm filename CT.MSG[PAR,SYS] blob
sn#553565 filedate 1980-12-22 generic text, type C, neo UTF8
COMMENT ⊗ VALID 00228 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00023 00002 ∂04-Sep-80 1145 Christopher L. Tucci <CSD.TUCCI at SU-SCORE> [CSD.DRF: alphatype wrap-up]
C00025 00003 ∂04-Sep-80 1659 REP I'll be in
C00026 00004 ∂05-Sep-80 0930 JJK
C00027 00005 ∂05-Sep-80 1334 ALS Questions on EFTP
C00029 00006 ∂05-Sep-80 1530 AVB
C00030 00007 ∂05-Sep-80 2313 VRP on TTY161 2313 via SU-TIP
C00033 00008 ∂06-Sep-80 1204 DON EFTP
C00035 00009 ∂07-Sep-80 1021 VRP via SU-TIP Many happy returns!
C00036 00010 ∂07-Sep-80 2245 DON EFTP
C00037 00011 ∂08-Sep-80 0355 VRP on TTY161 0355 via SU-TIP
C00040 00012 ∂08-Sep-80 0415 VRP on TTY161 0415 via SU-TIP
C00041 00013 ∂08-Sep-80 1205 LTP Mission impossible
C00042 00014 ∂08-Sep-80 1308 FRG Time Sheet
C00043 00015 Message from VRP on TTY161 1327 via SU-TIP
C00046 00016 ∂09-Sep-80 1610 KDO
C00048 00017 ∂09-Sep-80 1911 CSD.KDO at SU-SCORE (Ken Olum) Bad code
C00049 00018 ∂10-Sep-80 0349 KDO I fixed EFTP
C00051 00019 ∂10-Sep-80 0356 KDO I Put up new eftp
C00052 00020 ∂10-Sep-80 1224 REP Another font
C00053 00021 ∂10-Sep-80 2233 ROZ
C00054 00022 ∂11-Sep-80 1227 ROZ
C00055 00023 ∂11-Sep-80 1458 JQJ via SU-SCORE eftp
C00056 00024 ∂11-Sep-80 1620 J. Q. Johnson <Admin.JQJ at SU-SCORE>
C00057 00025 ∂12-Sep-80 2044 KDO EFTP
C00058 00026 ∂13-Sep-80 1721 CT Transfering files to the VAXes.
C00060 00027 ∂14-Sep-80 0940 FB via SU-SCORE Shasta account
C00061 00028 ∂14-Sep-80 1019 CT
C00063 00029 ∂14-Sep-80 1902 CSD.NOWICKI at SU-SCORE Re: putting new fonts on spruce
C00065 00030 ∂14-Sep-80 2123 VRP on TTY36 2123
C00066 00031 ∂15-Sep-80 1443 DEK
C00068 00032 ∂16-Sep-80 0024 FML EFTP
C00069 00033 ∂16-Sep-80 0834 REG taff meeting
C00070 00034 ∂16-Sep-80 1542 PAM
C00071 00035 ∂16-Sep-80 2040 KDO I have to go home
C00072 00036 ∂17-Sep-80 1808 MRC new monitor command
C00073 00037 ∂17-Sep-80 2111 REP fonts...miles and miles of fonts
C00074 00038 ∂18-Sep-80 1350 DEK urgent request!
C00075 00039 ∂20-Sep-80 0645 DEK eureka
C00077 00040 ∂20-Sep-80 0651 DEK addendum
C00079 00041 ∂20-Sep-80 1307 DRF via SU-SCORE hi
C00080 00042 ∂21-Sep-80 1232 EBM at MIT-XX Re: Dover spooler
C00084 00043 ∂21-Sep-80 1331 DEK you rang?
C00085 00044 ∂21-Sep-80 1815 MOON at MIT-MC (David A. Moon) Dover spooler
C00086 00045 ∂21-Sep-80 2246 RMS at MIT-AI (Richard M. Stallman) Dover spooler
C00089 00046 ∂22-Sep-80 1217 LTP ALF directory
C00090 00047 ∂22-Sep-80 1640 LTP
C00091 00048 ∂22-Sep-80 2139 DEK via SU-TIP today
C00092 00049 ∂23-Sep-80 0140 REP Meeting & What I need urgently
C00094 00050 ∂23-Sep-80 1130 REP Sorry...
C00095 00051 ∂24-Sep-80 1113 CT KAREL.MF
C00096 00052 ∂24-Sep-80 1116 JP dover
C00098 00053 ∂24-Sep-80 1439 REP The dark side of CMTT fonts
C00099 00054 ∂24-Sep-80 2258 MRC DOVER program
C00100 00055 ∂24-Sep-80 2308 MRC DOVER
C00101 00056 ∂24-Sep-80 2351 JED DOVER
C00102 00057 ∂25-Sep-80 0856 J. Q. Johnson <Admin.JQJ at SU-SCORE> eftp
C00104 00058 ∂25-Sep-80 1243 JQJ on TTY163 1243 via SU-SCORE
C00105 00059 ∂25-Sep-80 1342 J. Q. Johnson <Admin.JQJ at SU-SCORE> [Chris Tucci <CT at SU-AI>:]
C00108 00060 ∂25-Sep-80 1342 J. Q. Johnson <Admin.JQJ at SU-SCORE>
C00109 00061 ∂25-Sep-80 1347 J. Q. Johnson <Admin.JQJ at SU-SCORE> Re: dover crash, cntinued.
C00110 00062 ∂25-Sep-80 1440 JMC dover
C00111 00063 ∂25-Sep-80 1540 JQJ on TTY162 1540 via SU-SCORE
C00112 00064 ∂25-Sep-80 1542 Bmoore at SRI-KL DOVER
C00113 00065 ∂25-Sep-80 1627 J. Q. Johnson <Admin.JQJ at SU-SCORE>
C00114 00066 ∂25-Sep-80 1756 PB dover
C00115 00067 ∂25-Sep-80 1756 TWE Dover
C00116 00068 ∂25-Sep-80 1814 TWE dover
C00117 00069 ∂25-Sep-80 1816 REP When you ask Knuth about the KAREL.MF, you might...
C00118 00070 ∂26-Sep-80 1108 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover program
C00119 00071 ∂26-Sep-80 1132 Brian K. Reid <CSL.BKR at SU-SCORE> Dover use
C00121 00072 ∂26-Sep-80 1541 Brian K. Reid <CSL.BKR at SU-SCORE>
C00122 00073 ∂26-Sep-80 1756 J. Q. Johnson <Admin.JQJ at SU-SCORE>
C00124 00074 ∂26-Sep-80 1801 Brian K. Reid <CSL.BKR at SU-SCORE>
C00125 00075 ∂27-Sep-80 1539 REP One run today
C00126 00076 ∂29-Sep-80 1609 ME EFTP command scanning
C00128 00077 ∂29-Sep-80 1639 ME E gripe response
C00131 00078 ∂29-Sep-80 1857 REG CSD-CF Staff meeting
C00132 00079 ∂30-Sep-80 0000 VRP
C00133 00080 ∂30-Sep-80 1823 ME
C00134 00081 ∂30-Sep-80 1824 Mark Crispin <Admin.MRC at SU-SCORE> Re: EFTP command scanning
C00135 00082 ∂30-Sep-80 2045 Mark Crispin <Admin.MRC at SU-SCORE> DOVER bug
C00137 00083 ∂30-Sep-80 2346 Craig Everhart at CMU-10A Dover program?
C00139 00084 ∂30-Sep-80 2347 Craig Everhart at CMU-10A Ah, yes.
C00140 00085 ∂01-Oct-80 0001 CT Expired plan
C00141 00086 ∂01-Oct-80 1444 CWR fonts
C00142 00087 ∂01-Oct-80 2033 WOO fonts
C00143 00088 ∂02-Oct-80 0734 Avron Barr <csd.barr at SU-SCORE> Fonts
C00144 00089 ∂02-Oct-80 1043 WIN via SU-SCORE Dover Bug
C00145 00090 ∂02-Oct-80 1120 REG CSD-CF Staff meeting is today!
C00146 00091 ∂02-Oct-80 1555 Craig Everhart at CMU-10A
C00147 00092 ∂03-Oct-80 0103 VRP via Dialnet Ethernet problem back
C00148 00093 ∂03-Oct-80 0613 The Mailer <OPERATOR at SU-SCORE> Message of 3-Oct-80 05:57:13
C00150 00094 ∂03-Oct-80 0857 REG Bug in EFTP
C00151 00095 ∂03-Oct-80 0912 REG WAITS problem / EFTP
C00152 00096 ∂03-Oct-80 0921 REG WAITS/ EFTP
C00153 00097 ∂03-Oct-80 0930 VRP via Dialnet ethernet
C00154 00098 ∂03-Oct-80 1343 WIN via SU-SCORE EFTP gets memory errors
C00155 00099 ∂03-Oct-80 1534 Brian K. Reid <CSL.BKR at SU-SCORE> Fonts.Widths file on SAIL
C00156 00100 ∂03-Oct-80 1539 AVB dover tabs
C00157 00101 ∂03-Oct-80 1649 GFS dover losing
C00158 00102 ∂03-Oct-80 2341 VRP via Dialnet eftp revived
C00159 00103 ∂04-Oct-80 1219 Brian K. Reid <CSL.BKR at SU-SCORE> why .PRE
C00161 00104 ∂04-Oct-80 1219 VRP via Dialnet pup server
C00162 00105 ∂04-Oct-80 1721 ME DOVER
C00163 00106 ∂05-Oct-80 2152 MRC
C00164 00107 ∂06-Oct-80 1412 ME eftp question
C00165 00108 ∂06-Oct-80 1456 MOG eftp broken
C00166 00109 ∂06-Oct-80 1720 CSD.GARDNER at SU-SCORE Dover problem
C00167 00110 ∂06-Oct-80 1837 Keith A. Lantz <CSL.LANTZ at SU-SCORE> eftp and dover
C00170 00111 ∂06-Oct-80 1847 KAL via SU-SCORE eftp cont.
C00171 00112 ∂07-Oct-80 1100 LTP Sink
C00172 00113 ∂07-Oct-80 1430 GFS illegal press file
C00173 00114 ∂07-Oct-80 1615 FJM broken EFTP
C00175 00115 ∂07-Oct-80 2238 VRP via Dialnet eftp
C00176 00116 ∂08-Oct-80 0327 MRC
C00177 00117 ∂08-Oct-80 0642 Avron Barr <csd.barr at SU-SCORE> font queston
C00178 00118 ∂08-Oct-80 0953 Avron Barr <csd.barr at SU-SCORE> Alphatype work
C00179 00119 ∂08-Oct-80 1231 ME EFTP trouble
C00180 00120 ∂08-Oct-80 1703 Avron Barr <csd.barr at SU-SCORE> Alphatype work
C00182 00121 ∂08-Oct-80 1721 JMC loop or something
C00183 00122 ∂08-Oct-80 2132 KAL via SU-SCORE dover
C00184 00123 ∂09-Oct-80 1135 ROB forwarded from GRIPES (BUG-eftp)
C00185 00124 ∂09-Oct-80 1337 PAM
C00186 00125 ∂11-Oct-80 2108 JED EFTP hassles
C00188 00126 ∂12-Oct-80 1725 ME looping DOVER program
C00189 00127 ∂12-Oct-80 1834 DEK alphtyping
C00191 00128 ∂12-Oct-80 1900 WIN Dover still does not work
C00193 00129 ∂13-Oct-80 0030 Mark Crispin <Admin.MRC at SU-SCORE> Re: Dover still does not work
C00197 00130 ∂13-Oct-80 1059 VRP eftp
C00199 00131 ∂13-Oct-80 1102 VRP eftp
C00200 00132 ∂13-Oct-80 1128 VRP eftp
C00201 00133 ∂13-Oct-80 1539 ME DOVER
C00204 00134 ∂14-Oct-80 0133 MDP Bad "password"
C00207 00135 ∂14-Oct-80 1102 REG notice is short, but this is only a reminder:
C00208 00136 ∂14-Oct-80 1300 ME DOVER and E dir
C00210 00137 ∂14-Oct-80 1741 AVB
C00213 00138 ∂14-Oct-80 1743 CSD.NOWICKI at SU-SCORE [CSD.NOWICKI at SU-SCORE: In the SAIL tradition]
C00216 00139 ∂15-Oct-80 1146 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul) Bill Nowicki's comments on press files
C00218 00140 ∂15-Oct-80 1200 Baskett at PARC-MAXC Press files
C00221 00141 ∂15-Oct-80 1201 Baskett at PARC-MAXC Press files and the Dover spooler
C00224 00142 ∂15-Oct-80 1205 CSD.GARDNER at SU-SCORE New problem with Dover
C00225 00143 ∂15-Oct-80 1301 ME DOVER
C00227 00144 ∂15-Oct-80 2037 REG
C00230 00145 ∂15-Oct-80 2057 REG
C00232 00146 ∂15-Oct-80 2145 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul) People and Software
C00236 00147 ∂16-Oct-80 0037 JLS TEX Press Output at SAIL
C00237 00148 ∂16-Oct-80 0942 JLS Dover Bug?
C00239 00149 ∂16-Oct-80 1019 JED .PRE files
C00240 00150 ∂16-Oct-80 1151 Fagan at SUMEX-AIM dover
C00241 00151 ∂16-Oct-80 1502 Bob Moore <BMOORE at SRI-KL> dover
C00243 00152 ∂16-Oct-80 1608 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
C00244 00153 ∂16-Oct-80 1622 CSD.NOWICKI at SU-SCORE [Ralph Gorin <REG at SU-AI>:]
C00246 00154 ∂16-Oct-80 1946 CSD.GARDNER at SU-SCORE
C00248 00155 ∂17-Oct-80 1131 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover.
C00249 00156 ∂17-Oct-80 1243 MSM
C00251 00157 ∂17-Oct-80 1455 FFL
C00252 00158 ∂17-Oct-80 1747 MRC via SU-SCORE DOVER
C00253 00159 ∂17-Oct-80 1858 CSD.NOWICKI at SU-SCORE Dover fonts
C00254 00160 ∂18-Oct-80 1710 ME DOVER
C00255 00161 ∂18-Oct-80 1905 WIN Bad pups again
C00256 00162 ∂20-Oct-80 0917 REG DOVER tabs and spacing
C00257 00163 ∂20-Oct-80 0920 REG more about TABS and spaces in DOVER.
C00258 00164 ∂20-Oct-80 0923 REG Finally, the truth
C00259 00165 ∂20-Oct-80 1859 CSD.GARDNER at SU-SCORE
C00261 00166 ∂21-Oct-80 0058 PAM Dover lossage
C00263 00167 ∂21-Oct-80 1003 DEK two things
C00265 00168 ∂21-Oct-80 1110 KDO What is the state of the Ethernet
C00266 00169 ∂21-Oct-80 1113 KDO DOVER suggestion
C00267 00170 ∂21-Oct-80 1146 KDO via SU-TIP DOVER big file bug
C00268 00171 ∂21-Oct-80 1759 Mark Crispin <Admin.MRC at SU-SCORE> Re: What is the state of the Ethernet
C00269 00172 ∂22-Oct-80 1914 MRC DOVER
C00270 00173 ∂22-Oct-80 2036 Brian K. Reid <CSL.BKR at SU-SCORE>
C00273 00174 ∂22-Oct-80 2042 Mark Crispin <Admin.MRC at SU-SCORE>
C00274 00175 ∂22-Oct-80 2042 Mark Crispin <Admin.MRC at SU-SCORE>
C00280 00176 ∂23-Oct-80 0812 Aikins at SUMEX-AIM Know anyone for this?
C00281 00177 ∂23-Oct-80 1312 Bill Nowicki <WIN at SU-AI> Bad Checksum
C00282 00178 ∂24-Oct-80 1211 LTP
C00283 00179 ∂25-Oct-80 1435 Purger
C00284 00180 ∂26-Oct-80 0923 DLB DOVER (at SAIL) problems...
C00287 00181 ∂26-Oct-80 1343 WDG eftp
C00288 00182 ∂26-Oct-80 1348 CSD.NOWICKI at SU-SCORE Re: SAIL 8 font.
C00289 00183 ∂26-Oct-80 1500 PAM
C00290 00184 ∂26-Oct-80 1522 Mark Crispin <Admin.MRC at SU-SCORE> Re: SAIL 8 font.
C00291 00185 ∂27-Oct-80 1453 LTP Chinese sample...
C00292 00186 ∂27-Oct-80 1513 GFS dover
C00293 00187 ∂27-Oct-80 1554 KDO More Dover bugs
C00294 00188 ∂27-Oct-80 2357 PAM DOver lossage again
C00296 00189 ∂29-Oct-80 0007 PAM
C00297 00190 ∂30-Oct-80 0910 IAZ I have problems with your GFI file - and Ishall try
C00298 00191 ∂30-Oct-80 2331 REP Final run of Karel
C00299 00192 ∂31-Oct-80 1030 LTP Chemicals
C00300 00193 ∂03-Nov-80 1507 ME staff meeting
C00301 00194 ∂04-Nov-80 1233 DRF alphatype unavailability
C00302 00195 ∂04-Nov-80 1727 EJG Illegal press file generated from large text file
C00305 00196 ∂06-Nov-80 0852 Avron Barr <csd.barr at SU-SCORE> new font
C00306 00197 ∂06-Nov-80 2159 ROB What??!?
C00307 00198 ∂11-Nov-80 1510 LTP Need to see you soon
C00308 00199 ∂13-Nov-80 1227 LTP
C00309 00200 ∂14-Nov-80 0844 HAL at MIT-MC (Hal Abelson)
C00312 00201 ∂14-Nov-80 1048 LTP
C00313 00202 ∂14-Nov-80 1249 AVB dover
C00315 00203 ∂15-Nov-80 1733 Avron Barr <csd.barr at SU-SCORE> Font question
C00316 00204 ∂18-Nov-80 0804 REG Staff Meeting
C00317 00205 ∂18-Nov-80 1104 LTP eftp
C00318 00206 ∂18-Nov-80 1256 Avron Barr <csd.barr at SU-SCORE> font hacking
C00319 00207 ∂19-Nov-80 1144 LTP on TTY77 (at TV-134) 1144
C00320 00208 ∂21-Nov-80 1415 HB gfi files for shasta
C00321 00209 ∂22-Nov-80 0718 HAL at MIT-MC (Hal Abelson) printing our book
C00322 00210 ∂24-Nov-80 1043 HB Missing gfi's
C00323 00211 ∂29-Nov-80 2138 ROB
C00327 00212 ∂30-Nov-80 0058 ME
C00328 00213 ∂30-Nov-80 1521 VRP via Dialnet ethernet
C00329 00214 ∂01-Dec-80 2330 PEG DOVER
C00330 00215 ∂01-Dec-80 2348 PEG eftp...
C00331 00216 ∂02-Dec-80 1636 PEG eftp
C00332 00217 ∂02-Dec-80 1650 ROD UDPs
C00333 00218 ∂03-Dec-80 0036 ME
C00334 00219 ∂03-Dec-80 1021 GFS dover bug
C00335 00220 ∂03-Dec-80 1306 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover error messages
C00336 00221 ∂03-Dec-80 2206 YM EFTP
C00337 00222 ∂07-Dec-80 0221 RMS The Help file for the dover program
C00338 00223 ∂07-Dec-80 0511 DON DOVER[3,2]
C00339 00224 ∂07-Dec-80 1316 DON cleaning up
C00340 00225 ∂07-Dec-80 1646 HB GFI files for shasta
C00341 00226 ∂09-Dec-80 1029 HB eftp to transfer DVI files to shasta
C00342 00227 ∂09-Dec-80 2341 MOG eftp
C00343 00228 ∂12-Dec-80 1206 ME EFTP
C00345 ENDMK
C⊗;
∂04-Sep-80 1145 Christopher L. Tucci <CSD.TUCCI at SU-SCORE> [CSD.DRF: alphatype wrap-up]
Date: 4 Sep 1980 1144-PDT
From: Christopher L. Tucci <CSD.TUCCI at SU-SCORE>
Subject: [CSD.DRF: alphatype wrap-up]
To: ct at SU-AI
Date: 3 Sep 1980 2028-PDT
From: CSD.DRF
Subject: alphatype wrap-up
To: csd.tucci
Everyone here was very pleased with the job you did. Send an itimized bill
to Sam Whidden, AMS, 201 Charles St., Prov., RI 02904. It should tell the
hours you worked, and the rate (seperate for "overtime"). I suggest you
talk to Luis about a good rate to charge; probably more than normal, and
be sure to include ALL the time you spent on this, not just the time in
the Alphatype room. Oh, it would be helpfull if you sent along a tape
with a dump of <TEX.FONTS> with it. Thanks, and see you soon.
--drf
-------
---------------
-------
hi. I was wrong about the 75 sheets. I printed up exactly 102 sheets.
see you later,
== chris. ==
∂04-Sep-80 1659 REP I'll be in
very early on Friday. I am going to make a call to Wiley at about 8:30, so I'd
like to have a pretty good idea what is going to work by then.
Rich
∂05-Sep-80 0930 JJK
To: CT, KDO
Hooray for EFTP!
∂05-Sep-80 1334 ALS Questions on EFTP
To: CT, KDO
1. What happens if 2 people try to transfer files at the same time,
particularly if they use the same file name.
2. Have you thought of writing an E macro to a) decipher the formatting
info at the end of each paragraph from the ALTO, primarily as to fonts
that are used to get regular WAITS characters and fix up the text and b)
to go the other way and add font info for the inverse purpose?
The line length problem from ALTO to WAITS is easily handled by a macro to
insert extra blank lines between the lines from the ALTO and then to JUST
or JFILL the entire page. I have found that this works.
∂05-Sep-80 1530 AVB
will you be adding name lookup for eftp from sail soon?
CT -- no, we will not. The Alto EFTP is not able to handle
it, and we cannot do anything at the other end. However,
we may produce a more sophisticated File Transfer Protocol
sometime in the future.
== chris. ==
∂05-Sep-80 2313 VRP on TTY161 2313 via SU-TIP
I've been sending loads of files from diablo to sail, mostly with great
success. A couple of minor problems:
1. How come eftp on diablo sends a whole bunch of control-G's to the
monitor when it's done? Actually I'm not sure if it is eftp that's
doing it, or whether some other process snarfs TTY27 while I'm away
handling the sail end of the transaction (I start up the diablo end
via sail's tty27).
2. How do you get exactly one cr inserted for each lf? Everything I've
tried gives me either lf alone (no cr) or something that looks
suspiciously like cr lf lf. The latter is actually ok for me (I'm
producing 68000 downloadable hex and the extra lf does no harm - I get
the extra lf using the options reqtl at the diablo end). Is it possible
that saying t at the diablo end plus aSCII at the sail end is the
problem?
CT -- 1.) To be honest, I really don't know exactly how the EFTP runs
on the VAX. SAIL does not send the ↑G.
2.) What we do here is insert one Line Feed after each carriage
return when going into SAIL. This is because .TTY files on
the Alto do not have any Line feeds in them. What I can do
is prompt you to see if you want the extra line feed there.
Look for that in a couple of days.
∂06-Sep-80 1204 DON EFTP
To: CT, KDO
Even without trying it, I have a gripe (assuming your documentation is
accurate). What if two people are sending files from different Altos
to the same site (e.g., SAIL)? There seems to be nothing in the SERVER
end at SAIL that specifies from which Alto to receive the file, so each
might get the other's file (or perhaps even a mixture of the two files?).
-- Don.
∂05-Sep-80 1334 ALS Questions on EFTP
To: CT, KDO
1. What happens if 2 people try to transfer files at the same time,
particularly if they use the same file name.
CT -- only one person can RECEIVE at a time, since the Alto sends to
socket 20. But we pick a unique socket for SENDING from, so
many people can send at once (to different Altos), if it is
any consolation.
== chris. ==
∂07-Sep-80 1021 VRP via SU-TIP Many happy returns!
Glad to hear you put in the lreturns. I'll experiment some more to
see if I can't win better.
∂07-Sep-80 2245 DON EFTP
To: CT
CC: KDO
You should mention in the documentation that only one person can receive
at any given time on any given host. You might even mention what the
program does if you try to set up a server and someone else at that site
already had one going.
∂08-Sep-80 0355 VRP on TTY161 0355 via SU-TIP
There's something wrong with what you're doing with cr's and lf's in eftp.
When I send a file with lf and no cr, using aSCII mode, you don't put in
the missing cr. If I send both lf and cr, you put in an extra lf. Are
you sure you aren't interchanging the roles of cr and lf somewhere?
PLEASE RE-READ MY REPLY TO YOUR LAST LETTER:
∂05-Sep-80 2313 VRP on TTY161 2313 via SU-TIP
I've been sending loads of files from diablo to sail, mostly with great
success. A couple of minor problems:
1. How come eftp on diablo sends a whole bunch of control-G's to the
monitor when it's done? Actually I'm not sure if it is eftp that's
doing it, or whether some other process snarfs TTY27 while I'm away
handling the sail end of the transaction (I start up the diablo end
via sail's tty27).
2. How do you get exactly one cr inserted for each lf? Everything I've
tried gives me either lf alone (no cr) or something that looks
suspiciously like cr lf lf. The latter is actually ok for me (I'm
producing 68000 downloadable hex and the extra lf does no harm - I get
the extra lf using the options reqtl at the diablo end). Is it possible
that saying t at the diablo end plus aSCII at the sail end is the
problem?
CT -- 1.) To be honest, I really don't know exactly how the EFTP runs
on the VAX. SAIL does not send the ↑G.
2.) What we do here is insert one Line Feed after each carriage
return when going into SAIL. This is because .TTY files on
the Alto do not have any Line feeds in them. What I can do
is prompt you to see if you want the extra line feed there.
Look for that in a couple of days.
∂08-Sep-80 0415 VRP on TTY161 0415 via SU-TIP
I'd much prefer that eftp quit after each transfer.
∂08-Sep-80 1205 LTP Mission impossible
I would like you to work with us in a graphics project, please come see me.
Thanx, Luis.
∂08-Sep-80 1308 FRG Time Sheet
Chris, Please come see Susan Hill, Rm.260 Margaret Jacks, concerning your
time sheet. Thank you. jake
;; Message from VRP on TTY161 1327 via SU-TIP
Sorry I didn't read your reply more carefully, I ended up accusing you of
making the mistake that I was making. What would be nicer than a prompt
about whether the extra line feed is wanted would be to add a little
extra smarts to the part that reads in the destination
file name. If the user types in / (or some such character not the start
of any file name) then the following word (delimited by a space) should
be read as a string of options each overriding some default. Thus for
someone transmitting an Alto file to the file foo, the response to
the destination file name prompt would be
/l foo
indicating to insert extra line feeds.
Alternatively the syntax foo/l might be closer to local practice.
What would also be nice is to be able to do all of this with one command.
It should be possible to say
receive foo/la
(or receive foo/l/a , a meaning ascii)
and
transmit foo/a
Incidentally, I know that image mode tends to be the default in ftp's,
on the ground that image mode is more efficient, but wouldn't a better
criterion be based on usage: what proportion of requests want ascii,
image, or don't-care-because-source-and-dest-are-compatible? I would
guess mostly ascii is wanted, but I don't have the stats.
∂09-Sep-80 1610 KDO
∂09-Sep-80 1330 CT DSPOOL
To: KDO at SU-AI
CC: csd.tucci at SU-SCORE
Hi. Guess what we are supposed to start working on?
Right! A Dover spooler for SAIL. What it would entail
would be to see if a file is in PRESS format, if not,
make it into PRESS format, then EFTP it over to the
Dover. Right now there are a couple programs of note
floating around: one converts XGP files into DVI files,
and (from what you said) one converts DVI files into
PRESS files. Hmmm. Interesting?
== chris. ==
KDO - Interesting indeed. However EFTP needs more work (I.E. cleanup, ABort
processing, etc, unless you've done it already. A person can presumably
send XGP files to the dover by using XGPDVI and then DVI%score and then
EFTP, perhaps the first thing we should do is to document this process.
∂09-Sep-80 1911 CSD.KDO at SU-SCORE (Ken Olum) Bad code
Date: 9 Sep 1980 1910-PDT
From: CSD.KDO at SU-SCORE (Ken Olum)
Subject: Bad code
To: ct at SU-AI
How am I supposed to figure out what "Disable Line Feeds" means? Even though
I know what the two options are, I don't know if disable line feeds means
convert cr to crlf or disable this conversion.
Ken
-------
∂10-Sep-80 0349 KDO I fixed EFTP
to call a procedure YESNO to ask for replies and to change the prompts so
they are more understandable. I also inverted the polarity of lfmode so that
LFMODE not 0 means special process linefeeds and LFMODE = 0 means pass them
normally.
Also fixed a bug whereby a file on exactly 512*N bytes long caused an extra
0-length packet to be sent.
We still have a problem, although maybe it is not our fault. Sending from
SAIL to SAIL and somewhat from SAIL to an ALTO generates Bad Checksum
errors, seemingly unpredictably. Sending to SAIL is ok, because the
packets don't get acknowledged and thus get sent again, but it generates
a fatal error sending to an alto (maybe this is old bad checksums on replies).
Also, there is a strange mode whereby packets get sent about 1 every 5 seconds,
which seems to be connected with starting the alto transmitting before starting
a server at SAIL, but I don't understand what is wrong.
(I don't know if the bad checksum packets contain good data or not)
I probably won't be in tomorrow, but I'm not sure.
Ken
∂10-Sep-80 0356 KDO I Put up new eftp
even though it has bugs because new version bugs seem to be entirely a subset
of old version bugs.
Ken
∂10-Sep-80 1224 REP Another font
Chris, could you please create a cmb12 and cmti12 for me from cmr12 within the
next week. I really need all the letters (i.e. the J and W must work). Let me
know when you get it done. Oh, I need versions for both the Alphatype and for
the xpg. If you have any trouble, drop me a line.
rich
CT -- CMB12 is now out on [xgp,sys] in all its glory...
∂10-Sep-80 2233 ROZ
To: CT, KDO
Does EFTP work to the VAX's on the ethernet?
CT -- yes. try sending to host 300 or 301 (i am not
sure which is Diablo or which is Shasta.
∂11-Sep-80 1227 ROZ
What is the host number for the ISL-VLSI VAX (St. Helens) ? m.
CT -- I don't know...maybe someone around the VAXes would...sorry.
∂11-Sep-80 1458 JQJ via SU-SCORE eftp
I've been getting quite a few BAD PUP 4 and CHECKSUM INCORRECT errors trying
to move files to the altos. Any ideas what the problems are? What data s
should I collect to make your debugging job easier?
CT -- are you sure you started the Alto EFTP up first? Usually that only
happens if you try to send before you try to receive...
∂11-Sep-80 1620 J. Q. Johnson <Admin.JQJ at SU-SCORE>
Date: 11 Sep 1980 1619-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
To: CT at SU-AI
In-Reply-To: Your message of 11-Sep-80 1523-PDT
OK, that gives me some data to go on. The error results in or is caused by a
timeout on the alto side, and I'll look to see which happens first. Any more
data I should be gathering?
-------
∂12-Sep-80 2044 KDO EFTP
works with 32 bit stuff, all correctly (now asks for R or L). Tried to
send .DVI files to the VAX, both from TEX at SCORE and from XGPDVI and
then to run DVI on them, and in both cases you get ?PST not found. Info
about .DVI files is now on DVI.TXT[1,KDO]. The file originally comes from
score but I got it from /USR/WIN/DVI.TXT@SHASTA.
(sending .DVI file to score and back gets you good file in return)
Good luck...
Ken
∂13-Sep-80 1721 CT Transfering files to the VAXes.
To: KDO
CC: CT
Hi:
I finally figured out why our files do not work
correctly on Unix. When we transfer our words 32
bits at a time, they end up swapped from the way
DVI expects them because the format of the .DVI files
is supposed to take the LOW ORDER BYTE first from every
word. Therefore, wordwise we are o.k., but within each
16-bit word on Shasta, the bytes are swapped; it is
therefore conceivable that we are sending a PST BEFORE
we are sending a EOP, which is very confusing to the
poor DVI program.
My question, then,is this: do we change "take-right-32-bit-mode"
to do what the DVI format needs, or do we make a "DVI-to-VAX"
mode that swaps the bytes? Hmmm...well, if I do not hear
from you by tomorrow morning, I will fo ahead with the
"DVI-to-VAX" business.
Also, if I have time tomorrow, I might TRY to get some sort
of option parser going...I think we are getting in sore
need of being able to type switches, as opposed to interrogating
the poor user for 15 minutes before he runs the program..
Adios,
== chris. ==
∂14-Sep-80 0940 FB via SU-SCORE Shasta account
I assume you got my Score mail about your new Shasta account.
∂14-Sep-80 1019 CT
To: WIN at SU-AI, CT at SU-AI, csd.nowicki at SU-SCORE
Subject: putting new fonts on spruce
To: nowicki
cc: tucci
hi:
Ken Olum and I are trying to put new fonts for the Dover out. After
floundering around with Metafont, FTP, FNTACR, FTP again, EFTP, and
PrePress, I have finally created these files:
1.) CMR10.AC
2.) CMR10.AL
3.) CMR10.ORBITS
4.) CMR10.WIDTHS
I have also played around with PrePress Merge, etc., but still can't seem
to get our TEX file printed. The error is:
ILLEGAL PRESS FILE 1: FONT SET 1 OR FONT 0.
If I could talk with you about this process, I would be greatly relieved.
Any time that I could meet with you would be fine. I will be in at 7am
every day this week and will probably stay until around 3-5pm. If you
could get ahold of me at SCORE or SAIL it would be great.
Also, how hard would it be to put out DVI on SCORE's TOPS-20 C? IF it is
possible it might make things easier for us, but no big deal...
thanks, and see you later,
-- chris. --
∂14-Sep-80 1902 CSD.NOWICKI at SU-SCORE Re: putting new fonts on spruce
Date: 14 Sep 1980 1856-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: Re: putting new fonts on spruce
To: CT at SU-AI
In-Reply-To: Your message of 14-Sep-80 1019-PDT
I'm glad to see someone is finally interested in doing this.
I assume you have seen my short document on it, [Lassen]<UnixDocs>UnixDocs.Press.
Unfortunately, I'm in Livermore right now. I plan to be back at Stanford
a week from next thursday (Sept 25). Let's get together then and figure things out.
I had used CMR10 as a test of my software last winter, and had gotten it all to work.
My DVI program needs some adjustments to handle DVI files produced on SCORE,
since the .widths are needed, and Score puts the full file spec in the DVI file
instead of just the font name.
I didn't even know Score had a C compiler. If we can do I/O on byte streams
we'll be all set. Does sucha beast exist on SAIL? I guess not.
Note that some useful programs to figure things out are the list option of the
PREPRESS program and the PRINTPRESS program, I think its called.
-- Bill
-------
∂14-Sep-80 2123 VRP on TTY36 2123
I suspect eftp is not to blame for the ↑G's I'm getting. It must be
something that DIAL is sending up the link when I'm not there.
I get them when I reset tty27.
CT -- that is very good news indeed!
∂15-Sep-80 1443 DEK
The following .TFX files have changed very slightly. (There was an error
in the ligature/kerning table, causing .02777em of extra white space
between the letters ho, mo, no.) I have changed them here at SAIL.
Please move these .TFX[1,3] files to SCORE:
cmb5
cmb6
cmb7
cmb8
cmb9
cmb10
cmb12
cmdunh
cmff
cmfib
cmr5
cmr6
cmr7
cmr8
cmr9
cmr10
cmr12
cms7
cms8
cms9
cms10
cmsc9
cmsc10
dmr7
dmr10
dmb7
dmb10
dms10
If TEXPRE was used to preload a lot of standard .TFX files, at SCORE as
we do here at SAIL, you should make a new TEX too using TEXPRE and then TEXSYS.
If you don't know how to do that, I guess it's OK to wait until DRF returns.
CT -- This would make a slight change, as you can see. Do you want me to
hold off on it until we put your book out, or just go ahead?
I also found out more on this backlash compensation business. Dave Fuchs
is planning to change the matrix sides to go completely from left to
right on the Alphatype. He will be back within two weeks. What I
figure to be the best thing to do would be to go ahead and do everything
first, then re-do the sheets that look bad after he gets back. What do
you think?
== chris. ==
∂16-Sep-80 0024 FML EFTP
What if there isn't enough room on the Alto disk for a file sent from Sail?
(seems to hang the system)
CT -- this should not happen. When the Alto disk gets full, the EFTP on
the Alto end should send an ABORT pup over to SAIL, along with an error
message. This seemed to work when we tested it out. Did this hang SAIL?
your Alto? both?
== chris. ==
∂16-Sep-80 0834 REG taff meeting
To: csd.hill at SU-SCORE, bosack at SU-SCORE, ROY at SU-AI,
ROB at SU-AI, MRC at SU-AI, csd.mogul at SU-SCORE, CT at SU-AI
CC: csd-cf at SU-AI
Notice is short, but we shall meet again for the usual bi-monthly.
Today at 1:30 in 252
∂16-Sep-80 1542 PAM
To: CT at SU-AI, RPG at SU-AI, DON at SU-AI
Last night (10 pm on 15 Sept), DON, RPG, and I were all logged in
on modem lines and suffered the problem that SAIL would only accept
about one character every 10 seconds, although it would send chars to
us at normal speed. The load average was 0.20, and the system seemed
otherwise healthy.. Does EFTP cause this grief in any known failure mode??
∂16-Sep-80 2040 KDO I have to go home
Haven't worked on CMR10 stuff except to generate CMR.PRESS on [IFS]<KDO> which
is ABCDEFGHIJKLMNOPQRSTUVWXYZ in CMR10 and tells me that all the character codes
are not found. I would suggest extracting the font from spruce.fonts and seeing
if it is really there, or comparing it with some other font similarly extracted.
I'll be in tomorrow afternoon sometime.
Ken
∂17-Sep-80 1808 MRC new monitor command
To: ME, ROB, REG, CT, KDO
The next system will have a DOVER command, to run the spooler currently
being developed by KDO and CT.
∂17-Sep-80 2111 REP fonts...miles and miles of fonts
Did you ever finish the last font I asked you to write (I don't even remember
what it was). One more favor...could you try getting xgp and alphatype versions
of CMTT (which is cmt10) except I now need cmt12. I ya, I just remembered
that you were going to darken cmtt. Hold off on cmt12 until I see how the darkening
worked....
So back to the original question...did you finish darkening cmtt?
Rich
∂18-Sep-80 1350 DEK urgent request!
Mike Plass is going to be making a couple new fonts and a new version
of TEX (to be called JTEX at SCORE), for use in preparing examples of
TEX's justification algorithm. I'll need to print this either Friday or
Saturday, since I want to take it to New York with me next week! Mike
will need some help getting this up on SCORE, so I hope you will be able
to help him tomorrow (Friday) morning. OK?
CT -- sure. I will be here at 7am as usual.
∂20-Sep-80 0645 DEK eureka
I woke up a few minutes realizing what the problem must be, and I wrote
MFP a note about it. It's something he can handle, there's no need
to involve DRF at this stage, so you don't need to make the phone calls.
In my file called TESTS.TMP[1,dek] is a list of the five files that I
want to be run off on the Alphatype, together with another file
ACPHDR.TEX[v2,dek] that is called in by one of the others. It would be
a good idea for you to FTP these six files (well, seven including
TESTS.TMP) to SCORE as soon as possible since MRC is taking SAIL down
tonight. Then you can run them as soon as MFP gets his program revised.
I believe they are all ready to go. They generate a total of 12 pages
of output, which will fit nicely on two sheets of film.
Call me at home (493-7875) if you have any problems. Thanks much
∂20-Sep-80 0651 DEK addendum
At the beginning of my last MSG I meant to say:
I woke up a few minutes ago realizing...
Also, I forgot to mention that you'll have to edit the file
MATH.TEX[1,dek] slightly after bringing it to SCORE: where it
says \input acphdr[v2,dek] you will have to change it to simply
\input acphdr (of course).
When TEXing that file math.tex, it will give you two error messages,
something like "You've already defined font 63 to be tnynum so you
can't redefine it to be cmsss8"; just hit carriage return to get past
these messages, since I don't really want cmsss8 etc. on this run even
though acphdr thinks I do.
∂20-Sep-80 1307 DRF via SU-SCORE hi
I'm in Ann Arbor. Try calling 313-763-6828 if you want me.
∂21-Sep-80 1232 EBM at MIT-XX Re: Dover spooler
Date: 21 Sep 1980 1527-EDT
From: EBM at MIT-XX
Subject: Re: Dover spooler
To: CT at SU-AI
In-Reply-To: Your message of 21-Sep-80 0149-EDT
I wrote the spooler, and do some maintenance of :DOVER, the program
that spools/creates PRESS files (from text files). Our methods for
shipment will probably not help you much, because we ship over the
CHAOS network, one of our local nets, to a gateway onto the Ethernet.
Special software was written by Dave Moon to permit us to ship a
file to a "pseudo-site" on the CHAOS net in the CHAOS net protocol,
and to have the protocol converted to EFTP in the gateway. So,
shipment is trivial (except for error handling, which is not too
bad either), because transferring a file on the CHAOS net is very
simple. Just as a side remark, the spooler is written in C, with
some use of low-level routines to interface to the operating system
(ITS), and is remarkably reliable, considering the modest effort
expended upon it.
As for the :DOVER program, that was written by RMS @ AI,
and the method for creating PRESS files from text files is
straightforward. However, the program will also accept files
destined for our XGP and convert them, in a very limited way. The
main limitation is that the fonts available for the two are entirely
different, and even if the thing knew of reasonable substitutions
(it was designed to do that, but the substitution tables have never
been created), the effects would not typically be what the user wanted.
I think the goal was to have some of the old formatters produce the
same format output file as they always used to, but supply the correct
font width and name info to them for the Dover. Anyway, creating
a PRESS file is not that hairy, it just takes a bunch of code.
The source is in MC:SYSEN1;DOVER > if you want to take a look. I
have written more sophisticated programs to create PRESS files, in
CLU,
-------
∂21-Sep-80 1331 DEK you rang?
I tried to call you back but you had stepped out. Try me again
493-7875, thanks
∂21-Sep-80 1815 MOON at MIT-MC (David A. Moon) Dover spooler
Date: 21 SEP 1980 2026-EDT
From: MOON at MIT-MC (David A. Moon)
Subject: Dover spooler
To: CT at SU-AI
Read the program. The file is MC:SYSEN1;DOVER >.
∂21-Sep-80 2246 RMS at MIT-AI (Richard M. Stallman) Dover spooler
Date: 22 SEP 1980 0144-EDT
From: RMS at MIT-AI (Richard M. Stallman)
Subject: Dover spooler
To: CT at SU-AI
:DOVER isn't normally a spooler. It turns the text file data into
press file data and outputs it to the dover all at once.
It goes through a gateway in a PDP-11. It used to talk PUP to the
dover and send the pups to the gateway to be sent on the ethernet.
It is much faster to use a chaosnet stream connection to the gateway,
which does the pup stuff. This is relevant to you because it means
that the present-day code isn't good for you, but there was earlier
code that might be more directly suitable for you to translate.
:DOVER does spool on MC, and a separate job sends the press file.
There is a directory on MC that holds spool files, each describing
some printing to do. It gives the name of a press file, which is also
likely to be on that directory. The spooled files are all press
files; :DOVER converts text files into press files when it is
spooling.
:DOVER also understands XGP files. This code should be suitable for
you to gobble. You would have to change the code that looks for
switches at the front of the file, since our format is different
∂22-Sep-80 1217 LTP ALF directory
CC: admin.mrc at SU-SCORE, LTP at SU-AI, DEK at SU-AI, CT at SU-AI
We need to keep the ALF directory protected so taht only a few people have
access to it. I prefer to have it out of the access of the rest of the
TEX project group, so could you please make sure that it can only be accessed
by using a password? (let it be DEK originally)
Thank you, Luis.
∂22-Sep-80 1640 LTP
∂22-Sep-80 1637 Mark Crispin <Admin.MRC at SU-SCORE> Re: ALF directory
Date: 22 Sep 1980 1635-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Subject: Re: ALF directory
To: LTP at SU-AI
In-Reply-To: Your message of 22-Sep-80 1217-PDT
ALF is now inaccessible by TEX group members; and its password has been
set to DEK, per your request.
-------
∂22-Sep-80 2139 DEK via SU-TIP today
I reran one page, and instead of using a whole sheet of paper I used
only the top half. So you'll find a half sheet in the
lite-tite box.
Have fun,
∂23-Sep-80 0140 REP Meeting & What I need urgently
Chris,
Hope moving in wasn't too much of a hassle. Remember that we are supposed to
meet at about 10:00 today (tuesday). I want to get a few (about 8) sample pages
off to Wiley by express mail later today, so we should run these off in the early
afternoon. Before we can do this, I need you to create 2 fonts and bring my
KAREL.MF file over (remember that those characters still didn't come out. You
should try reMETAFONTING the file.). Also, I'd like to have a bold face CMTT
font (as we discussed before). Let's call it CMTB10. Also, I'd like a 12 point
version called CMBT12. If you could start these three fonts when you get in,
I'd appreciate it. Again, I'll need them for about 2 sheets worth of proofs by
the afternoon.
See you soon (yawn)...rich
CT -- hi. I still haven't moved in yet, although I did move out. all
of my stuff is currently in a van right now. I don't see you around
right now, but it looks as if you were up fairly late. I will try to
start the fonts up. CMTB12 might take a while. I don't know if I can
finish it by this afternoon.
∂23-Sep-80 1130 REP Sorry...
I did oversleep. It was a late night.
The KAREL and bold CMTT font (CMTB10) are the most crucial, CMTB12 can wait.
Let me know when you get back so we can plan the rest of today.
Rich
∂24-Sep-80 1113 CT KAREL.MF
To: REP
CC: CT
Hi. Don is in New York this week. I am not sure when he will
be back, but I assume within a few days. He sent me a message
stating that he would be there "this week," and I forgot to
tell you last night.
== chris. ==
∂24-Sep-80 1116 JP dover
chris,
what kind of files can be printed by the dover program? do they have
to be press files or are you converting xgp into press? And also,
how can I try it out?
thanks.
jorge
CT -- Hi. I am not quite finished with the program yet. In any case,
here is what is SUPPOSED to do:
1. Take a text file.
2. Create a press file.
3. Send the press file to the Dover.
To run the program (yes, it WILL work when I finish the code...)
just type DOVER to the monitor.
Within a few weeks, we should be able to:
1. Actually SPOOL files.
2. Get TEX output printed.
Don't hold me to these "few weeks" too closely, as I am starting
school again, but I hope to work on this as much as possible. I hope
I have answered your questions; if you have any more, just send me
another message. See you,
== chris. ==
∂24-Sep-80 1439 REP The dark side of CMTT fonts
Just to repeat:
Keep CMBT10 and CMBT12.
Make a font whose darkness is between CMTT and CMBT10.
Make one more font whose darkness is between CMTT and the previous font.
This gives us a (0, 1/4, 1/2, 1) interval between CMTT and CMBT10.
Could you run off a sample with each of these fonts by friday afternoon.
I expet typesetting to commence by Saturday of Sunday. Maybe we could put a
bit more effort into finding why KAREL.MF doesn't work to well too.
Rich
∂24-Sep-80 2258 MRC DOVER program
I found and fixed a bug in DOVER causing it not to recognize uppercase
DOVER commands. Sigh.
∂24-Sep-80 2308 MRC DOVER
DOVER is generating bad press files, the message is
[603] Illegal Press file 1124: Improper number of records listed in document
directory (Dir)
CT -- Mark, do you know which file caused that to happen? I have been
trying it out this morning, and all seems well....I just put it back
out on sys:, with a warning that it may generate illegal press files.
∂24-Sep-80 2351 JED DOVER
Oops, it seems to have disappeared, or been removed:
.dover
? DOVER.DMP not found
↑C
.
Jim
∂25-Sep-80 0856 J. Q. Johnson <Admin.JQJ at SU-SCORE> eftp
Date: 25 Sep 1980 0850-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Subject: eftp
To: ct at SU-AI
It seems that EFTP flushes typeahead, which I find very undesirable. Consider
my situation: I want to have a MIC control file at SCORE that contains
several EFTP commands. Unfortunately, it doesn't complete, because after
asking for the destination number the program apparently flushes my remaining
commands (which MIC had typed much faster than they could be executed, since
the TN protocol has no provision for flow control).
How about a better command scanner to EFTP that doesn't do that silly
command completion?
-------
CT -- Basically, it does not clear the terminal input buffer except
when you make a mistake in typein. The only exception to this
exception is that it would clear after RECEIVE (I did change that.)
However, you seem to indicate that it was clearing after a SEND.
∂25-Sep-80 1243 JQJ on TTY163 1243 via SU-SCORE
DOVER crashes with a Halt at user 2113
CT -- If you would not mind, could you do me a favour? Please
re-try running DOVER in hand-hold mode (i.e. just type DOVER
to the monitor -- it will prompt for a file) using the same file.
This will give you an error message hopefully. Then tell
me the name of the press file so I can pick it apart and see
what exactly is wrong.
I had some problems with large files, but thought that I had
fixed those....oh, well..thanks for the message.
∂25-Sep-80 1342 J. Q. Johnson <Admin.JQJ at SU-SCORE> [Chris Tucci <CT at SU-AI>:]
Date: 25 Sep 1980 1339-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Subject: [Chris Tucci <CT at SU-AI>:]
To: ct at SU-AI
Yes, it was on a SEND. Maybe I was just losing because I was exceeding the
typeahead buffer, but I was only 50 or so characters ahead, and assumed that
the buffer was about 160 chars. The file in question is <ADMIN.JQJ>ODOVER.MIC
---------------
Mail-from: ARPANET site SU-AI rcvd at 25-Sep-80 1320-PDT
Date: 25 Sep 1980 1307-PDT
From: Chris Tucci <CT at SU-AI>
To: admin.jqj at SU-SCORE
∂25-Sep-80 0856 J. Q. Johnson <Admin.JQJ at SU-SCORE> eftp
Date: 25 Sep 1980 0850-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Subject: eftp
To: ct at SU-AI
It seems that EFTP flushes typeahead, which I find very undesirable. Consider
my situation: I want to have a MIC control file at SCORE that contains
several EFTP commands. Unfortunately, it doesn't complete, because after
asking for the destination number the program apparently flushes my remaining
commands (which MIC had typed much faster than they could be executed, since
the TN protocol has no provision for flow control).
How about a better command scanner to EFTP that doesn't do that silly
command completion?
-------
CT -- Basically, it does not clear the terminal input buffer except
when you make a mistake in typein. The only exception to this
exception is that it would clear after RECEIVE (I did change that.)
However, you seem to indicate that it was clearing after a SEND.
---------------
-------
∂25-Sep-80 1342 J. Q. Johnson <Admin.JQJ at SU-SCORE>
Date: 25 Sep 1980 1337-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
To: CT at SU-AI
In-Reply-To: Your message of 25-Sep-80 1312-PDT
I get the same effect when I run DOVER in handheld mode: Halt at user 2113
followed by ↑C. The file in question is INTRO.PRT[1,JQJ].
-------
∂25-Sep-80 1347 J. Q. Johnson <Admin.JQJ at SU-SCORE> Re: dover crash, cntinued.
Date: 25 Sep 1980 1342-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Subject: Re: dover crash, cntinued.
To: CT at SU-AI
In-Reply-To: Your message of 25-Sep-80 1330-PDT
The text file is <CSD.QUEENIE>INTRO.MSS @ score. I'll ftp it to INTRO.MSS[1,JQJ]
for you to compare with.
-------
∂25-Sep-80 1440 JMC dover
Until program tells you that it has done the listing, the writeup
should tell you that the program doesn't.
CT -- right. sorry about that; I was suppressing all display from DOVER FOO.BAR mode.
∂25-Sep-80 1540 JQJ on TTY162 1540 via SU-SCORE
"file not found on PRESS file I just created!"
CT -- were you sending a PRESS file or a TEXT file?
∂25-Sep-80 1542 Bmoore at SRI-KL DOVER
Date: 25 Sep 1980 1541-PDT
From: Bmoore at SRI-KL
Subject: DOVER
To: ct at SAIL
cc: bmoore
In your list of things to be implemented in the near future you did not
list being able to print XGP files. Are there any active plans to allow
this?
Thanks,
Bob Moore
-------
CT -- yes. In the near future. (I will add it to the list.)
∂25-Sep-80 1627 J. Q. Johnson <Admin.JQJ at SU-SCORE>
Date: 25 Sep 1980 1623-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
To: CT at SU-AI
In-Reply-To: Your message of 25-Sep-80 1613-PDT
I was sending a press file, INTRO.PRT[1,JQJ]. It shouldn't have had to
do ANYTHING. Looks to me as if it got confused, perhaps it figured out in one
place that the thing was PRESS and didn't need converting, then later decided
that it had to read the converted file that wasn't.
-------
∂25-Sep-80 1756 PB dover
Lookout! Here come the complaints.
Viz., I observe that in my text file, printing on the Dover resulted in
} printing as ~ and _ having 0 width (i.e. underprinting following char).
I suppose the latter may be an intrinsic property of the font (probably a good
one, but not compatible with SAIL usage), but the former seems curable by
a substitution (I haven't noticed what does come out as }). Otherwise
THANKS!
--peter
CT -- I will fix this first problem today.
∂25-Sep-80 1756 TWE Dover
I was unable to print the file ns.me[s,doc]. I tried
dover twice, and eftp once. Dover said bad press file format
mumble directory. EFTP said bad "password". Help!!
CT -- o.k...that bug has now been fixed. thanks for the message.
∂25-Sep-80 1814 TWE dover
foo and foo.prs on [1,twe] are copies of ns.me[s,doc] and
ns.prs. Same problem, illegal number of records listed in
document directory.
Tim
∂25-Sep-80 1816 REP When you ask Knuth about the KAREL.MF, you might...
...mention that I need to produce the proof pages by Wednesday or Thursday.
This might make him pay more attention to the problem (although we haven't had
trouble with this before). Will you have a sample of the CMBT fonts by tomorrow
afternoon?
Rich
∂26-Sep-80 1108 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover program
Date: 26 Sep 1980 1105-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Subject: dover program
To: ct at SU-AI
Guess what? Brian Reid is going to start calling press files *.press instead
of *.prt here. So if people aren't careful they are going to get files at
SUAI called *.PRE instead of *.PRT or *.PRS. More incompatability. Would
it be possible in dover to have a switch that defaulted but which could be
used to override the extension-determination?
-------
∂26-Sep-80 1132 Brian K. Reid <CSL.BKR at SU-SCORE> Dover use
Date: 26 Sep 1980 1125-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
Subject: Dover use
To: CT at SU-AI
Scribe produces files on the 20 with extension PRESS, which when they
are FTP'd to Sail get the extension clipped to PRE. The Dover program
does not seem to realize that PRE is a press file. Could you either
(1) look for the extension PRE, or (2) better yet look for 8-bit byte
size, or (3) even better yet, look at the file to see if it is a
press file, by checking to see if it has the magic signature in it and
that the last page is a directory page.
Brian
-------
CT -- I will check for .PRE extensions, although I am kind of against
doing it for the reason that if you look at FOO.PRE, it is not
OBVIOUS what kind of file you are dealing with. (.TEX, .SAI, etc. are
good examples of OBVIOUS file extensions.) I think that for three
letters, .PRS expresses the fact that the file is a press file best.
In other words,I will check for the .PRE files, because you asked me
to, but wouldn't it make much more sense just to say:
*RETR FOO.PRS←LONGFOO.PRESS ?
See you later,
Chris.
∂26-Sep-80 1541 Brian K. Reid <CSL.BKR at SU-SCORE>
Date: 26 Sep 1980 1535-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
To: CT at SU-AI
In-Reply-To: Your message of 26-Sep-80 1145-PDT
The right thing to do is to ignore the extension and look inside the file
and see whether or not it is a press file. With random-access I/O that
is a very simple operation, and it is what all other press file spooling
programs of my acquaintance do.
-------
CT -- right you are. this should be done by tonight some time.
∂26-Sep-80 1756 J. Q. Johnson <Admin.JQJ at SU-SCORE>
Date: 26 Sep 1980 1751-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
To: CT at SU-AI, csl.bkr at SU-SCORE
In-Reply-To: Your message of 26-Sep-80 1145-PDT
Brian is certainly right that the optimal strategy is to look at the file
and decide if it is press format, using an algorithm that is sophisticated
enough that it doesn't make the mistake too often of thinking its an ascii
file if it isn't. Remember our discussion -- the cost of thinking a press
file is really ascii is much higher than the cost of thinking an ascii file
is press. Getting it wrong once in about 1000 times is probably quite
acceptable, especially if you can interpret the message that you get back
from the dover ("not press format" or whatever) and if you have sent a file
without empressment try pressing it.
I understand fully that you don't want to invest too much effort in this.,
though.
Incidentally, is the empressment routine sufficiently modular that we will
be able to pull it out and use it in the score dover spooler if an when?
-------
∂26-Sep-80 1801 Brian K. Reid <CSL.BKR at SU-SCORE>
Date: 26 Sep 1980 1755-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
To: Admin.JQJ at SU-SCORE, CT at SU-AI
In-Reply-To: Your message of 26-Sep-80 1751-PDT
CMU has a beautifully functional empressment program that runs on Tops-20
that Score can steal. It produces output in any font, any orientation,
any number of columns, with running heads, substitution, and all sorts
of options like page selection. MIT-XX has one, too, but it hasn't had
the year and a half to mature that CMU's has.
-------
∂27-Sep-80 1539 REP One run today
To: CT at SU-AI, csd.tucci at SU-SCORE
Cris, I want to run off one sheet of alphatype paper today. Let me know when
you get in. I am expecting to do the big run on Wednesday, with a carry over
to Thursday if necessary. Hope you can make it during this time. If not--and
Dave not back--I should really write down how to do everything.
Rich
∂29-Sep-80 1609 ME EFTP command scanning
To: CT at SU-AI
CC: MRC at SU-AI
∂29-Sep-80 1534 CT EFTP monitor command.
To: MRC at SU-AI
CC: ME at SU-AI
Mark:
Would it be possible to get an EFTP command at SAIL? I just
modified EFTP to RESCAN the input line so (theoretically) one
will be able to say:
.EFTP FOO.BAR TO 116 /il
etc. This will be a lot less cumbersome than the present
Spanish Inquisition/U.S. Census method of interrogating the
user for information...
Chris.
ME - Until an EFTP command can be added, there is no reason you can't have
EFTP rescan the monitor command that says "R EFTP;FOO.BAR TO ...". Note
the semicolon which makes the monitor ignore the rest of the command line,
but you can start scanning at the semicolon after you have noticed that
the line starts with the R command. Many programs do this.
We can add an EFTP command when the next system is put up.
∂29-Sep-80 1639 ME E gripe response
∂20-Sep-80 2048 CT line editor over-ride.
To: BUG-E
Would it be possible to implement some sort of line editor
over-ride that performs top-level E commands in the middle
of a line? For instance, you are in the middle of the line
and you want to exit and save the changes. But when you type
<ctrl>-<meta>-E (which is a VERY natural thing to do), you
simply get the frustrating message:
Sorry -- Command illegal from line editor here -- <ctrl><meta>e
Also, if you want to kill a line in the middle, <ctrl><meta>d
gets you the same sort of response.
Now I realise that people can say: "oh, what the hell is so bad
about just hitting CR or <ctrl>CR. CT must be spoiled....etc."
but this is costing me more than one extra key-stroke. Because
I invariably do this once every other editing session. So I do
αβE, then CR, then αβE, and I invariably get really frustrated.
I am not sure what public opinion on this is, but I would bet that
more often than not the message usually goes unappreciated.
Chris Tucci
ME - Well, no it would not be possible to do that at this point,
and I'm not in favor of being able to exit from within the line
editor anyway. Sorry, but to use E, you really have to realize
that being in the line editor is special, and there are various
things you can't do from there, mainly for safety.
∂29-Sep-80 1857 REG CSD-CF Staff meeting
To: "@STAFF.DIS[CSD,REG]" at SU-AI
Will be held Thursday, Oct 2, instead of our customary Tuesday meeting.
1:30 in room 252.
∂30-Sep-80 0000 VRP
How come r eftp does an Ill mem ref at user 1337? I can't correlate it
with anything abnormal, I'm just sending a file to sail from diablo
using eftp. Let me know what diagnostics would be helpful.
∂30-Sep-80 1823 ME
Your DOVER documentation fails to even mention that it writes a .PRS
file on the user's directory, which the user might like to delete (if not
the program).
∂30-Sep-80 1824 Mark Crispin <Admin.MRC at SU-SCORE> Re: EFTP command scanning
Date: 30 Sep 1980 1819-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
Subject: Re: EFTP command scanning
To: ME at SU-AI, CT at SU-AI
cc: MRC at SU-AI
In-Reply-To: Your message of 29-Sep-80 1609-PDT
The next version of WAITS will have an EFTP command. ME is right though
that EFTP should scan for ";" if not invoked by an EFTP or TEST command.
DOVER already has this feature.
-------
∂30-Sep-80 2045 Mark Crispin <Admin.MRC at SU-SCORE> DOVER bug
Date: 30 Sep 1980 2043-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
Subject: DOVER bug
To: CT at SU-AI
ALPHAM.DOC[1,MRC] still causes DOVER to generate a bad PRESS file. The
message is
[302] Illegal Press file 1762: Internal pointer exceeds file bounds.
This is the file which has bare CR's. The other file with bare LF's now
works like a charm.
-------
CT -- yes, I discovered this one 2 days ago, when someone else brought
me a file with bare CR's. I will try to work on it this weekend.
∂30-Sep-80 2346 Craig Everhart at CMU-10A Dover program?
Date: 1 October 1980 0245-EDT (Wednesday)
From: Craig Everhart at CMU-10A
To: CT at SU-AI
Subject: Dover program?
Message-Id: <01Oct80 024500 CE10@CMU-10A>
I saw your note on the SAIL bboard, so I TNed to Sail and read your Dover
document, then the program. I think I'd be remiss if I didn't tell you that
I wrote a convert-and-queue Dover program for CMU PDP-10s last fall, with
many options and things, that might just save you lots of work. We ship
things to the Ether totally differently around here, but the text-to-Press
conversion is the same.
Craig Everhart
CT -- Is there any way for me to get a hold of the code? I am still
getting a few bugs, and to see how you did it would be really helpful.
∂30-Sep-80 2347 Craig Everhart at CMU-10A Ah, yes.
Date: 1 October 1980 0245-EDT (Wednesday)
From: Craig Everhart at CMU-10A
To: CT at SU-AI
Subject: Ah, yes.
Message-Id: <01Oct80 024553 CE10@CMU-10A>
The document for my program is on DOC:DOVER.DOC on CMUA. HLP:DOVER.HLP
is briefer and more of a summary.
CT -- thanks for this info!
∂01-Oct-80 0001 CT Expired plan
Your plan has just expired. You might want to make a new one.
Here is the text of the old plan:
home phone 327-5071 after 8:00 p.m.
I work full time around MJH, and can usually be found in the general
vicinity of the Alphatype Darkroom. Mail to CT@Sail, CSD.TUCCI@Score,
TUCCI on the Ether, or S.SCT at LOTS.
∂01-Oct-80 1444 CWR fonts
Chris, I need to write a tape with the tex.fonts on it. I can't use something
like tapeio, or dumper either since it's going to a non-dec site. Could you
recommend some method for accomplishing this? Any format i specify using
one of the canned programs gives me massive truncations, since it's a binary
formatted file. Are there any programs around that already do this?
Thank you.
Chuck
ps: dinner soon?
\
∂01-Oct-80 2033 WOO fonts
To: CT at SU-AI, csd.tucci at SU-SCORE
this message is from rich pattis call me as soon as you get in thursday morning
856 0948 urgent
∂02-Oct-80 0734 Avron Barr <csd.barr at SU-SCORE> Fonts
Date: 2 Oct 1980 0733-PDT
From: Avron Barr <csd.barr at SU-SCORE>
Subject: Fonts
To: csd.tucci at SU-SCORE
cc: ct at SU-AI
Hi Chris,
I guess you are back in school now and getting busy. Any chance
you could help us out with converting that graphics font to meta-font
next week? It's critical now. Also, do you know the name of the
.TFX version of the standard SAIL font--I can't find it. Is there a
font catalogue for TEX?
.Av
-------
∂02-Oct-80 1043 WIN via SU-SCORE Dover Bug
The file PREPRE.PRE[1,WIN] is a press file that causes the Dover
program to hang for a while, then get Ill mem ref at user 1541.
-- Bill
CT -- thanks. I will not have time to work on this until this
weekend, but I hope to have a marathon bug-fixing session then...
∂02-Oct-80 1120 REG CSD-CF Staff meeting is today!
To: "@STAFF.DIS[CSD,REG]" at SU-AI
At 1:30 in room 252.
∂02-Oct-80 1555 Craig Everhart at CMU-10A
Date: 2 October 1980 1820-EDT (Thursday)
From: Craig Everhart at CMU-10A
To: Chris Tucci <CT at SU-AI>
In-Reply-To: Chris Tucci's message of 2 Oct 80 09:38-EST
Message-Id: <02Oct80 182044 CE10@CMU-10A>
All the files on [S200DV00], basically. I think the ones of greatest interest
to you are Fonts.Bli, Press.Bli, maybe Dover.Bli and DovScn.Mac. IPCQue.Bli
does queueing to our spooler daemon. DovSt.Bli has a start address, a version
number, and an edit history, and nothing else. I'll send more information
if you like, but check out the document (from DOC:).
Craig
∂03-Oct-80 0103 VRP via Dialnet Ethernet problem back
The reason your bug went away was that diablo's ethernet gave up
altogether. Now that it's fixed, the Ill mem ref at user 1337 is back.
If this could be fixed fairly soon it would be much appreciated,
there is no other satisfactory way for me to get data to sail, and I'm
kind of in a bind at the moment as a result.
∂03-Oct-80 0613 The Mailer <OPERATOR at SU-SCORE> Message of 3-Oct-80 05:57:13
Date: 3 Oct 1980 0557-PDT
From: The Mailer <OPERATOR at SU-SCORE>
To: CSD.BARR at SU-SCORE
Subject: Message of 3-Oct-80 05:57:13
Remailed-date: 3 Oct 1980 0609-PDT
Remailed-from: Avron Barr <csd.barr at SU-SCORE>
Remailed-to: ct at SU-AI
Message failed for the following:
tc at SU-AI: 450 I don't know anybody named tc
------------
Date: 3 Oct 1980 0553-PDT
From: Avron Barr <csd.barr at SU-SCORE>
Subject: another question
To: csd.tucci at SU-SCORE
cc: tc at SU-AI, csd.dkanerva at SU-SCORE
Chris--Another font query: Is there an alphatype font that is both
fixed-with (like GACHA or FIX or TTY) and also has the full
SAIL character set? How do I find out what fonts are available
in general?
.Av
-------
-------
∂03-Oct-80 0857 REG Bug in EFTP
in a Receive command, ascii mode, it jumped to 150320 without having
transferred anything, as soon as I started EFTP on the Alto.
∂03-Oct-80 0912 REG WAITS problem / EFTP
To: admin.mrc at SU-SCORE, CT at SU-AI
Mark:
I have just seen the ACs and job data area of EFTP wiped out.
It looks like there's Ethernet data spread throughout. I think you
just allowed a user mode transfer to start at zero.
Moreover, I don't think the transfer specified zero as a starting location,
so I don't think Chris is doing anything wrong here.
∂03-Oct-80 0921 REG WAITS/ EFTP
To: admin.mrc at SU-SCORE, CT at SU-AI
I was wrong when I implied that the IN UUO clobbered the protected
part of the job data area. It didn't. But it quite happily clobbered
rest.
∂03-Oct-80 0930 VRP via Dialnet ethernet
Well, the Ill mem ref at user 1337 also happens when sending files from
Shasta (as opposed to Diablo) to Sail. Has anything in the code changed
in the last week?
∂03-Oct-80 1343 WIN via SU-SCORE EFTP gets memory errors
To: ME, CT
Something was done to EFTP around thursday afternoon that makes it
no longer work. It sits around for a while and then prints ill mem ref
or something like that. Could we at least go back to a previous version
or something for a quick fix?
Thanks, Bill
∂03-Oct-80 1534 Brian K. Reid <CSL.BKR at SU-SCORE> Fonts.Widths file on SAIL
Date: 3 Oct 1980 1529-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
Subject: Fonts.Widths file on SAIL
To: win at SU-AI, ct at SU-AI, ltp at SU-AI
Where is the Dover Fonts.Widths file kept on SAIL?
-------
∂03-Oct-80 1539 AVB dover tabs
Chris: The Tabs now work most of the time but not always yet.
Printing P.DOC[1,AVB] on the dover leads to excess indentation
of the two header lines in the document by about 4 fixed width spaces,
and also the Dollar columns in the lower portion of the document
are flushed too far right (at least some of them).
Could it be that the fixed width space is still wrong?
∂03-Oct-80 1649 GFS dover losing
I tried to Dover my file, regdef.doc[mei,gfs], and lost, both in hand-holding
mode and a one line command.
Improper number of records listed in document directory (Dir)
∂03-Oct-80 2341 VRP via Dialnet eftp revived
Fantastic, it works again. What's going on?
CT -- Something was wrong in the PUP server, causing all sorts of havoc.
Note that I did not change the program on sys: at all in the last
two weeks or so.
∂04-Oct-80 1219 Brian K. Reid <CSL.BKR at SU-SCORE> why .PRE
Date: 4 Oct 1980 1219-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
Subject: why .PRE
To: CT at SU-AI
This is a trivium, but I thought I'd explain why CMU has used .PRE
for the past 2 years as the extension for press files. The idea was
to minimize the amount of tomfoolery needed for moving files around
from one place to another. There is an automatic file motion server
that can move files from, say, IFS or a Twenex machine to any of the
others, including the Tops-10. When file names are moved, the
convention is always just to truncate the file names, so that PRESS
becomes PRE, Bravo becomes BRA, and so on. While PRS may be more
mnemonic, PRE is more compatible.
-------
∂04-Oct-80 1219 VRP via Dialnet pup server
The rotten cur! Some sonofabitch been hacking the pup server, I wonder?
Anyway, I'm really glad it's back.
CT -- Mark Crispin has been trying to get RTP/BSP and buffer-mode I/O going.
∂04-Oct-80 1721 ME DOVER
DOVER no longer seems to work. I get only the title page out,
and it says "Unknown Error". Earlier I got "Illegal Press file 1064: Improper
number of records listed in document directory (DIR)". This was from just
trying to print a simple text file.
Also, DOVER should use 19 disk input buffer and 19 disk output buffers
for optimum disk throughput.
∂05-Oct-80 2152 MRC
EFTP monitor command installed
∂06-Oct-80 1412 ME eftp question
To: MRC at SU-AI, CT at SU-AI
∂06-Oct-80 0926 FJM eftp broken
To: ROB, ME
EFTP gives an ill UUO message preceded by file lookup problems. Is the
system sick, or was EFTP changed over the weekend? Please advise. - Jack
∂06-Oct-80 1456 MOG eftp broken
when I try sending an ascii-mode file to host #300 (although
I don't thik that it matters), I get an
"Illegal UUO (76040000607) at user 1634"
-Jeff
∂06-Oct-80 1720 CSD.GARDNER at SU-SCORE Dover problem
Date: 6 Oct 1980 1714-PDT
From: CSD.GARDNER at SU-SCORE
Subject: Dover problem
To: ct at SU-AI
I just tried twice to print a file from Sail, was told by the system that
it worked, and then found only a title page with the message:
[602] Illegal press file 1067: Bad "password". Sent from host 0# 302#.
What's the problem? --Anne Gardner
-------
∂06-Oct-80 1837 Keith A. Lantz <CSL.LANTZ at SU-SCORE> eftp and dover
Date: 6 Oct 1980 1836-PDT
From: Keith A. Lantz <CSL.LANTZ at SU-SCORE>
Subject: eftp and dover
To: ct at SU-AI
cc: csl.lantz at SU-SCORE
1) DOVER no longer seems to work with TTY output. I received the
following error (but had to delete the file to make space for more...):
[603]illegal press file 1071: improper number of records listed in
document directory (Dir)
2) The manner in which EFTP has been changing out from under users
is unacceptable. One day it works; the next day it works differently
and slower; the next day it doesn't work at all. As with any program
that some courageous souls would like to use, you should give your
experimental versions different names -- e.g. XEFTP, or whatever.
The main reason I need it at the moment relates to (1) above -- if I
can't DOVER my output from SAIL, I have to do it from an Alto. If
I can't get the file to an Alto, I get annoyed. If you want some help
with EFTP perhaps you should drop by and talk to Brian Reid and/or I;
this is one very small thing we weren't exactly hired for, but it is
becoming a real nuisance.
Keith
-------
CT -- The fact is that I have not changed EFTP for over 3 weeks (on
sys). The Pup Server is now being modified (not by me) so that we
can get an FTP going. This is the reason for all the changes in
behaviour of the programs. When the Pup Server is finally done, all
will be well (?) again. Sorry for the inconvenience, but I am not
really to blame.
Chris.
∂06-Oct-80 1847 KAL via SU-SCORE eftp cont.
The problem with DOVER is probably one with EFTP, yes? An even
better reason not to screw around with a working (sic)
product.
Tangential point: I have lots of experience building decent user interfaces for programs
running under TOPS-10. Most of my stuff is in SAIL and includes file
iterators, etc. It should be transportable to WAITS with minor
effort and would make programs like DOVER and EFTP much
nicer to bear with. I hate WAITS and TOPS-10, but am willing
to help make them a little more bearable.
Keith
∂07-Oct-80 1100 LTP Sink
Is it ready? Please confirm that to MAry, at 256MJH. Thanx, Luis
∂07-Oct-80 1430 GFS illegal press file
I lost trying to DOVER my E file "REGDEF.DOC[MEI,GFS]".
The error message was
[603]Illegal Press File 1075:Improper number of records listed in document
directory (Dir)
I anxiously await your reply.
CT -- there has been a rash of this lately. I will check up on it this
afternoon.
∂07-Oct-80 1615 FJM broken EFTP
Chris - EFTP broke over the weekend. It
types "LOOK-UP failed ..." before the file name is typed in (bug
in your rescanning code, probably), and then completes the
dialog but produces an ill UUO (76040000607) at user 1634. This happens
reproducibly when I try to ship bribe.ima from [1,fjm] on SAIL to
alto #105 in image mode, Left bits. It didn't happen Saturday;
then the problem was checksum errors (bad PUP) occurring early or later in
the transmission and causing EFTP to give up. I tried 10 times or so and
never made it to the end of the file. Or does EFTP always cough on the
last PUP? Please fix the ill uuo bug asap. Also, if there's any way you
can enhance EFTP to resend bad PUPs (I have no idea what the protocol
is), you will be doing an immense service to those of us who
need to move long files to the ifs. I've been waiting several days
now to move this one. Please acknowledge this message. Thanks. - Jack
P.S. I'm a new Research Affiliate in HPP, in case you're wondering who I
am. I just started work 9/24/80.
∂07-Oct-80 2238 VRP via Dialnet eftp
More peculiar behavior: after I have transferred one file, it acts
like I typed everything over again and starts trying to receive a
second file.
∂08-Oct-80 0327 MRC
EFTP.FAI[1,CT] does not generate a working version of EFTP. Please
look into it. I fixed a couple of bugs causing it to illegal UUO, but
now it complains about bad checksums.
∂08-Oct-80 0642 Avron Barr <csd.barr at SU-SCORE> font queston
Date: 8 Oct 1980 0637-PDT
From: Avron Barr <csd.barr at SU-SCORE>
Subject: font queston
To: csd.tucci at SU-SCORE
cc: ct at SU-AI
Chris,
I noticed that font DMR7 only has a .TFX file, and not
a .VNT file--is there some reason that I shouldn't use that font?
.Av
-------
∂08-Oct-80 0953 Avron Barr <csd.barr at SU-SCORE> Alphatype work
Date: 8 Oct 1980 0951-PDT
From: Avron Barr <csd.barr at SU-SCORE>
Subject: Alphatype work
To: ct at SU-AI
cc: csd.tucci at SU-SCORE, csd.dkanerva at SU-SCORE, csd.drf at SU-SCORE
Chris,
I would like to get two Alphatype output sheets (12 pages)
from a DVI file I7ve created. Will you have any time today to
do them? I want them to show to my publisher who is going to
stop by tomorrow morning. I think that the file contains only
fonts that are available, but we7ll have to check.
.Av
-------
∂08-Oct-80 1231 ME EFTP trouble
To: BUG-EFTP
∂07-Oct-80 2246 BVM EFTP lossage
To: BUG-eftp
It craps out with Illeg UUO (76040000607) at user 1634,
apparently independent of mode. Also, during the interaction it
complains that LOOKUP failed even before I get a chance to type the
filename
EFTP.OLD still seems to work, though.
∂08-Oct-80 1703 Avron Barr <csd.barr at SU-SCORE> Alphatype work
Date: 8 Oct 1980 1658-PDT
From: Avron Barr <csd.barr at SU-SCORE>
Subject: Alphatype work
To: csd.tucci at SU-SCORE
cc: ct at SU-AI, csd.dkanerva at SU-SCORE
Chris,
Just in case you show up tonight, here is the scoop on the
files I would like a sample from. They are both on <CSD.HANDBOOK>
at SCORE. The preferred one is
SEARCH-DMR.DVI
but it uses the DMR7 font which may not be available. If it's not
then
SEARCHN.DVI
is the same file without that font. As I mentioned in a previous
note, I would like only two Alphatype sheets (12 text pages) to
show to my publisher tomorrow. If you get time to do it, and
something goes wrong or you need my help, I will be at one of the
following numbers tonight: 328-3277 (early) and 322-2233 (late).
Thanks.
.Av
P.S. you never got back to me about converting that Graphics font?
-------
∂08-Oct-80 1721 JMC loop or something
To: BUG-DOVER
DOVER FOLPRM.REF[AIM,DOC] ran for 11 minutes and had produced a
monstrous number of records when I killed it on the advice of ME.
The file is about 22K.
∂08-Oct-80 2132 KAL via SU-SCORE dover
I hear persistent rumors that dover really does work with press
files. Hence, why not with normal text files (as documented)?
But then, EFTP doesn't really work so how can dover work? In any case
who can I single out for ethernet bungles -- Mark Crispin?
∂09-Oct-80 1135 ROB forwarded from GRIPES (BUG-eftp)
∂07-Oct-80 2246 BVM EFTP lossage
To: BUG-eftp
It craps out with Illeg UUO (76040000607) at user 1634,
apparently independent of mode. Also, during the interaction it
complains that LOOKUP failed even before I get a chance to type the
filename
EFTP.OLD still seems to work, though.
∂09-Oct-80 1337 PAM
you loser! I tried to Dover terms.mit[up,doc] and it wrote
30K sectors befoore I deleted the file and blew the core image away...
Que Pasa?
∂11-Oct-80 2108 JED EFTP hassles
Hi. Some problems occurred while I was using EFTP tonight.
First, my Alto seemed to have a variable host number. That is, sometimes
it said 0#111#, and sometimes it said 52#111#. What does this number
mean, anyway? What problems does it cause if it's non-zero? Why should
it change (seemingly randomly) during the course of a terminal session?
Anyway, I tried sending a file to the Alto anyway, with the following
results:
Bye
↑C
.r eftp
*send
Local file name:lewis.73a
Mode: aSCII mode
Convert CRLF to CR (normal to send to Altos)?y
Host number (octal):111
.!.!.!.!.!.!.!.!.!..!.?Received bad pup 4 from 0#111#
Error: Bad Checksum
;;right here the file had finished, but the last PUP had a bad checksum.
Exit
↑C
..r eftp ; try again
*send
Local file name:lewis.73a
Mode: aSCII mode
Convert CRLF to CR (normal to send to Altos)?y
Host number (octal):111
.?Received bad pup 4 from 0#111#
Error: No such port
;;right here the Alto host was not started first.
Exit
↑C
But, when I went to the Alto, I found that the file had in fact
arrived there, correctly. I'm not sure whether it was sent on the first
or the second attempt at EFTP.
Jim
∂12-Oct-80 1725 ME looping DOVER program
When I try to DOVER the file RFC759.TXT, the DOVER program gets to the end
of the input .TXT file and then loops indefinitely writing (all zero?) data
in the .PRS file, making it longer and longer without end.
∂12-Oct-80 1834 DEK alphtyping
The log book in darkroom shows no entries since August. Hasn't anybody
even changed the water since then? Such things belong in the log book.
But from the looks of the water, it might not have been changed!
I brought in two buckets that are "clean" (except I didn't take time
to clean the buckets themselves...)
I had trouble with newda. First it apparently now puts 4/page instead
of 6, so I ruined my output with things overprinted. Then it
somehow got a JSYS error and the next time I ran it, its accounting
file seems to have disappeared. (One of the times I ran it, I forgot
to type "res" first. But before the JSYS error, I got into a new section,
where it asked me if I wanted to see the list of accounts. I said
no, then it asked me for my name again, then the JSYS error happened.
DRF changed the default, I think, for Rich Pattis's book, and probably
forgot to change it back.
∂12-Oct-80 1900 WIN Dover still does not work
To: MRC at SU-AI, CT at SU-AI
I have given up trying to use Dover. What happened to it???
I seemed to be able to eftp a file to host 201 just fine, however.
More important: You have not implemented the PUP protocols
correctly! Several times I've had eftp bomb out with the message
"Bad PUP received". According to the protocols, you should IGNORE
all PUPS with bad checksums, out-of-sequence byte IDs, wrong socket
numbers, etc. This makes the software much more robust. Just timeout
and retransmit after a while.
Is there any chance of getting real ftp and telnet service? I seem
to remember someone claiming it would take "a few weeks" after working
Ethernet hardware was installed to get ftp and telnet working.
It's been six months.
-- Bill
∂13-Oct-80 0030 Mark Crispin <Admin.MRC at SU-SCORE> Re: Dover still does not work
Date: 13 Oct 1980 0028-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
Subject: Re: Dover still does not work
To: WIN at SU-AI, MRC at SU-AI, CT at SU-AI
In-Reply-To: Your message of 12-Oct-80 1900-PDT
Bill -
Chris Tucci is responsible for both the DOVER and the EFTP programs. You
should talk with him about getting them working. Your assistance would
be of far greater utility than just your complaints. Remember, the code
on WAITS is being done from scratch, unlike the efforts on the VAXen.
The delay with full Ethernet service on WAITS has mostly been due to
delays on my part. I would rather not elaborate, but I have had a lot of
problems in my personal life recently. The past 1 1/2 months have been a
very trying time for me; other people with more sense would have had a
nervous breakdown by now and be done with it. It has been enough of an
effort for me just to continue showing up and keeping SCORE alive and
happy. Every time I get back on my feet to the point where I can start
working on this again, something else has gone wrong; most recently on
Friday night.
All of this has been external to Stanford, but it affects my state of mind
when I'm on the job and necessarily affects my rate of progress. Ralph
knows the details of what's wrong and isn't hassling me about it; I suggest
you do the same. The fact remains that there still IS some progress being
made; just not at my normal rate. Once some of the terrific external
pressures I am facing are lifted I am hoping to get back to what should be
my normal performance.
I would prefer that the contents of this message or any of its subject
matter not be passed on to any third party. I felt though that I had to
answer your comment, inappropriate and unfair as it was.
-- Mark --
-------
∂13-Oct-80 1059 VRP eftp
I tried sending a file to 104 before 104 was ready to receive it,
but eftp went ahead with it. It typed 14 periods, with no !'s, then
said ?Received bad pup 4 from 0#104#.
Error: No such port. What's it all mean?
CT -- oh, goodie. This is an easy one. "No such port" is the Alto's
way of saying "I am not sure we are at the beginning of the file -- I
may have missed something." This apparently is a bug in the Alto version
of EFTP, in other words, it should probably check the i.d. number of the
PUP but it does not. It just sends out that VERY LUCID AND CLEAR error
message!! (no such port? I am still trying to figure out why they chose it...)
∂13-Oct-80 1102 VRP eftp
Whoops, forgot, receiver is done first. Sorry.
∂13-Oct-80 1128 VRP eftp
Help, I get ?Received bad pup 33 from 0#104#, what does it mean?
CT -- 33 is the abort PUP. This probably means that your Alto disk
was full or some other strange reason prompted the Alto EFTP to send
an ABORT.
∂13-Oct-80 1539 ME DOVER
Besides my one RFC759 file that loops forever in the DOVER program,
I have 4 other RFC7xx files that can't be DOVERed. They consistently
get errors like:
Illegal Press file 1642: Improper number of records listed in document directory
Illegal Press file 1637: Bad "password".
Also, the current default (and only available) font doesn't have all of the
SAIL character set. It would be very nice if the default font did have all
the SAIL characters.
And the date printed on the title page when DOVERing from SAIL has the
day of the month wrong.
And finally, I would like to get some responses on these questions so that
I know what is going on. Thanks.
CT -- for some reason (still unknown to me, the DOVER program gets caught
in some loop. That much I know. The only possible loop in the whole
program is where it keeps reading characters until EOF. So, for some reason,
EOF is not being detected. Note that this loop only happens on text files
of around 55 pages or more.
Your other problems may stem from the fact that the files are not E files
or that they are longer than usual. I know that this is really disgusting,
but I have not had a real opportunity since school started to actually
sit down and take the code apart to find out what is causing all of these
errors. As soon as I get "adjusted," I hope to finish this once and for all.
I have heard tell that there is a font on the Dover that does have the
full character set. I will try to find it, if it exists, I will make it
the default.
OK, here is the story on the day of the month: last month, when I figured
out the date, I followed the procedure in the UUO manual, and the day of
the month was not right (I think it was off by 3 days or something.)
So I compensated for it. I will change this also.
∂14-Oct-80 0133 MDP Bad "password"
To: BUG-dover at SU-AI, CT at SU-AI, OR.Stone at SU-SCORE
The DOVER program failed on an ordinary text file I ftp'd over from score in
ascii mode. At SAIL it went
.dover
Local file name:ibm.txt
.!.!.!.!.!.!.!.!.!
Dover accepted file. Please pick up your output.
The Dover was less encouraging...
Problems encountered:
[602] Illegal Press file 1341: Bad "password". Sent from host 0#302#.
When I put in an E directory, the Dover bought it. I have had no similar
problems with other files, short and long, with and without directory pages.
CT -- the DOVER program does not like non-E files very much. I am not sure
why yet. Hopefully I will pretty soon!
∂14-Oct-80 1102 REG notice is short, but this is only a reminder:
To: "@STAFF.DIS[CSD,REG]" at SU-AI
staff meeting today, 1:30, at 252.
∂14-Oct-80 1300 ME DOVER and E dir
How come E directories have anything to do with the DOVER losing? Does
the DOVER program use the E directory for some purpose? If so, for what,
and how does it work with non-E files? Perhaps you are using the directory
to calculate the length of the file instead of using the info from the
LOOKUP??
Or perhaps the problem is that you lose if the file ends in the middle of
a disk record (which E files are guaranteed not to do, whether or not
they have directories). There is no promise that disk files will have a
length that is a multiple of 128 words (the record size is 128 words).
Most files in fact won't, except for E files.
CT -- I think it has something to do with this last point. For instance,
if we take a run-of-the-mill non-E file, sometimes it will work and sometimes
it will not. I have n
∂14-Oct-80 1741 AVB
could you convert .PRS to .PRE ??
∂14-Oct-80 1734 CSD.NOWICKI at SU-SCORE In the SAIL tradition
Date: 14 Oct 1980 1727-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: In the SAIL tradition
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List: ;
I just found out that the powers that be at Sail have decreed that press
files shall have extensions of .PRS. In fact, the DOVER program will
send a file that ends in .PRS directly to the Dover, even if it is not a valid
PRESS file, and if it is a valid press file, but has a different extension
it will generate reems of blank paper. This is an enormous Kludge, with
a capitol K. The least we could do is use .PRE, like CMU has done for
almost a year, the only other site I know of with a losing file naming system.
I routinly FTP 3-4 files a day from either CMU, MIT, or MAXC (note the Sail
FTP automatically truncates .PRESS to .PRE), and would like to be able to
print them without renaming them first!
PLEASE, change this before people get to used to it, and we become
incompatible with everybody else in the world on yet another point.
Also Dover should do a litle testing first to make sure the file
what it thinks it is (text or press). Don't print a file that generates
several pages with no printing characters on them, and ship files that
have a correct password.
-- Bill
-------
∂14-Oct-80 1743 CSD.NOWICKI at SU-SCORE [CSD.NOWICKI at SU-SCORE: In the SAIL tradition]
Date: 14 Oct 1980 1738-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: [CSD.NOWICKI at SU-SCORE: In the SAIL tradition]
To: ct at SU-AI
I hope these flames are hot enough.
---------------
Mail-from: ARPANET site SU-AI rcvd at 14-Oct-80 1734-PDT
Date: 14 Oct 1980 1727-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: In the SAIL tradition
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List: ;
I just found out that the powers that be at Sail have decreed that press
files shall have extensions of .PRS. In fact, the DOVER program will
send a file that ends in .PRS directly to the Dover, even if it is not a valid
PRESS file, and if it is a valid press file, but has a different extension
it will generate reems of blank paper. This is an enormous Kludge, with
a capitol K. The least we could do is use .PRE, like CMU has done for
almost a year, the only other site I know of with a losing file naming system.
I routinly FTP 3-4 files a day from either CMU, MIT, or MAXC (note the Sail
FTP automatically truncates .PRESS to .PRE), and would like to be able to
print them without renaming them first!
PLEASE, change this before people get to used to it, and we become
incompatible with everybody else in the world on yet another point.
Also Dover should do a litle testing first to make sure the file
what it thinks it is (text or press). Don't print a file that generates
several pages with no printing characters on them, and ship files that
have a correct password.
-- Bill
-------
---------------
-------
∂15-Oct-80 1146 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul) Bill Nowicki's comments on press files
Date: 15 Oct 1980 1142-PDT
From: CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
Subject: Bill Nowicki's comments on press files
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List:
cc: ct at SU-AI
(1) .PRE is certainly a better choice than .PRS; this is not all
that important, but why choose the lesser of two?
(2) There is no reason why DOVER (on SAIL) cannot check the
"press file password" to determine the press-hood of a file;
it takes virtually no time if random access to the file is
available, and it will work right something like 32767/32768ths
of the time. Empress on the Altos, "print" on the IFS, and
CZ on Unix all do this without any difficulty (in fact, CZ
can tell if it's got a text file, press file, or byte-swapped
press file!). If the maintainer(s) of DOVER will speak up, I
can tell them how to do this.
-Jeff
-------
CT -- I am "speaking up"...I will be in to see you this afternoon.
∂15-Oct-80 1200 Baskett at PARC-MAXC Press files
Date: 15 Oct 1980 11:58 PDT
From: Baskett at PARC-MAXC
Subject: Press files
To: REG@Sail,ME@Sail,MRC@Sail,CT@Sail
cc:
Please forward this (hopefully helpful) information to whoever.
Forest
------------------------------------------------------------
Mail-from: Arpanet host SU-AI rcvd at 15-OCT-80 1147-PDT
Date: 15 Oct 1980 1142-PDT
From: CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
Subject: Bill Nowicki's comments on press files
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List:
cc: ct at SU-AI
(1) .PRE is certainly a better choice than .PRS; this is not all
that important, but why choose the lesser of two?
(2) There is no reason why DOVER (on SAIL) cannot check the
"press file password" to determine the press-hood of a file;
it takes virtually no time if random access to the file is
available, and it will work right something like 32767/32768ths
of the time. Empress on the Altos, "print" on the IFS, and
CZ on Unix all do this without any difficulty (in fact, CZ
can tell if it's got a text file, press file, or byte-swapped
press file!). If the maintainer(s) of DOVER will speak up, I
can tell them how to do this.
-Jeff
-------
------------------------------------------------------------
∂15-Oct-80 1201 Baskett at PARC-MAXC Press files and the Dover spooler
Date: 15 Oct 1980 12:00 PDT
From: Baskett at PARC-MAXC
Subject: Press files and the Dover spooler
To: CT@Sail
cc:
Apropos of the previous message. -Forest
------------------------------------------------------------
Mail-from: Arpanet host SU-AI rcvd at 14-OCT-80 1735-PDT
Date: 14 Oct 1980 1727-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: In the SAIL tradition
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List: ;
I just found out that the powers that be at Sail have decreed that press
files shall have extensions of .PRS. In fact, the DOVER program will
send a file that ends in .PRS directly to the Dover, even if it is not a valid
PRESS file, and if it is a valid press file, but has a different extension
it will generate reems of blank paper. This is an enormous Kludge, with
a capitol K. The least we could do is use .PRE, like CMU has done for
almost a year, the only other site I know of with a losing file naming system.
I routinly FTP 3-4 files a day from either CMU, MIT, or MAXC (note the Sail
FTP automatically truncates .PRESS to .PRE), and would like to be able to
print them without renaming them first!
PLEASE, change this before people get to used to it, and we become
incompatible with everybody else in the world on yet another point.
Also Dover should do a litle testing first to make sure the file
what it thinks it is (text or press). Don't print a file that generates
several pages with no printing characters on them, and ship files that
have a correct password.
-- Bill
-------
------------------------------------------------------------
∂15-Oct-80 1205 CSD.GARDNER at SU-SCORE New problem with Dover
Date: 15 Oct 1980 1204-PDT
From: CSD.GARDNER at SU-SCORE
Subject: New problem with Dover
To: ct at SU-AI
Yesterday, on two different files, the Dover program said it accepted my
file, then produced a suitable number of entirely blank pages. PLEASE
tell me what's going on.
Anne Gardner
-------
CT -- sounds as if it thinks your file is a press file already, or vice
versa. Could you tell me the name of the files and let me take a quick
look at them? I might be able to find out what is wrong quicker that way.
∂15-Oct-80 1301 ME DOVER
I completely agree with the comments on DOVER by Jeff Mogul and Bill Nowicki
regarding the use of .PRE instead of .PRS and checking of the input file
for what format it is in instead of just assuming that the format is implicit
from the extension. Also, you should try to get some of the worst bugs out
of DOVER as quickly as possible, since many people are already using it,
or trying to. When you think you have fixed some, please tell me so I can
try various files again. I will be glad to help if you have any trouble.
CT -- I will try to finish up that debugging tomorrow afternoon on that one
annoying problem.
∂15-Oct-80 2037 REG
CT: I think this warrants your serious attention.
∂14-Oct-80 1734 CSD.NOWICKI at SU-SCORE In the SAIL tradition
Date: 14 Oct 1980 1727-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: In the SAIL tradition
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List: ;
I just found out that the powers that be at Sail have decreed that press
files shall have extensions of .PRS. In fact, the DOVER program will
send a file that ends in .PRS directly to the Dover, even if it is not a valid
PRESS file, and if it is a valid press file, but has a different extension
it will generate reems of blank paper. This is an enormous Kludge, with
a capitol K. The least we could do is use .PRE, like CMU has done for
almost a year, the only other site I know of with a losing file naming system.
I routinly FTP 3-4 files a day from either CMU, MIT, or MAXC (note the Sail
FTP automatically truncates .PRESS to .PRE), and would like to be able to
print them without renaming them first!
PLEASE, change this before people get to used to it, and we become
incompatible with everybody else in the world on yet another point.
Also Dover should do a litle testing first to make sure the file
what it thinks it is (text or press). Don't print a file that generates
several pages with no printing characters on them, and ship files that
have a correct password.
-- Bill
-------
∂15-Oct-80 2057 REG
∂15-Oct-80 1146 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul) Bill Nowicki's comments on press files
Date: 15 Oct 1980 1142-PDT
From: CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
Subject: Bill Nowicki's comments on press files
To: <CSL.SUN>SUN.DIS = Stanford University Network Meeting Mailing List:
cc: ct at SU-AI
(1) .PRE is certainly a better choice than .PRS; this is not all
that important, but why choose the lesser of two?
(2) There is no reason why DOVER (on SAIL) cannot check the
"press file password" to determine the press-hood of a file;
it takes virtually no time if random access to the file is
available, and it will work right something like 32767/32768ths
of the time. Empress on the Altos, "print" on the IFS, and
CZ on Unix all do this without any difficulty (in fact, CZ
can tell if it's got a text file, press file, or byte-swapped
press file!). If the maintainer(s) of DOVER will speak up, I
can tell them how to do this.
-Jeff
-------
∂15-Oct-80 2145 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul) People and Software
Date: 15 Oct 1980 2139-PDT
From: CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
Subject: People and Software
To: fb at SU-SCORE, mrc at SU-SCORE, csd.nowicki at SU-SCORE, ct at SU-AI,
vrp at SU-AI, lantz at SU-SCORE
In-Reply-To: Your message of 15-Oct-80 1807-PDT
My comments on the MRC/WIN/CT/SUN/me mess:
(1) It is an oversight on somebody's part that MRC was
not on the SUN distribution list ... my guess is that VRP, being
a newcomer, is not yet aware of all of the people around here.
I believe that the tradition at MIT of announcing the formation
of a mailing list, and the address of its maintainer, on
*MSGS (== BBOARD) is what should have been done ... then, people
are responsible for getting themselves onto lists.
I will suggest to VRP (I suspect he is maintaining the SUN list)
that he should make such an announcement, but I will leave the
decision to him (so that he may avoid being swamped with requests.)
(2) Re: software improvement requests; I believe that a system
designer/implementer should always accept (and usually comment
upon) non-abusive requests for improvements, or bug reports.
Additionally, the rest of the world should be kept moderately
informed about such reports, to avoid duplication and
stewing in the dark. On the other hand, I think that abuse
and invective (e.g., "stupid", "ill-designed", "cretinous
hack", etc.) are only valid when directed at an unknown
group of people who would otherwise not pay much attention.
In short, only use insulting language when there is no doubt
but that the insulter does not know who the insultee is.
I think we have a general problem here with software
which is not well-designed and well-built. It is hard
to know when and how to criticize mistakes, since some people
appreciate honest criticism (I include myself, even if I may
defend my position excessively), while some people take even
mild criticism as unspeakable insult and refuse to deal with
the problem. Although from time to time, throw-away software
may be useful, it certainly should not become our main product,
and I do not believe that a long-term project should be immune
to criticism ... we may get stuck with something.
Sorry for the flame ... I did not send this to the entire
distribution list, since I'm not sure who is on it, and
I don't think that everyone is interested in our problems.
-Jeff
-------
∂16-Oct-80 0037 JLS TEX Press Output at SAIL
To: CT at SU-AI, LTP at SU-AI, DEK at SU-AI,
guibas at PARC-MAXC, RAS at SU-AI, DRF at SU-AI, IAZ at SU-AI
CC: JLS at SU-AI
Is any one working in producing TEX press output at SAIL, So that
one can use the DOVER?
Juan
PS: I attempted to so using TEXPRS[TEX,DEK] but get DOVER
errors (Bad Password, or some such)..
∂16-Oct-80 0942 JLS Dover Bug?
16-OCT-80 0936
Filnam Ext PPN Size Written Time Pro Writer Reference % Dumped
RFC760 TXT 1 ME 16.4 9-JAN-80 2034 000 1 ME FTP 14-OCT-80 1 P1634>
RFC759 TXT 1 ME 30.2 23-AUG-80 1151 000 1 ME E 14-OCT-80 7 P1770>
RFC772 TXT 1 ME 12.2 1-OCT-80 1303 000 1 ME FTP 14-OCT-80 13 P1798
RFC771 TXT 1 ME 3.7 1-OCT-80 1304 000 1 ME FTP 14-OCT-80 26 P1798
RFC770 TXT 1 ME 5.3 1-OCT-80 1304 000 1 ME FTP 14-OCT-80 26 P1798
TOTAL= 68.0
#↑C
↑C
.dover rfc760.txt[1,me]
? Error in job 41
Ill mem ref at user 2100
↑C
.dover rfc760.txt[1,me]
? Error in job 41
Ill mem ref at user 2100
↑C
.dover rfc759.txt[1,me]
? Error in job 41
Ill mem ref at user 2100
↑C
.r ppsav
∂16-Oct-80 1019 JED .PRE files
What was wrong with .PRS? It's a lot more consistent with other
abbreviations (e.g., .LSP, .TXT) which are formed by dropping vowels.
∂16-Oct-80 1151 Fagan at SUMEX-AIM dover
Date: 16 Oct 1980 1150-PDT
From: Fagan at SUMEX-AIM
Subject: dover
To: ct at SAIL
Thanks for letting me know. I was wondering what happened...
Larry
-------
∂16-Oct-80 1502 Bob Moore <BMOORE at SRI-KL> dover
Date: 16 Oct 1980 1449-PDT
From: Bob Moore <BMOORE at SRI-KL>
Subject: dover
To: ct at SU-AI
cc: bmoore at SRI-KL
I would like a rough estimate of when you think it will be possible to
print XGP files on the Dover. I have just completed a large report using
PUB and the XGP, and naturally, I would like the published version to
look as nice as possible.
I hope you will interpret this as a simple request for information and not
as a complaint.
--Bob Moore
-------
CT -- hi. Right now we have an XGP-DVI program, and we have a little-used
(mainly because it is on Unix) DVI-PRESS converter, which I hope to transfer
to SAIL and start up (of course, the only use of this program is the algorithm
because SAIL does not support C.) So my estimate is about a month or so at
least. I don't work here full time anymore because of school, but will try
to get going on this when I get ALL the bugs out of DOVER (there are many!)
∂16-Oct-80 1608 CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
Date: 16 Oct 1980 1229-PDT
From: CSL.JLH.MOGUL at SU-SCORE (Jeffrey Mogul)
To: CT at SU-AI
In-Reply-To: Your message of 16-Oct-80 0837-PDT
I should make it clear that I cannot tell you how to random access
on Sail (ask MRC or another Sail wizard how to do that) ... I
can only tell you what the "password" should be and where it
should be.
-Jeff
-------
CT -- I found out, thanks.
∂16-Oct-80 1622 CSD.NOWICKI at SU-SCORE [Ralph Gorin <REG at SU-AI>:]
Date: 16 Oct 1980 1042-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: [Ralph Gorin <REG at SU-AI>:]
To: ct at SU-AI
Chalk one up to reasonableness. Please change the extension for press
files to .pre. I found out another good reason. Brian Reid's scribe
program that runs on score generates .PRESS files, which, when FTPed
get an extension of .PRE.
-- Bill
---------------
Mail-from: ARPANET site SU-AI rcvd at 15-Oct-80 2037-PDT
Date: 15 Oct 1980 2038-PDT
From: Ralph Gorin <REG at SU-AI>
To: csd.nowicki at SU-SCORE
"reams" and "capital" are misspelled.
You have a strange idea of who the powers that be at SAIL are, and what
they can decree. I happen to agree with your idea. Now what?
---------------
-------
∂16-Oct-80 1946 CSD.GARDNER at SU-SCORE
Date: 16 Oct 1980 1125-PDT
From: CSD.GARDNER at SU-SCORE
To: CT at SU-AI
In-Reply-To: Your message of 15-Oct-80 1335-PDT
The two files that gave me blank output were FREFS.TXT and F6C3.TXT on
[1,AVG]. The FREFS.PRS file is still there too, if you want to look at
it.
Also: Am I correct in thinking a file must be formatted by E for Dover to
work? Doing this seemed to cure the problem I had earlier; but it's an
extra step since I normally use Emacs.
Thanks-- Anne
-------
CT -- try them again now. I did a lot of debugging yesterday so hopefully
most files should work. If they still don't please send me another message.
∂17-Oct-80 1131 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover.
Date: 17 Oct 1980 1127-PDT
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Address: 26970 Beaver Lane, Los Altos Hills, CA 94022
Phone: (415) 494-0860
Subject: dover.
To: ct at SU-AI
I just now sent a file (non press) to the dover via your dover program.
I got no sign at all of any output, however. This despite a "Dover
accepted file" message. Odd....
-------
∂17-Oct-80 1243 MSM
.dover
Local file name:flshdt.dat
.!.!.!.!.!.!.!.!.!.!.!.!.!
Dover accepted file. Please pick up your output.
Exit
↑C
I tried to print out a file using DOVER and the message on the DD
was as above. The output was only the cover page and a message on it that
said
Problems encountered:
[603] Illegal Press file 1375: Improper number of records listed in document directory[Dir]
I tried it with and without the E directory page, and the same error occurs.
I was on ppn [1,MSM] at the time.
What am I doing wrong?
Shahid
CT -- chances are you are not doing anything wrong and it is simply another
of my many bugs in the program. I am going to attempt another massive de-
bugging session this weekend.
∂17-Oct-80 1455 FFL
To: CT, FFL
Your copy of ESSENTIAL E is in Room 358.
CT -- thank you!! This is the first non-depressing message I have received
in over a month!
∂17-Oct-80 1747 MRC via SU-SCORE DOVER
I have made a small change to DOVER. Please install a new version
if you havne't made any other changes to it...
∂17-Oct-80 1858 CSD.NOWICKI at SU-SCORE Dover fonts
Date: 17 Oct 1980 1857-PDT
From: CSD.NOWICKI at SU-SCORE
Subject: Dover fonts
To: me at SU-AI
cc: ct at SU-AI
I just remembered you asked a while ago about getting all
Sail characters. The font SAIL8 is available for this, and Chris
has said that will be the new default.
Mark Crispin knows of some other possibly useful fonts from MIT.
Just get AC format versions (or even OC) and they can be put in to
Spruce.Fonts with a minimum amount of effort.
-- Bill
-------
∂18-Oct-80 1710 ME DOVER
I finally got one of my five RFC files to print on the Dover, but the
other four still get one of these errors:
Bad "password"
Internal pointer exceeds file bounds
The latter error occurs with RFC759.TXT, RFC760.TXT and RFC770.TXT
and the former with RFC771.TXT, all on [1,ME].
Also, you still seem to have some trouble with the date. I bet the
problem is that you are doing some part of the date calculation with
octal numbers instead of decimal, or vice versa.
∂18-Oct-80 1905 WIN Bad pups again
I repeat: PUPs with bad checksums should be ignored.
After a usual timeout it will be retransmitted.
I keep getting:
?received bad pup 4 from 0#103#
Please fix.
∂20-Oct-80 0917 REG DOVER tabs and spacing
Tabs and spacing are rather screwed up in DOVEr.
try spooling the file DTEST[CSD,REG] for an example.
Note the third line in DTEST, the line of Y's, uses the space
character, not tab, between characters.
∂20-Oct-80 0920 REG more about TABS and spaces in DOVER.
I think I was wrong about space being wrong. Tabs are screwed up.
Also, the DIGITS are smaller than the other characters.
∂20-Oct-80 0923 REG Finally, the truth
My previous messages were both wrong. Now, I think the problem is
that the SPACE character is wider than the other characters.
∂20-Oct-80 1859 CSD.GARDNER at SU-SCORE
Date: 20 Oct 1980 1850-PDT
From: CSD.GARDNER at SU-SCORE
To: CT at SU-AI
In-Reply-To: Your message of 17-Oct-80 1525-PDT
Thanks for your message. I finally managed to get output for all my files,
but there are still a couple of problems. One is a bug: things come out
horribly unaligned, even though there are no tabs in the text and they
appear to be aligned on the screen. For instance, look at F0.TXT[1,AVG]
and then compare the Dover listing.
The other problem is the new font. The original one was fine, and I
meant to distribute copies of my paper produced on the Dover instead of
the XGP. But the tiny one looks ridiculous. How do I get back to Gacha 10
(if that's what it was)?
Anne
-------
∂21-Oct-80 0058 PAM Dover lossage
I tried unsuccessfully to DOVER a 40 page E file (with directory) today.
It got the "internal pointer exceeds bounds" message. I broke the file
into 4 pieces, and DOVERed three successfully, one unsuccessfully. I
further divided the failing file until the section that failed was only
1.5 pages of text. Dividing that failing file produced two files which
worked okay. I noticed that all the header pages for failing files have
"Stanford" printed at the top in a large font, whereas all the winning
files had "Stanford" at the top in a smaller font. Is this an artifact or
a clue??
PS files available on request for debugging..Paul Martin
PPS between each iterative attempt with successively smaller files, I
deleted *.PRE from my directory to avoid conflicting versions
problems..Paul
∂21-Oct-80 1003 DEK two things
Somebody (Juan?) sent a message asking for TEX output to Dover from SAIL.
I think we should wait with that change until Lyle Ramshaw finishes the
new Dover fonts at Xerox. I will be installing a new TEX then since he is
going to change the format of TFX files, and the new TEX will generate
PRESS files.
Another thing: Please make for me a SAMPLE of the CHIN8 font. I need only
the 128 characters, not the "text" that normally appears below; this is
for a CS report about Chinese characters that is going to be finished soon.
If you can't do it this week, please find somebody else who can, thanks.
∂21-Oct-80 1110 KDO What is the state of the Ethernet
To: CT at SU-AI, MRC at SU-AI
Since I haven't worked on anything in a long time, could one of you folks tell
me what the state of the world is? In particular, rumor has it there is an
XGP→PRESS converter existing or being written at SAIL; is that true or not?
Also, what needs to be done before we have real FTP and TELNET and nice things
like that?
Ken
∂21-Oct-80 1113 KDO DOVER suggestion
We stole the filename parsing code from the wrong place. It can't parse ∂ filehack.
That's not too bad, but it can't parse ↓ KDO↓ either, so that I have to copy
my mail file in order to dover it.
Ken
∂21-Oct-80 1146 KDO via SU-TIP DOVER big file bug
Still trying to spool my mail file, DOVER tells me Dover accepted file, but
the DOVER says "Illegal password," and evidently the whole file is screwed
up, because user name and so on are missing.
DOVERing option.txt works just fine, so I am guessing it is file size.
The files are MAIL. and MAIL.PRE (produced by DOVER, DOVERing it is the same)
Ken
∂21-Oct-80 1759 Mark Crispin <Admin.MRC at SU-SCORE> Re: What is the state of the Ethernet
Date: 21 Oct 1980 1757-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
Subject: Re: What is the state of the Ethernet
To: KDO at SU-AI, CT at SU-AI, MRC at SU-AI
In-Reply-To: Your message of 21-Oct-80 1110-PDT
I don't know anything about an XGP->PRESS converter. As for real FTP/TELNET,
it awaits my completing an implementation of BSP/RTP in the monitor.
-- Mark --
-------
∂22-Oct-80 1914 MRC DOVER
DOVER is unable to Pressify FUCK.YOU[PUP,MRC], a perfectly straightforward
text file. Please look into this. The error is "invalid password".
∂22-Oct-80 2036 Brian K. Reid <CSL.BKR at SU-SCORE>
Date: 22 Oct 1980 1630-PDT
From: Brian K. Reid <CSL.BKR at SU-SCORE>
To: Admin.MRC at SU-SCORE, CSD.Nowicki at SU-SCORE, CT at SU-AI,
Lantz at SU-SCORE
I thought I'd pass on the documentation of CMU's Dover program so you can
see what a truly superior Dover interface looks like. It is written in
Bliss, and could probably be brought up on Score, but it would need to
have its use of Tops-10 SCAN converted to the Tops-20 COMND Jsys. I realize
that some of you don't like Bliss, but for those of us who prefer getting
work done to programming in morally correct languages, it is a very fine
way to write high-quality system software that is easy to debug. And its
compiler can beat the pants off any other compiler ever written in terms
of code quality produced from a high-level language.
The essence of the CMU interface is that there is a hack in the
operating system that allows PUPs to be sent from a user program, and
then everything else, both the Dover spooler and the FTP programs, are
written in Sail or Bliss and use the PUP UUO.
By the way, an easy hack to get around the bug in SAIL10 is to use the
width of the "0" character and not the width of the space character as
the amount of horizontal white space to leave when an octal 40 code is
found in the file being pressified.
Brian
∂22-Oct-80 2042 Mark Crispin <Admin.MRC at SU-SCORE>
Date: 22 Oct 1980 2013-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
To: CSL.BKR at SU-SCORE, CSD.Nowicki at SU-SCORE, CT at SU-AI, Lantz at SU-SCORE
In-Reply-To: Your message of 22-Oct-80 1630-PDT
PS: By all means if CMU's DOVER program can be brought up at SAIL
and made to work in a reasonable fashion at SAIL it should. Do I
hear any of the BLISS experts volunteering?
-------
∂22-Oct-80 2042 Mark Crispin <Admin.MRC at SU-SCORE>
Date: 22 Oct 1980 2012-PDT
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
To: CSL.BKR at SU-SCORE, CSD.Nowicki at SU-SCORE, CT at SU-AI, Lantz at SU-SCORE
In-Reply-To: Your message of 22-Oct-80 1630-PDT
Brian -
I thought I had explained this to you, but your comment
about how you "prefer getting work done to programming in morally
correct languages" indicates you haven't. Nothing whatsoever has
been done to even start work on Ethernet software on SCORE. The
work has been all on SAIL, and the reason for the delay has
nothing to do with what programming language the software is
written in.
It is because I alone am working on the Pup device support,
and a very junior programmer (Chris Tucci) is working on the
DOVER program. I finished the necessary operating system work to
support the DOVER program at the end of August; so I feel the
delay in getting winning Dover software up is not my fault. The
delay which is my fault is that BSP support is still incomplete,
and the excuse I plead is that I didn't want to totally neglect
SCORE and I had a lot of bad things happen to me last September
which threw me for a loop. I wouldn't say that it is Chris'
fault either; there are a lot of people who know more about
things like the format of Press files and Dover software than
Chris or I do, but who have yet to volunteer their programming
assistance.
The WAITS operating system has support (I prefer to consider
it to be somewhat better than a "hack") to allow Pups to be sent
and received by a user program. That is what I finished in
August. What I'm working on now is BSP. I don't see how all
that is different than CMU, although perhaps our implementation
is striving for a bit more than CMU's is since we are planning to
use Ethernet for a lot more.
I have no objection whatsoever to bringing up CMU's Dover
program on SAIL, if it can be done. I just do not feel that new
programs at Stanford should be written in BLISS. Also, I should
note that Stanford does NOT have a BLISS-36 license, and
therefore we can NOT run the BLISS-36 compiler here. That means
we are stuck with BLISS-10, which produces incredibly cretinous
code (so much for that issue) and may not work at SAIL at all.
There is also the maintainability issue. Will CMU be
willing to support their Dover program on SAIL, and fix bugs in
it as they come up? If not, what are we to do when it breaks?
None of the system staff around here know BLISS.
Personally, I would have preferred a program written in SAIL
having the functionality of CMU's DOVER program over the DOVER
program we have now. I certainly never told Chris that he had to
write DOVER in assembly language. I have even suggested that we
could try converting CMU's DOVER program into SAIL (although
TOPS-10's SCAN is going to be a pain in the wazoo).
Finally, I don't know if I've told you this, but I've
verified that Press files which print fine elsewhere lose here
with the SAIL8 font. You'll certainly agree that @ works just
fine at CMU; after all, it was CMU who wrote the Press support in
@!
-- Mark --
-------
∂23-Oct-80 0812 Aikins at SUMEX-AIM Know anyone for this?
Date: 23 Oct 1980 0811-PDT
From: Aikins at SUMEX-AIM
Subject: Know anyone for this?
To: CT at SAIL
Hi. It turns out that the draft of the article on resolution
theorem proving that was written for the handbook is pretty bad--
according to Nilsson and Bob Moore. Now I'm looking for someone who
could write a fairly short, direct article on that subject. Do you
know of anyone who would be good?
Jan
-------
∂23-Oct-80 1312 Bill Nowicki <WIN at SU-AI> Bad Checksum
Mail-from: ARPANET site SU-AI rcvd at 23-Oct-80 1053-PDT
Date: 23 Oct 1980 1051-PDT
From: Bill Nowicki <WIN at SU-AI>
Subject: Bad Checksum
To: MRC at SU-AI
Remailed-date: 23 Oct 1980 1312-PDT
Remailed-from: Mark Crispin <Admin.MRC at SU-SCORE>
Remailed-to: CT at SU-AI
Every time I try to use eftp it bombs out with Received bad
pup from 104# Error: Bad Checksum.
I cannot do any debugging of press files until this works.
-- Bill
∂24-Oct-80 1211 LTP
∂23-Oct-80 1556 CT GFI files.
Hi. I have a GFI file on my area, CMR10.GFI; is there
any way of telling if it is o.k.?
///////////////////////////////////////////////////////
Please talk to Heather Brown, she is going to be the direct user...
She actually needs them now. Thanx, Luis
∂25-Oct-80 1435 Purger
You are exceeding your disk quota.
Files that occupy space beyond your quota are subject to purging.
If you don't delete some of your files, the purger will.
Your disk quota is: 150
Your files occupy 286
∂26-Oct-80 0923 DLB DOVER (at SAIL) problems...
I tried to dover these files, which I still have, but they both failed.
The first came back with (on Dover cover sheet):
(no file name, no username etc..., and at SAIL no press file was there afterwards.)
[602] Illegal Press file 1070: Bad "password". Sent from host 0 # 302 # .
the second came back with:
(file name ERRORS.PRE, created 10-30-1980 09:07:52, for [ 1,DLB])
[302] Illegal Press file 1071: Internal pointer exceeds file bounds.
In both cases, all I got was the cover sheets. I still have them too, if you want.
What I did at SAIL follows...
- Dan
------------------------------------------
Exit
↑C
.copy *.*←errors.tmp[sav,dlb
Exit
↑C
.dover
Local file name:lndmrk.fin
.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!
Dover accepted file. Please pick up your output.
Exit
↑C
.dover
Local file name:errors.tmp
..!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!.!
Dover accepted file. Please pick up your output.
Exit
↑C
.del *.pre
Deleted:ERRORS PRE
Total space reclaimed = 2.1 1 Blk 94% Free blocks = 1395
Exit
↑C
.undel errors.pre
Looking ...
Here's one:
ERRORS.PRE[1,DLB] 2176 wds, last written 26-OCT-80 0907 pro=55 mode=10
File deleted by 1,DLB with COPY
Is this the one you want? y
Recovering ...
File recovered: ERRORS.PRE[1,DLB]
Exit
↑C
.r ppsav
∂26-Oct-80 1343 WDG eftp
To: CT, KDO
could not send file from alto 117 to sail at 1342 10/26. folloed instructions
from READing EFTP. program ran but nothing happened. was able to send file
from sail to alto 117.
Bill Gropp
∂26-Oct-80 1348 CSD.NOWICKI at SU-SCORE Re: SAIL 8 font.
Date: 26 Oct 1980 1347-PST
From: CSD.NOWICKI at SU-SCORE
Subject: Re: SAIL 8 font.
To: CT at SU-AI
In-Reply-To: Your message of 26-Oct-80 1337-PST
Thanks for fixing this. You are not contributing to the trouble,
but are one of the people who solve more problems than complaining about
them. I still would like to get rid of the "Bad Pup" messages,
however.
-- Bill
-------
∂26-Oct-80 1500 PAM
page 7 of nss[spl,sys] appears to do what you want..Paul
∂26-Oct-80 1522 Mark Crispin <Admin.MRC at SU-SCORE> Re: SAIL 8 font.
Date: 26 Oct 1980 1522-PST
From: Mark Crispin <Admin.MRC at SU-SCORE>
Postal-Address: 12155 Edgecliff Place; Los Altos Hills, CA 94022
Stanford-Phone: (415) 497-1407
Subject: Re: SAIL 8 font.
To: CT at SU-AI
In-Reply-To: Your message of 26-Oct-80 1337-PST
Chris -
Thanks for all your good work. Do you think you can fix DOVER so that
it will correctly print those RFCnnn.TXT files on [1,MRC]?
Presently it bitches about incorrect password.
-- Mark --
-------
∂27-Oct-80 1453 LTP Chinese sample...
Don needs it...
∂27-Oct-80 1513 GFS dover
I just tried to dover over a file and got the following error:
[802] illegal press file 1257: entity list has odd number of bytes. (scan)
It worked once but not now.
∂27-Oct-80 1554 KDO More Dover bugs
Dover put me in a state of Bad pup 33 Receiver busy, which was printed as
"ABORT: CEIVER BUSY" for some reason, then Dover seems to be unavailable now,
will you wait? I said yes andd the same message was printed out and the same
question asked. After a while I ↑C'd out and tried to DOVER the .PRE file. That
complained about receiver busy for a while, then sent the file, but did not type
any .!'s...Oh well, the file didn't print anyway, same error as before
Ken
P.S. What about MRC's DOVER code in ATSIGN?
∂27-Oct-80 2357 PAM DOver lossage again
On Monday, the Dover was out of paper all afternoon, so DOVER would just wait
forever. I went to check it out, saw the red lights, loaded it with paper, and
tried again. Meanwhile, all the backlog from other lusers was coming out.
I dovered thrydp.>, pointr.>, abs.>, and bigvw.> all from [dis,pam]. The
abs file made it through, all others got error msg as follows:
Problems encountered:
[802] Illegal Press file 1303:Entity list ahs odd number of bytes.(Scan)
All files were E files, some with index page and some without.... Any ideas?
Paul
CT -- try it again...should work now...
∂29-Oct-80 0007 PAM
do still need the file loser.ct[dis,pam]? it was short file that wouldn't DOVER
∂30-Oct-80 0910 IAZ I have problems with your GFI file - and Ishall try
and find you later this morning.
CT -- there were many bugs in it. I think I have fixed them all. See Heather
for details.
∂30-Oct-80 2331 REP Final run of Karel
To: DRF, CT
I expect to do the final run of Karel on Saturday. Is there a chance that one
of you could show up in the moring and get me started, and then generally hold
my hand until it is finished. I estimate that the whole mess will take about
6 hours (based on last time). Any takers? If not, I'd like to run off a chapter
tomorrow, just to see how the whole thing is done so I can doit myself on Saturday.
Happy Haloween..rich
CT -- I can come in at 7:30am; Since I have work at the Med Center at 6am,
7:30 would be fine, but any later would be "idle time"...
∂31-Oct-80 1030 LTP Chemicals
I would like you to reorder them when needed. Please see me aboout accounts etc.
Thanx, Luis
∂03-Nov-80 1507 ME staff meeting
To: "@STAFF.DIS[CSD,REG]" at SU-AI
There will be a staff meeting Tuesday, 4 Nov 80 (election day), at
2:30pm in MJH 252. Note that the time is an hour later than usual.
∂04-Nov-80 1233 DRF alphatype unavailability
To: REP, CT
Is it ok with you guys if the alphatype becomes unavailable
starting this Friday, for some indeterminate amount of time?
Complain soon if not.
--drf
∂04-Nov-80 1727 EJG Illegal press file generated from large text file
To: BUG-Dover at SU-AI
DOVER seems to have problems with large text files (mine, anyway).
I have a text file generated with COPY/HEAD which when DOVERized
appears at the terminal to successfully list, but generates only
a dover header page containing the message:
Problems encountered:
[301] Illegal Press file 1117: Internal pointer exceeds file bounds.
I will leave the file which produces this problem in DOVER.BUG[1,EJG]
for a while, until I get tired of it pushing me over my allocation
in these times of scarce disk space. I can bypass the problem by
using page ranges on COPY/HEAD to split the file into two smaller
ones which both list OK. I would appreciate being informed when this
is fixed so I can stop being paranoid about having to split large
files before listing.
A SAIL terminal log of the DOVERizing appears below:
----------------------------------------------------------------------
.da
Tuesday 4-Nov-80 17:11 70F 21C
.di/fu dover.*
04-Nov-80 1711
Filnam Ext PPN Size Written Time Pro Writer Reference--% Dumped
[1,EJG] UFD Pro=405; Def Pro=005
DOVER BUG 1EJG 26.5 04-Nov-80 1709 005 1EJG COPY 04-Nov-80
Total= 26.5
Exit
↑C
.dover dover.bug
14 DOVER PRE 1EJG 162 163 W
14 DOVER BUG 1EJG 213 97 R
14 DOVER PRE 1EJG 364 334 R
Dover accepted file. Please pick up your output.
Exit
↑C
.da
Tuesday 4-Nov-80 17:14 70F 21C
.r ppsav
∂06-Nov-80 0852 Avron Barr <csd.barr at SU-SCORE> new font
Date: 6 Nov 1980 0850-PST
From: Avron Barr <csd.barr at SU-SCORE>
Subject: new font
To: csd.tucci at SU-SCORE, ct at SU-AI
Chris,
What's your status these days. Are you too busy to create this
graphics font? Know anybody else who might be interested?
.Av
-------
∂06-Nov-80 2159 ROB What??!?
To: BUG-eftp at SU-AI
--------------------
.as udp1
UDP1 Assigned
.r eftp
*receive
Local file name:udp1:bmp.irf[1,cis]
Illegal File specification
Local file name:↑C
↑C
.as udp1 dsk
UDP1(SCRATC) Assigned as DSK
.s
*receive
Local file name:bmp.irf[1,cis]
Mode: iMAGE mode
--------------------
Do you mean that EFTP doesn't understand about device names?
- Rob
∂11-Nov-80 1510 LTP Need to see you soon
Call me or sen msg asap. Thanx Luis
∂13-Nov-80 1227 LTP
Chris keep this format for tomorrows discussion with Heather, thanx.
comment
GNTblock→ starter gntname GNTfile
starter→ 00FF00FFH
gntname→ {8 bytes, left justified, padded with NULL (0) bytes}
GNT transfer→ [ GNTblock]+ endoftransfer
endoftransfer→ 55555555H
;
∂14-Nov-80 0844 HAL at MIT-MC (Hal Abelson)
Date: 14 NOV 1980 1143-EST
From: HAL at MIT-MC (Hal Abelson)
To: CT at SU-AI
CC: MJS at MIT-MC, RZ at MIT-MC, HAL at MIT-MC, LTP at SU-AI
LTP referred me to you about getting our book set on the alphatype.
We would like to run off one sheet as soon as possible, just to check
the fonts. Then we would like to do chapter 1 (about 50 pages). Then
the rest of the book in pieces, between now and January.
How do we get this process started? tell me the name of the file and
the fonts that you need for the sample.
Here is a list of the fonts we need:
CMR: sizes 5 6 7 8 9 10 11 14
CMI: 5 6 7 8 10
CMB: 8 9 10 11 14
CMS: 8 9 10 11 14
CMSY: 5 6 7 8 9
CMTT 8 10
We can provide you with metafont sources of any that you do not have,
so that they can be installed on the Alphatype.
O.K. I will send you a list of the metafont files we need. The large
fonts (bigger than 12-pt.) might be difficult; I have had trouble in
the past getting the whole fonts because metafont runs out of space.
Some questions:
Q1. We can tex things at score over the arpanet. What directories do we
have to know about, and where would you like us to stick the output?
A1. You should probably put the .DVI or .TEX files out on <4SCRATCH>
the day that I need them (the scratch directories are purged every night)
or (better yet) I could simply FTP them over when I am ready to start.
Q2. Do we have to do anything special about controlling how pages are
arranged on an Alphatype sheet, or does your software take care of
that?
A2. No, we will take care of that here.
Q3. How do we arrange for payment?
Q3. We will charge you $10 per sheet (note: probably 4 or 6 pages, depending
on how big your pages are) and send you a bill.
∂14-Nov-80 1048 LTP
Chris: come and see me this afternoon regarding HAL's book printing.
Heather expects you at 2:15.
∂14-Nov-80 1249 AVB dover
Chris -
it would be nice if a user does not have to keep typing "Y" to
"Dover seems to be unavailable right now, will you wait?"
if the dover is buys. Instead, the dover program, like empress on the alto,
should continue retrying by itself, maybe every few seconds,
maybe printing a message for each attempt.
Since the user really wants to print whatever he gives to Dover,
it seems pointless to ask him about his intentions again and again.
If the request really does not get serviced, I would ↑C out of the program
if I get tired of it.
Good luck - Andy
CT -- o.k. I made the change.
∂15-Nov-80 1733 Avron Barr <csd.barr at SU-SCORE> Font question
Date: 15 Nov 1980 1730-PST
From: Avron Barr <csd.barr at SU-SCORE>
Subject: Font question
To: ct at SU-AI
Chris,
The font CMSY10 seems to have a bug in it. According to
Dave Fuchs, somehow the XGP version is different from the Varian
version, and also different from the XGP version of CMSY9, in
that the character on '40 is not a "←",. Do you know how this
happened, or how to make a new verions of CMSY10 for the XGP?
.Av
-------
∂18-Nov-80 0804 REG Staff Meeting
To: "@STAFF.DIS[CSD,REG]" at SU-AI
3 pm, today, tuesday. In room 220 (I think)
∂18-Nov-80 1104 LTP eftp
Keeps saying "something's wrong!"
∂18-Nov-80 1256 Avron Barr <csd.barr at SU-SCORE> font hacking
Date: 18 Nov 1980 1254-PST
From: Avron Barr <csd.barr at SU-SCORE>
Subject: font hacking
To: csd.tucci at SU-SCORE
cc: ct at SU-AI
Chris,
I still haven't gotten around to creating that graphics
font. Any chance you'd be interested in doing it before thanksgiving?
Any chance you'd like me to forget it and stop bugging you?
.Av
-------
∂19-Nov-80 1144 LTP on TTY77 (at TV-134) 1144
Please get those tfx files or else I'll send you to Folsom Prison...
∂21-Nov-80 1415 HB gfi files for shasta
Chris - the gfi files for shasta are needed urgently now.
Please can you add cmtt8 to the list of fonts I gave you
last week. Thanks. Heather.
(you can send me a message on hb on sail)
∂22-Nov-80 0718 HAL at MIT-MC (Hal Abelson) printing our book
Date: 22 NOV 1980 1018-EST
From: HAL at MIT-MC (Hal Abelson)
Subject: printing our book
To: CT at SU-AI
Don Knuth told us that there was a last-minute font change, so we will
probably generate a new set of fonts for our use here at MIT.
Meantime, send us the list of fonts that you need. We should be ready
to send you some sample pages in a week or so.
∂24-Nov-80 1043 HB Missing gfi's
We are missing:
cmathx
bdr40
sta200
cmss10
Please get them ready by Tuesday afternoon. Thanx, Luis.
∂29-Nov-80 2138 ROB
To: bug-dover
CC: ME
Here are the totals for a DIR *.PRE[*,*]. We are currently down to 134
blocks. The total of 881.8 Kwords represents 78 blocks. This usually
shouldn't be a problem, but it is likely that people will start to use
the DOVER program faster than we aquire more disk space.
I propose that the DOVER program delete the .PRE file as the default. The
optional switch /-D would tell DOVER to save the .PRE file after spooling.
- Rob
--------------------
.DIR *.PRE[*,*] (somewhat edited)
[TMP,DLB] Total= 2.0
[BAC,LCS] Total= 1.0
[BIL,LCS] Total= 4.2
[1,LS] Total= 51.7
[CSD,REG] Total= 16.3
[1,JQJ] Total= 24.1
[1,TWE] Total= 1.1
[1,WIN] Total= 28.2
[1,MLT] Total= 14.0
[1,KDO] Total= 896
[1,CXF] Total= 11.1
[1,ROB] Total= 1.8
[1,RRT] Total= 1.3
[1,AJS] Total= 3.2
[1,TLH] Total= 52.5
[THE,DBG] Total= 73.2
[1,CET] Total= 6.3
[1,OP] Total= 3.6
[1,JJF] Total= 39.5
[1,TRG] Total= 4.6
[S80,JMC] Total= 4.5
[1,MJN] Total= 8.6
[1,AVB] Total= 96.3
[1,SMX] Total= 23.7
[1,FRG] Total= 1.1
[ISI,SJW] Total= 6.3
[TLP,RSC] Total= 3.0
[NET,RMS] Total= 37.2
[COM,RPG] Total= 15.2
[F80,JMC] Total= 10.3
[1,BKR] Total= 8.0
[1,ME] Total= 768
[1,SGN] Total= 19.3
[THE,RPG] Total= 1.2
[1,SJW] Total= 12.6
[1,CT] Total= 9.3
[1,RWF] Total= 8.7
[1,MAI] Total= 65.5
[1,KLC] Total= 4.6
[MEI,GFS] Total= 11.8
[DAS,ROB] Total= 1.7
[1,ARK] Total= 21.3
[1,BR] Total= 28.3
[1,RDG] Total= 3.1
[1,FC] Total= 896
[RDG,DBL] Total= 640
[1,DHG] Total= 3.5
[TEX,MFP] Total= 9.0
[1,LMF] Total= 2.2
[1,HSS] Total= 1.7
[1,DBG] Total= 256
[PER,CSD] Total= 1.1
[1,WDG] Total= 4.8
[PAG,LCS] Total= 1.7
[1,JER] Total= 1.7
[1,RPG] Total= 31.3
[BUR,CSD] Total= 8.3
[1,SSO] Total= 42.3
[PAS,SJF] Total= 1.6
[UP,DOC] Total= 19.2
[1,HN] Total= 6.5
Grand Total= 881.8
∂30-Nov-80 0058 ME
I agree with Rob that the default should be to delete the PRE file
as soon as it has been transmitted.
∂30-Nov-80 1521 VRP via Dialnet ethernet
eftp doesn't seem to be able to either send or receive
packets from the vaxes
∂01-Dec-80 2330 PEG DOVER
DOVER apparently went into an infinite loop when attempting to transfer the file
SLIM.PRE[1,PEG] (I'll leave it around for awhile). The file was generated by
SCRIBE on SCORE and FTPed to SAIL in image mode. Of 110 blocks in the file,
DOVER read 3 (by the file info line on the DD).
Phil Gerring
∂01-Dec-80 2348 PEG eftp...
Ahem. Same problem with EFTP, same file, now renamed to SLIM.BUG[1,PEG].
∂02-Dec-80 1636 PEG eftp
Why does EFTP not accept device designators in front of filenames? Since one of
the major uses (at present) for EFTP is to transfer extremely large files from
the CIS VAXen, it would seem reasonable to be able to write to a UDP...
∂02-Dec-80 1650 ROD UDPs
To: BUG-eftp at SU-AI
EFTP can't handle UDPs. Since it is mainly used to transfer absolutely gibiferous
VLSI files to SAIL, and filles disk, and doesn't stop to complain but sits
around wainting for suckers to delete files, then eats the tracks straight away,
it should work o UDPs so that users don't destroy the rest of the people trying
to work on SAIL. I seem to have to stop someone using EFTP about every three or
four days cause they have eaten all the disk and are continuaing while people
frantically delete files, and EFTP just gobbles all the disk up.
∂03-Dec-80 0036 ME
To: CT at SU-AI
CC: MRC at SU-AI
I have moved EFTP.FAI and DOVER.FAI onto [CSP,SYS] where
sources for system programs normally live. I may hack EFTP to use UDPs.
∂03-Dec-80 1021 GFS dover bug
DOVER seems to put the FILE PPN on the output as the FOR: value, whereas
it really should put the LOGIN PPN there. This morning a lot of people are
getting copies of AVB's Ethernet documentatation and all the copies are
coming out FOR: AVB. Also, the file name is different. I DOVERed the file
Eiover.pre and actually got ei.overview.press. All you can do is look for a
file that looks like what you wanted and rip it off.
∂03-Dec-80 1306 J. Q. Johnson <Admin.JQJ at SU-SCORE> dover error messages
Date: 3 Dec 1980 1026-PST
From: J. Q. Johnson <Admin.JQJ at SU-SCORE>
Address: 26970 Beaver La., Los Altos Hills, CA 94022; Phone: (415) 494-0860
Subject: dover error messages
To: ct at SU-AI
I frequently get:
?Received bad pup 33 from 50#201#
Abort: ceiver busy, please try again later.?Received bad pup 33 ......
Two minor complaints:
1/ this reply probably means the transfer should be aborted. It doesn't
seem to be.
2/ please fix that "ceiver" and make it "Receiver"!
-------
∂03-Dec-80 2206 YM EFTP
To: CT
CC: KDO
Hi,
Is there going to be EFTP version to talk from SAIL directly to IFS the way it
is done by the Alto FTP? I think that usually when you want to transfer file
to SAIL you want to do something with it on SAIL and you don't need to use the
Alto (assuming that the file is on IFS). More than this sometime you want while
you are logged in at SAIL to find some file in some directory before you
transfer it. Essentially the above says that the existence of such EFTP will
make IFS a file server for SAIL as well as for the Alto (and VAXen).
-Yoni
∂07-Dec-80 0221 RMS The Help file for the dover program
To: BUG-dover
is essentially missing. It says that the program will
be installed in the future, probably september.
It's after september now. Meanwhile, I wanted to make
multiple copies of a printout without having to send it
several times. I guessed that /copies:5 would do it
but it doesn't. Perhaps it ought to.
∂07-Dec-80 0511 DON DOVER[3,2]
Shouldn't the HELP DOVER text be updated?
∂07-Dec-80 1316 DON cleaning up
To: BUG-dover
Why doesn't the DOVER program delete the PRE file when it's done?
∂07-Dec-80 1646 HB GFI files for shasta
Chris - Thanks for the last batch of GFI files. I need one
more batch of them (at least I hope these will really be
the last). Please can you prepare GFI files for the fonts
cmbig, cmr18, cmr40, cmss18, cmti8,cmti9,cmtt0s and cmtt8n.
or , if you can give me instructions for preparing them
myself I can do it without bothering you each time.
Heather.
∂09-Dec-80 1029 HB eftp to transfer DVI files to shasta
Chris - We need your version of eftp with the right and laft
options for imge mode so that we can transfer DVI files to
shasta. Can you get that version in the system please or let
me know how to run it. Thanks, Heather.
∂09-Dec-80 2341 MOG eftp
I was in the middle of an EFTP (send, ascii, Yes, 300), when
it stopped printing .!.! and started printing ..... then
it printed Something's wrong!, and quit. I retried
several times, but all it printed was ...... Something's wrong!!
-Jeff