Date: Sun, 10 Sep 95 12:27:25 EDT From: Snuffles@kew.com Subject: UUPC-Info-Request Digest 1995 #32 To: uupc-info-digest@kew.com Message-ID: Reply-To: UUPC-Info-Request@kew.com UUPC-Info-Request Digest Sun, 10 Sep 95 Volume 1995: Issue 32 Today's Topics: 1.12o and OK = success Alpha test: 1.12o on ftp.clarkson.edu (2 msgs) bug report--uuxqt chokes on long addresses for rmail in X files DOS Based newsreaders How to install 1.12o? (2 msgs) News (2 msgs) Problems while calling from WinNet 2.20 System Alias file ?s (2 msgs) Timezone question Trailing dots on addresses UUPC V1.12o BigGPacketSize Windows NT version of UUPC with Win 95? (3 msgs) Word 6.0 source document... To subscribe to UUPC-Info-Digest, send the command in the body of a message to listserv@kew.com: subscribe uupc-info-Digest To signoff from UUPC-Info-Digest, use "signoff" instead of "subscribe". You can also send an "index" to the listserv to get an index of back issues and other files available for retrieval. ---------------------------------------------------------------------- Date: Sun, 10 Sep 1995 02:37:44 -0100 From: stefan@donald.ruhr.de Subject: 1.12o and OK = success To: UUPC/Extended mailing list The subject sais it all, uupc1.12o for NT solved all my problems and runs beautifull until now. Many thanx, Ciao Stefan -- "Leverkusen stoppt die Seuche !!!" (kicker nach dem Spiel BVB-LEV) ------------------------------ Date: Thu, 07 Sep 1995 21:11:52 -0400 From: Software@kew.com Subject: Alpha test: 1.12o on ftp.clarkson.edu To: UUPC/Extended mailing list On Tue, 5 Sep 95 23:43:43 +0000, "K.P.Kirchdoerfer" wrote: > XHelp@kew.com schrieb: > > upc12oaw.zip 199308 Word 6.0 source document > > Can You please provide a version in any other format readable with > OS/2 (ascii, tex, postcript...) ? ASCII and postscript for 1.12p, I promise. -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 Every Creature has within him the wild, uncontrollable urge to PUNT. ------------------------------ Date: Fri, 08 Sep 1995 07:13:19 -0400 From: Software@kew.com Subject: Alpha test: 1.12o on ftp.clarkson.edu To: UUPC/Extended mailing list On Thu, 7 Sep 1995 22:40:39 +0200 (MET DST), "Rene Baart" wrote: > XHelp@kew.com writes: > > > > upc12on1.zip 330623 Windows NT > > upc12on2.zip 415899 > > upc12on3.zip 361905 > > > > Does the NT version run under Windows 95 as well? This is what I have been told. Let me know how you make out. -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 "Nuke those VMBLOK's and them suckers is history" - J. J. McMahon, out of context ------------------------------ Date: Sat, 09 Sep 1995 17:39:46 -0400 From: XHelp@kew.com Subject: bug report--uuxqt chokes on long addresses for rmail in X files To: UUPC/Extended mailing list > I have found problem with the command line in the email command > files (the ones that go to the directory c:\uupc\spool\fugazi\x I > think). When I ran UUXQT it gave an error about not being able to > execute a shell and it has some truncated version of mequanint's > address in the error line. This error kept it from processing any > mail (I saw some mail waiting in c:\uupc\spool\fugazi\d). So I > went, looked at the mail control (or command?) file with the > earliest time. Noticed that it had a command line like: > > C rmail asmara.asmara.punchdown.org!archive asmara.asmara.punchdown.org!mequanint asmara.asmara.punchdown.org!robert asmara.asmara.punchdown.org!mequanin > > I deleted the last address from the command line, and then UUXQT ran > like a charm and delivered the mail. This is not a problem that is > going to be easy for the folks here to figure out if it happens again. > > ANALYSIS: > > the code in process() in uuxqt.c which reads the X file has a max line > length of BUFSIZ bytes (i am guessing that BUFSIZ is 1024 bytes in the > DOS build environment). lines longer than this will cause the code to > break. this is worth documenting, since it is entirely possible to get > more than 1024 bytes of addresses for some mailing lists. the mailer > at my end (smail) can be configured to a maximum number of address > bytes in a single message, so if the maximum value were noted in the > documentation then it would be easy to stay within that limit. > however, that is not the source of this problem (the addresses here > appear to be only 150 bytes). The DOS limit is actually 512. This is also the limit for many UNIX systems. > the code in shell() in uuxqt.c has a limit of 126 bytes for the command > name and arguments (at least with Turbo C in a DOS environment). This is actual DOS limit, the length is a fixed system structure. > This > should probably be documented as well (perhaps the limit is 253 bytes > with the Microsoft C compiler in a DOS enviornment ???). > > in any case, RMAIL gets special handling in shell(), and for rmail > under DOS the limit is 126 bytes for command and arguments, leaving > 120 bytes for addresses. Note the line: > > int rlen = IsDOS() ? 126 : sizeof commandBuf - 2; > > the code appears to detect and flag addresses longer than 126 bytes: . . . > although it unfortunately panics rather than simply skipping the file, > assuring that the same problem will recur the next time uuxqt is > invoked. returning the mail to the sender, or converting the address > to a shorter one (known to be bogus) would be a better solution for > unattended systems. I agree with this, except such handling requires smarts on the part of the system which I'm not really prepared to put into the DOS environment because of space considerations and the like. In short, it's not supposed to do that. This complaint can also be applied to the non-zero return code check from RMAIL further down -- there is a contest between small and complete, and complete lost. > the code implies that it handles address lists longer than 120 bytes > correctly by making multiple invocations of RMAIL, but in fact it does > not do this. instead, it invokes RMAIL only once, on the entire set of > addresses passed in (the outer while loop does not include the call to > execute() so only one call to RMAIL will be performed. in fact, the > code appears to be quite buggy. I didn't look at your code in detail, but I saved it lest you think the following is still unacceptable. It was in fact wrong, and someone submitted a fix in time for 1.12o, which was to terminate the inner loop within the shell() function before the check for firstpass(). This loop now looks like: while (( parameters != NULL ) && (result != -1 )) { KWBoolean firstPass = KWTrue; int left; size_t rlen = sizeof commandBuf - 2; *commandBuf = '\0'; rlen -= strlen( commandBuf ) + strlen( RMAIL ) + 1; /*--------------------------------------------------------------------*/ /* Copy addresses into the buffer */ /*--------------------------------------------------------------------*/ left = (int) (rlen - strlen( parameters )); while ((parameters != NULL) && (left > 0)) { char *next = strtok( NULL, ""); if ( *parameters == '-') /* Option flag for mail? */ printmsg(0,"Disallowed option %s ignored",parameters); else { /* Not option, add to param list */ strcat( commandBuf, " "); strcat( commandBuf, parameters ); left -= strlen( parameters ) + 1; firstPass = KWFalse; } /*--------------------------------------------------------------------*/ /* Step to next parameter */ /*--------------------------------------------------------------------*/ if ( next == NULL ) parameters = NULL; else parameters = strtok( next, WHITESPACE ); } /* while ( parameters != NULL ) */ if (firstPass) /* Did we process at least one addr? */ { /* No --> Serious problem! */ printmsg(0, "Address \"%s\" too long (%d chars)! %d available, short fall would be %d", parameters, strlen(parameters), rlen, left ); panic(); } /* if (*commandBuf = '\0') */ /*--------------------------------------------------------------------*/ /* Execute one command line of addresses */ /*--------------------------------------------------------------------*/ result = execute( RMAIL, commandBuf, bflag[F_WINDOWS] ? NULL : inname, outputName, KWTrue, KWFalse ); if ( result != 0 ) /* Did command execution fail? */ { printmsg(0,"shell: command \"%s %s\" returned error code %d", cmdname, commandBuf, result); panic(); } } /* while */ -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 "We are Microsoft. Unix is irrelevant. Openness is futile. Prepare to be assimilated." - anonymous .sig file ------------------------------ Date: Fri, 8 Sep 1995 07:49:04 5GMT From: cipav@mafalda.univalle.edu.co Subject: DOS Based newsreaders To: UUPC/Extended mailing list What are people using as DOS based news readers for UUPC (other than SNEWS 1.91)? Is it possible to interface PC-Pine to UUPC for example? Snews is very nice but only really suitable for leaf nodes. I am wondering if there is anything that will inject news into a UUPC based network that will take advantage of the new news distribution capability of UUPC Thanks in advance, Nicholas Waltham Fundacion CIPAV | Carrera 35 A Oeste # 3-66, (Centro para la Investigacion | Tejares de San Fernando, A.A. 20591, en Sistemas Sostenibles de | Cali, Colombia Produccion Agropecuaria) | Tel: +57-2 5542300 / +57-2 5542294 ------------------------------ Date: Sat, 9 Sep 95 0:00:06 (null) From: eric@terra.xs4all.nl Subject: How to install 1.12o? To: UUPC/Extended mailing list Hi, I just tried to install UUPC 1.12o but since there is no readable documentation, it failed. My news reader (trn) couldn't get the articles delivered localy, so I tried to fix it, didn't know what to do and returned to the backup I made. What should I do to get a working version of 1.12o? -- Eric Veldhuyzen TEAM OS/2 Eric.Veldhuyzen@si.hhs.nl CIS: [100010,3051] Eric@terra.xs4all.nl PGP-KeyID: 0xFB64FCB3 ************ FIGHT to keep your right to PRIVACY. Use PGP! ************ ------------------------------ Date: Sat, 09 Sep 1995 17:50:58 -0400 From: uupcinfo@kew.com Subject: How to install 1.12o? To: UUPC/Extended mailing list On Sat, 9 Sep 95 0:00:06 (null), eric@terra.xs4all.nl wrote: > I just tried to install UUPC 1.12o but since there is no readable > documentation, it failed. My news reader (trn) couldn't get the articles > delivered localy, so I tried to fix it, didn't know what to do and > returned to the backup I made. What should I do to get a working version > of 1.12o? The HP LJ 4si we borrowed to dump the docs will be confused to hear the docs are unreadable, as it spit out all 200 pages with nary a burp. (I assume it didn't burp, I actually kicked off the printing from 10 miles away and the printer was back in power save mode before I ever picked up the hardcopy.) 1.12p will include the usual formats of the docs. In any case, 1.12o works just like 1.12n, if you had that working 1.12o should work just the same. If you DIDN'T have 1.12n working and don't have Word 6.0, the same "don't try this at home" warnings apply. -ahd- -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 To sign off from uupc-info, send the command "signoff uupc-info" in the body of a message to listserv@kew.com. DO NOT send this request to the list itself! For human assistance with the list itself, send mail to snuffles@kew.com. Support Little League Roller Ball! ------------------------------ Date: Wed, 06 Sep 1995 23:16:43 -0500 From: plummer@altamira.theme.com Subject: News To: UUPC/Extended mailing list Is uupc 1.12n (and o, p, ...) able to post news as well as receive it? --Bill -- William W. Plummer, 7 Country Club Dr., Chelmsford, MA 01824-4505 508-256-9570 PP-ASEL, N1NGK, MA RE Broker plummer@theme.com ------------------------------ Date: Fri, 08 Sep 1995 08:20:52 -0400 From: uupcinfo@kew.com Subject: News To: UUPC/Extended mailing list On Wed, 06 Sep 1995 23:16:43 -0500, "William W. Plummer" wrote: > Is uupc 1.12n (and o, p, ...) able to post news as well as receive > it? --Bill Yes. inews is the news injector, you need a wrapper to create the headers. See pnews.cmd in the OS/2 archive for a sample wrapper. -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 To sign off from uupc-info, send the command "signoff uupc-info" in the body of a message to listserv@kew.com. DO NOT send this request to the list itself! For human assistance with the list itself, send mail to snuffles@kew.com. "AT&T is a modem test command." ------------------------------ Date: Sat, 09 Sep 1995 18:00:22 -0400 From: Software@kew.com Subject: Problems while calling from WinNet 2.20 To: UUPC/Extended mailing list On Fri, 08 Sep 1995 13:54:53, "Tahir Mufti" wrote: > As WinNet is quite a popular software, I was wondering if other > people had made similar complaints? No complaints on UUPC/extended and WinNET failing to initialize on the protocol message. Anyone? > I know for a fact that WinNet supports only the 'g' protocol. Is > there a way to tell UUPC to transmit only ^pg rather than the list > of protocols. Maybe WinNet gets confused because of that? Not without rebuild the source. Have you contacted the authors of WinNET? -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 "We are Microsoft. Unix is irrelevant. Openness is futile. Prepare to be assimilated." - anonymous .sig file ------------------------------ Date: Sun, 10 Sep 95 6:30:42 +0000 From: tannen@crusader.stat.com Subject: System Alias file ?s To: UUPC/Extended mailing list Hi, Up until now I have not had to have a system aliases file but now I do need one. I am a bit unclear on a few things: 1. Does the aliases file reside in the same directory as the uupc.rc file? (Based on the documentation I think it does) 2. Is the filename for the system aliase file suppose to be 'ALIASES'? 3. Can anyone tell me why the following lines in a system aliases file would not work? # *--------------------------------------------------------------------* # * The following aliases are used by DZwo * # *--------------------------------------------------------------------* prayer: | g:\dzwo\d2p.exe -# -d Prayer prayer-request: | g:\dzwo\d2p.exe -# -R prayer-archive: g:\usr\pub\prayer\prayers.1995 prayer-users: :include: g:\dzwo\prayer.mbr dzwo-error: g:\usr\tannen\dzwo.errors Thanks in advance. An ambassador for Christ David Tannen (tannen@crusader.stat.com) John 14:21 "Whoever has my commands and obeys them, he is the one who loves me. He who loves me will be loved by my Father, and I too will love him and show myself to him." ------------------------------ Date: Sun, 10 Sep 1995 12:26:08 -0400 From: uupcinfo@kew.com Subject: System Alias file ?s To: UUPC/Extended mailing list On Sun, 10 Sep 95 6:30:42 +0000, tannen@crusader.stat.com wrote: > Up until now I have not had to have a system aliases file but now I > do need one. I am a bit unclear on a few things: > > 1. Does the aliases file reside in the same directory as the > uupc.rc file? (Based on the documentation I think it does) Normally. it goes in the directory specified by ConfDir in UUPC.RC, which defaults to the directory the UUPC.RC file is in. > > 2. Is the filename for the system aliase file suppose to be > 'ALIASES'? Yes. > 3. Can anyone tell me why the following lines in a system aliases > file would not work? No. You don't say how the aliases file. You can see RMAIL's processing if you issue the following command: set uupcdebug=10 This may help determine exactly RMAIL is doing for or to you. > # *--------------------------------------------------------------------* > # * The following aliases are used by DZwo * > # *--------------------------------------------------------------------* > prayer: | g:\dzwo\d2p.exe -# -d Prayer > > prayer-request: | g:\dzwo\d2p.exe -# -R > > prayer-archive: g:\usr\pub\prayer\prayers.1995 > > prayer-users: :include: g:\dzwo\prayer.mbr > > dzwo-error: g:\usr\tannen\dzwo.errors -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 To sign off from uupc-info, send the command "signoff uupc-info" in the body of a message to listserv@kew.com. DO NOT send this request to the list itself! For human assistance with the list itself, send mail to snuffles@kew.com. "Then I heard that long whistle whine . . ." ------------------------------ Date: Fri, 8 Sep 1995 00:56:12 -0400 (EDT) From: DCantor@chqsplay.MV.COM Subject: Timezone question To: UUPC/Extended mailing list In UUPC Info Digest #30 and #31 there was a discussion about how to resolve the different ways that UUPC/Extended and Pegasus Mail use the TZ environment variable. Drew supplied one solution. I, too, am a user of SNUUPM, and I solved it by having (the equivalent of) the following statements in my AUTOEXEC.BAT file: SET TZ=EST5EDT rem Change next two statements when clocks change. SET TIMESTRING=EDT SET TIMEOFFSET=-0400 and editing the file %SNDRIVE%\USR\LIB\SNUUPM\UTL\MAIN.MNU as follows Mail set savetz=%tz% set tz=%timeoffset% (%timestring%) %sndrive%\usr\lib\snuupm\pmail.exe set tz=%savetz% David A. Cantor +1 860.444.7268 (444-RANT) 453 Bayonet St., #16 Connecticut has a new area code. New London, CT 06320 ------------------------------ Date: Thu, 07 Sep 1995 20:58:30 -0400 From: Software@kew.com Subject: Trailing dots on addresses To: UUPC/Extended mailing list On Wed, 6 Sep 1995 20:27:07 +0200, "Stephen Marquard" wrote: > Drew, > > Is there any particular reason why RMAIL should deliver > rmail scm@silver.wcape.school.za > locally, but spool > rmail scm@silver.wcape.school.za. > off to the mailserver? (assuming domain=silver.wcape.school.za in > uupc.rc) > > One can override this behaviour in HOSTPATH with > silver.wcape.school.za. = silver > but I can't think of a rational reason for the behaviour in the first > place. Sure. UUPC/extended treats that trailing period as significant. To explicitly ignore is more code, which has never been an issue. For example, look at how your address reached me (In the 'On Wed line, above.) -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 Every Creature has within him the wild, uncontrollable urge to PUNT. ------------------------------ Date: Sat, 09 Sep 1995 17:41:29 -0400 From: XHelp@kew.com Subject: UUPC V1.12o BigGPacketSize To: UUPC/Extended mailing list On Fri, 08 Sep 1995 11:01:42 -0500, "Drew Derbyshire - UUPC Help Desk" wrote: > I have just upgraded Digi-Tech from UUPC V1.12k to V1.12o and I > noticed that the connections were using G protocol as before, but > that the packet size was 512 instead of 1024. Adding a > BigGPacketSize=1024 option to the modem file fixed this, but the > documentation says that the default should be 1024. Either the docs > or the code needs to be changed to match each other (preferably the > code). No, the default is now 512, and that should be used unless you are paying by the packet. It's a long story, the details of which will be included in the 1.12p docs. -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 "We are Microsoft. Unix is irrelevant. Openness is futile. Prepare to be assimilated." - anonymous .sig file ------------------------------ Date: Fri, 8 Sep 1995 04:38:01 mst-8 From: leemal@pc.jaring.my Subject: Windows NT version of UUPC with Win 95? To: UUPC/Extended mailing list I would like to know whether the Windows NT version of UUPC will work with Windows 95. What are the advantages of doing this (if any) and what makes the NT version of UUPC different from the Win 3.1 version? Thanks.-- Lee Tet Woo ---> Happy Garden, 58200 Kuala Lumpur, Malaysia. ---> leemal@pc.jaring.my ---> The Unicus in #beginner on Undernet. --!@@1 111 Warni!@, ma@!er error. do yo!@ want to co!$#@nue? ------------------------------ Date: Thu, 07 Sep 1995 23:41:15 -0800 From: dave@smersh.watt.com Subject: Windows NT version of UUPC with Win 95? To: UUPC/Extended mailing list A while ago, I tried this, and it worked just fine. Knowing what I do about the code, I can predict that the code for setting the system time from the NIST probably will not work properly, because it uses the NT security APIs, which are not implemented on Windows 95. (You have to give yourself permission to change the system time, in case you were wondering.) The NT version is likely to have fewer resource limits than the Windows 3.1 version. I don't know if the performance is likely to be any better; frankly, I doubt it, since most of the time the app is busy is spent in the comm or network drivers. - Dave On Fri, 8 Sep 1995 04:38:01 mst-8, leemal@pc.jaring.my wrote: > I would like to know whether the Windows NT version of UUPC will work > with Windows 95. What are the advantages of doing this (if any) and > what makes the NT version of UUPC different from the Win 3.1 version? > > Thanks.-- > Lee Tet Woo ---> Happy Garden, 58200 Kuala Lumpur, Malaysia. > ---> leemal@pc.jaring.my > ---> The Unicus in #beginner on Undernet. > --!@@1 111 Warni!@, ma@!er error. do yo!@ want to co!$#@nue? -- Dave Watt (N1HMB) Home: 510-352-6799 dmwatt@smersh.watt.com San Leandro, California [A computer is] like an old testament god, with lots of rules and no mercy. - Joseph Campbell ------------------------------ Date: Fri, 08 Sep 1995 08:36:23 -0400 From: uupcinfo@kew.com Subject: Windows NT version of UUPC with Win 95? To: UUPC/Extended mailing list On Fri, 8 Sep 1995 04:38:01 mst-8, leemal@pc.jaring.my wrote: > I would like to know whether the Windows NT version of UUPC will work > with Windows 95. What are the advantages of doing this (if any) and > what makes the NT version of UUPC different from the Win 3.1 version? The Windows 3.1 version is a 16 bit pseudo Windows application, driven by Windows GUI messages, etc., and write it's output into a GUI window. Because the internal code doesn't handle the GUI interface well, we don't even ship MAIL or other some other programs. It mostly exists to allow running UUCICO under Windows using the Windows communications drivers. The OS/2 2,x and NT versions are both 32 bit character mode applications using native API's (OS/2 for OS/2, Win32 for NT). This allows the code to follow the natural logic flow of the original DOS application but to do so more efficiently and a superset of function, because the 32 bit API's have such things as good communications drivers and TCP/IP support. These also handle multitasking very well. In short, while the Windows 3.1 version is a limited subset of the original function squeezed into the Windows 3.1 environent, the OS/2 and NT versions are both full versions which do the job better then the original. -ahd- -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 To sign off from uupc-info, send the command "signoff uupc-info" in the body of a message to listserv@kew.com. DO NOT send this request to the list itself! For human assistance with the list itself, send mail to snuffles@kew.com. "AT&T is a modem test command." ------------------------------ Date: Fri, 08 Sep 1995 07:11:49 -0400 From: Software@kew.com Subject: Word 6.0 source document... To: UUPC/Extended mailing list On Thu, 7 Sep 1995 23:25:58 -0700 (PDT), (Tom Watson) johana!tsw@isp.net wrote: > Since you are going to provide this as postscript, I will volunteer to > confirm that it can be printed out on a variety of Laser Printers. I > can immediately access an Apple LaserWriter Plus (it is 24 inches away > as I write this), and an HP Laserjet 4 (there are other letters after > the model number, but it duplexes real neat, it is at work). We actually have access to LJ's and LaserWriters, although not here. We often run drafts on te same printers, although if someone would tell me why it tends to puke on SparcPrinters somewhere around the table for the MAIL subcommands. > I tried to read the Word 6.0 document with Word 5.1 for the Mac (with > translators) and it bombed (I tried many times). If I could get the > 'source' document in a more 'transportable' format, I could probably > convert it to a "Portable Postscript" format that could be copied > directly to a subject printer (i.e. Rich Text). If you want to try > this, just let me know, I'm sure we can work something out. RTF, at least as saved by Word 6.0, invariably loses format information over the native Word 6.0. We really won't put effort into to getting it into multiple formats -- getting people to buy the manual and support from us is not, after all, a bad thing for us, since it encourages us to keep up the good work. -ahd- -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-641-3452 "Nuke those VMBLOK's and them suckers is history" - J. J. McMahon, out of context ------------------------------ End of UUPC-Info-Request Digest ******************************