Date: Sun, 17 Dec 95 00:36:17 PST From: Snuffles@kew.com Subject: UUPC-Info-Request Digest 1995 #52 To: uupc-info-digest@kew.com Message-ID: Reply-To: UUPC-Info-Request@kew.com UUPC-Info-Request Digest Sun, 17 Dec 95 Volume 1995: Issue 52 Today's Topics: 100 users crashes MAIL! Film at 11! Announce: MS Mail (and Exchange) gateway to UUCP (fwd) expire Mailheaders (2 msgs) uucp; vs. Winsock? (5 msgs) uucp vs SLIP/PPP (2 msgs) UUPC-Info-Request Digest 1995 #51 uupc server over tcp bug uupc vs. Winsock.. 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. Note: Questions on UUPC/extended itself which are not of general interest should be sent to help@kew.com, not to the mailing list. Nor questions should be posted on Usenet, we don't read it. (Much.) ---------------------------------------------------------------------- Date: Tue, 12 Dec 1995 23:13:23 -0500 From: Software@kew.com Subject: 100 users crashes MAIL! Film at 11! To: UUPC/Extended mailing list Actually, it really is 11 PM as I write this. A number of users have reported the current releases of UUPC/extended still have a 100 user PASSWD file limit after I had lifted it in 1.12k or so. This is incorrect, it has a BUG. Specifically, for those with source, there is a limit check in usertabl.c (line 200, +- 20 lines) which compares (userElements-1) to allocUsers. This should actually be: if ( userElements == allocUsers ) The -1 causes the users table to overflow _before_ reallocation, causing a user not be copied when the array is copied by realloc (which gets more memory.) I will also be replacing a duplicate user check which inccurs a nasty performance penalty when loading (it scans the entire unsorted table for every user), and instead will check for duplicate users after the table is sorted by the C run-time library. This code will be available as source and beta test executables later this week. The release primarily be to test news fixes, including truncation of the active file and various failures caused by buffer overflows under DOS. Availability of the actual beta test will ONLY be announced on uupc-info and uupc-info-digest. Look for a real release towards the end of the month. This will be announced on all the mailing lists, per normal policy. -ahd- -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-279-9712 "Well some they do and some they don't And some you just can't tell . . ." - Supertramp ------------------------------ Date: Wed, 13 Dec 1995 09:24:02 +0000 (WET) From: johan@hsb.nest.nl Subject: Announce: MS Mail (and Exchange) gateway to UUCP (fwd) To: UUPC/Extended mailing list I thought this might be interesting for some of you on this mailingslist. Regards, Johan Johan Henselmans work: johan@netsense.nl tel: (+31)-(0)20-6261180 http://www.netsense.nl Netsense vof, Nieuwe Herengracht 149 1011 SG Amsterdam home: johan@hsb.nest.nl tel: (+31)-(0)20-6267538 Nieuwe Herengracht 147d 1011 SG Amsterdam ---------- Forwarded message ---------- Date: 11 Dec 1995 12:13:38 GMT From: Lee C. Bussy Newgroups: comp.mail.uucp Subject: Announce: MS Mail (and Exchange) gateway to UUCP I have posted a few replies lately but I guess I'll hit it all with one shot: SpinMail has just been released by the author (not me) and is a UUCP <> MS Mail gateway that runs unser 95 and as a service in NT or under WfWG. System reqirements are a UUCP connect, a 386 or better with 8 megs and 40 megs disk space. Installation is pretty damnned simple and the directions are quite good. This will not make you an instant UUCP expert! My wife did set it up once to get it running according to the directions but that's not what this is for. This is a powerfull tool and after it's set up (takes me about 15 minutes to install it now) it requires minimal maintenance. You can get some more infor at: http://www.cfw.com/~middletn/spinmail.html or DL both versions at: ftp://ftp.southwind.net/users/l/leeb No financial interest, just helping the author out for a while. -- -Lee C. Bussy/leeb@southwind.net * I feel sorry for people who don't drink. When they wake up, * it's as good as they are going to feel all day. ------------------------------ Date: Wed, 13 Dec 95 21:34:48 -0500 From: sysop@mome.apk.net Subject: expire To: UUPC/Extended mailing list peter@pgeck.sub.org wrote: > > Hi, > another question: How can I tell expire to leave news for a > different time in different newsgroups? I want to keep some news > for a long time, others only for some hours or days. > > Try: expire -e<#days> . It also appears to work if you do: expire -e4 alt.binaries.* Can anyone verify that it will actually accept wildcards? John ------------------------------ Date: Thu, 14 Dec 95 0:34:49 +0100 From: eric@terra.xs4all.nl Subject: Mailheaders To: UUPC/Extended mailing list Hi, My access provider just mentioned to me that they had to patch their rmail heavily to get all the mail delivered right. When I asked what was wrong, he told me that the header should start with From user date remote from node But I've noticed that the headers that are in my mails start with From terra.xs4all.nl!eric Thu Dec 14 00:22:52 1995 remote from terra So this means that the "terra.xs4all.nl!" part shouldn't be there. Since I don't want to edit all my headers manually before polling, I want to fix thisi automagically. How should I do this? Is this standard behavior of UUPC that can't be fixed or did I make a mistake somewhere? BTW I am using UUPC 1.12k OS/2 32bit -- 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: Thu, 14 Dec 95 19:33:50 -0500 From: sysop@mome.apk.net Subject: Mailheaders To: UUPC/Extended mailing list eric@terra.xs4all.nl wrote: > My access provider just mentioned to me that they had to patch their > rmail heavily to get all the mail delivered right. When I asked what > was wrong, he told me that the header should start with > > From user date remote from node > > But I've noticed that the headers that are in my mails start with > > From terra.xs4all.nl!eric Thu Dec 14 00:22:52 1995 remote from terra > > So this means that the "terra.xs4all.nl!" part shouldn't be there. Actually, my interpretation of _Managing uucp and Usenet_ by O'Reilly is that this is the correct way to do it. It is the "bang path" method, and it is used exclusively (AFAIK) in UUCP. Your provide should then add to this header. If your provide is xxxnet, then the adress becomes xxxnet!terra.xs4all.nl!eric. A good example is on p. 294 of the above reference in the chapter "Frequently Asked Questions". > BTW I am using UUPC 1.12k OS/2 32bit I honestly don't know if this will help you or not, but I had problems with a rather unusual provider before I moved, and upgrading solved some of the problems. John ------------------------------ Date: Mon, 11 Dec 1995 00:54:11 -0500 From: uupcinfo@kew.com Subject: uucp; vs. Winsock? To: UUPC/Extended mailing list On 11 Dec 1995 00:00:00 +0000, hajo@quijote.in-berlin.de wrote: > If you get some points, it is not for shortcomings of UUCP, but for > shortcomings of UUPC. > If you have major traffic both ways, PPP is faster, because UUPC does > not support uucp-i. True. Taylor UUCP blows our doors off. :-) (Although I know what Ian's dad runs on his PC, and it's not Taylor UUCP. Ian installed it for him, which led to a really bizarre phone call one night.) > rmail transport is utterly unefficient. But at least for the OS/2 platform, > this might get replaced by gbsmtp in 1996. Since every OS/2 is delivered > with a Sendmail Server, most of the work can be configured in a file that's > everybodies darling: sendmail.cf. :-) No! No! Not sendmail.cf! If you promise to never speak of sendmai.cf again. I'll look at 'i'. The problem, BTW, is memory and horse power on DOS. (But not for at least a year, I've got promises to keep for news and 'g' protocol!) BSMTP is also on my list, I knew how to do that before I knew UUCP. (BITNET, my haven of a decade ago invented it!) -- 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. VM programmers do it virtually all the time. ------------------------------ Date: 11 Dec 1995 00:00:00 +0000 From: hajo@quijote.in-berlin.de Subject: uucp; vs. Winsock? To: UUPC/Extended mailing list Hi Drew, Du schriebst am 11.12.95: > > Since every OS/2 is > > delivered with a Sendmail Server, most of the work can be configured in a > > file that's everybodies darling: sendmail.cf. :-) > > No! No! Not sendmail.cf! > > If you promise to never speak of sendmai.cf again. I'll look at 'i'. Don't panic, don't panic! _If_ you can talk somebody into porting smail to OS/2, I will tell the devil to go away... :-) hajo ------------------------------ Date: Mon, 11 Dec 95 13:03:25 CST From: guthrie@mum.edu Subject: uucp; vs. Winsock? To: UUPC/Extended mailing list Interesting, of course this assumes that you are not running uucp over TCP/IP, sop perhaps the answer here is that if one uses a point-to-=point uucp link, it more secure thatn a packet switched TCP/IP link. However (!), wouldn't the same security apply to any direct-dialed (to the host) PPP link! I.e. isn't this an issue of the physical link security, rather than the protocol run over that link? It is interesting to wonder why users of UUPC choose it, is it for these technical reasons, or historical reasons, or familiarity, or ..? I use(d) it for local compatibility, and historical reasons, but have now moved to TCP/IP. ------------------------------------------ Gregory Guthrie ------------------------------------------ On Mon, 11 Dec 95 09:57:28 EST bbeazley@cnsinc.com wrote: > Gregory, > > To answer your question of: > Why UUCP .vs. PPP? > Quite simple: > Security through a protocol switch! > > UUCP is not TCP/IP. Nor does it allow the services therein. Therefore, > although it may be more cumbersome to manage (for now) it is less > susceptible to hackers. And quite frankly, Firewalls were made for > breaking and boots were made for walking. > > Robert Beazley > CNS Consulting ------------------------------ Date: Mon, 11 Dec 95 13:09:07 CST From: guthrie@mum.edu Subject: uucp; vs. Winsock? To: UUPC/Extended mailing list Not to be a non-advocate, but... On Mon, 11 Dec 1995 08:41:07 -0500 John Donohue at Donohue&Co. wrote: >One thing it allows is multiple userids in a very simple manner. >I have seen a network installation where all the users are sending and >receiving mail for a total cost of $75/month. > >Also, a single user can have multiple userids for different purposes. >If I simply have one CompuServe ID, that's all I have. -- Of course :-) all TCP/IP tools also want this, and most support it. I know Eudora for example allows multiple users with various logins, all on one machine. I do not think it allows batch mail transfer from all these multiple acounts, at not easily, although one can probably coerce it to do it.. This area of multiple users on a PC is an interesting one, that the current tools and OS do not often support so well. ------------------------------------------ Gregory Guthrie ------------------------------------------ ------------------------------ Date: Tue, 12 Dec 1995 10:25:13 -0600 (CST) From: paulus@nextdown.pe.utexas.edu Subject: uucp; vs. Winsock? To: UUPC/Extended mailing list According to guthrie@mum.edu: + +It is interesting to wonder why users of UUPC choose it, is it for +these technical reasons, or historical reasons, or familiarity, or ..? my network uses UUPC due: - really robust and easy to setup for rural area of Indonesia where phone line is very noisy, and tcp/ip is still only at major cities and expensive, - independent, no risk of government ban of Internet in the future. - can also use tcp/ip via those major city ISPs, to go to the main uucp server in the USA (a linux box running Taylor UUCP) (this one uses Trumpet Winsock and UUCICO/windows version, using uucp protocol t) - can also go via all terminal server, which is very handy for my US members, instead of long-distance call to the server or a SLIP account (this one uses a slightly modified uucico of version 1.12p and protocol f) TTFN, Paulus -- Paulus Suryono Adisoemarta, N5SNN / YG1QN yono@parokinet.org n5snn@mail.utexas.edu paulus@nextdown.pe.utexas.edu ------------------------------ Date: Mon, 11 Dec 1995 19:26:04 -0700 From: wieringa@xroads.com Subject: uucp vs SLIP/PPP To: UUPC/Extended mailing list >I wonder why today would anyone choose uucp/pc over SLIP/PPP? >It is my impression that uucp came from the 1970's, remained in the >'80s, but that now in the '90s the standatd for networking has >changed. I get good uucp support free. I can run it "hands-off" quite reliably on a timed basis. I haven't found any other realiable free service at compares. --- Dave Wieringa wieringa@xroads.com Mesa, AZ (recently moved from MI) http://www.cps.msu.edu/~wieringa ------------------------------ Date: Mon, 11 Dec 1995 19:26:04 -0700 From: wieringa@xroads.com Subject: uucp vs SLIP/PPP To: UUPC/Extended mailing list >I wonder why today would anyone choose uucp/pc over SLIP/PPP? >It is my impression that uucp came from the 1970's, remained in the >'80s, but that now in the '90s the standatd for networking has >changed. I get good uucp support free. I can run it "hands-off" quite reliably on a timed basis. I haven't found any other realiable free service at compares. --- Dave Wieringa wieringa@xroads.com Mesa, AZ (recently moved from MI) http://www.cps.msu.edu/~wieringa ------------------------------ Date: Mon, 11 Dec 1995 19:34:22 -0700 From: wieringa@xroads.com Subject: UUPC-Info-Request Digest 1995 #51 To: UUPC/Extended mailing list >Date: Sat, 9 Dec 95 19:14:22 -0500 >From: uupc@mistik.express.net >Subject: Good news (sort of) >To: UUPC/Extended mailing list > ... >This one is false. You better go back to printing my manual and don't >write about things you are knowledgable with. ... >Since you are wrong on #4, and perharps slandering us, you better go >print the manuals and don't get involved in issues that are obviously >over your head. ... >:) It seems you can't keep what you read in your memory. The argument >was about us not being his service provider for internet access. [various other UNWARRANTED insults toward KEW deleted] Get a clue and get a life. You are WAY over your head here. Drew and his wife are good people who have put a lot of hard work into a good product. I'm sure they haven't made peanuts from their efforts, which is why they also have full-time jobs. I'm suprised they've put up with you at all. --- Dave Wieringa wieringa@xroads.com Mesa, AZ (recently moved from MI) http://www.cps.msu.edu/~wieringa ------------------------------ Date: Sat, 16 Dec 1995 19:18:00 -0500 From: Software@kew.com Subject: uupc server over tcp bug To: UUPC/Extended mailing list On Sat, 16 Dec 1995 15:27:33 -0500, "Mustafa Soysal" wrote: > Here is the doctored log (without the password and API errors) that > shows a serious problem. Anyone who got it to work? Note that the > other site doesn't have 'e'. 't' and 'g' are the only choices, however, > it seems to go for 't' on network connections. We could not get it to > try 'g'. Judging by your question, which is very reasonable with the detail given, I presuem you wanted this on the list . I've cc'ed this reply to it. My question is, what is the remote site using for UUCP? That looks very much like you are talking to the old (1.12k?) version of UUPC/extended; if it is neither that nor Taylor UUCP, I need to find a copy of the remote user's package to test with. > uupc uucico aborts when sending or receiving the first file for the same > excuse. > > > (1) Monitoring port tcptty device tcp for 546 minutes until user hits Ctrl-Break > (4) S state = I > (4) ==> Access to this service is only allowed for the subscribers. If you are > > (4) ==> not a paid subscriber, please disconnect NOW. > > (4) ==> > > (4) ==> > login: > (4) <== Uspinet > (4) ==> > Password: > ...password deleted by Mustafa Soysal > (0) login: login user Uspinet (spinet) at Sat Dec 16 15:17:28 1995 > (4) S state = A > (4) S state = K > (4) ==> ^pShere=mistik > (4) <== ^pSspinet > (2) 1st msg from remote = Sspinet > (0) h:/apps/comm/spool/locks.lck/.DUMB.TEST.NAME: The system cannot open the device or file specified. > (4) advancedFS: h:/apps/comm/spool/locks.lck/spinet.LCK resides on file system supporting long file names > (4) ==> ^pROK > (4) ==> ^pPgGfvet > (4) <== ^pUt > (3) setproto: wanted 't', have 'g' > (3) setproto: wanted 't', have 'G' > (3) setproto: wanted 't', have 'f' > (3) setproto: wanted 't', have 'v' > (3) setproto: wanted 't', have 'e' > (3) setproto: wanted 't', have 't' > (0) mistik called by spinet: network link, t protocol, z grade > (4) S state = M > (3) ImportPath: Mapped h:/apps/comm/spool/locks.lck/*status.LCS to h:/apps/comm/uupc/spool/locks.lck/7status.LCS > (4) process: Machine state is = m > (4) topenpk: Timeout = 90 sec, buffer size = 1024 bytes, packet size = 512 bytes > (4) process: Machine state is = l > (4) process: Machine state is = n > (2) <<< S D.spine2f4d D.spine2f4d spinet - D.spine2f4d 0666 dummy > (4) process: Machine state is = p > (3) ImportPath: Mapped D.spine2f4d to spinet/D/nohqvu > (0) Receiving "D.spine2f4d" as "D.spine2f4d" (spinet/D/nohqvu) > (2) Using temp name h:\apps\comm\uupc\spool\uupc1553.tmp > (2) >>> SY > (4) process: Machine state is = q > (0) tsread: EOF on recv() > (0) tgetpkt: Length read failed > (4) process: Machine state is = t > (0) process: Connection lost to spinet, previous system state = q > (4) S state = N > (4) ==> ^pOOOOOO > (0) tsread: EOF on recv() > (2) rmsg: Timeout waiting for sync > (4) ==> ^pOOOOOO > (0) 0 files sent, 0 files received, 3 bytes sent, 58 bytes received > (0) 2 packets transferred, 0 errors, connection time 0:13, 4 bytes/second > (4) S state = P > (4) S state = Q > (3) ImportPath: Mapped h:/apps/comm/spool/locks.lck/*status.LCS to h:/apps/comm/uupc/spool/locks.lck/7status.LCS > -- Drew Derbyshire UUPC/extended e-mail: software@kew.com Telephone: 617-279-9712 "I just program for Drew because Binkley Bunny (who is as snowy white as I am) and Gunther Bear used to help Drew program from the top of kendra's monitor. They went west in September [1990] with Mom so that Mom and Binkley could work on their PhDs. That made the Wonderworks understaffed, so at Christmas [1990] Mom asked me to live with Drew." - Snuffles P. Bear ------------------------------ Date: Mon, 11 Dec 95 12:55:00 CST From: guthrie@mum.edu Subject: uupc vs. Winsock.. To: UUPC/Extended mailing list On Sun, 10 Dec 1995 18:52:14 -0800 Dave Watt wrote: >UUCP is only a protocol. The protocol works over both modem lines > and network connections like PPP. -- yes; i.e. it has variants of protocol that it uses over each. >The UUCP protocol itself has some virtues as well. For one thing, > it's more efficient than the currently used network protocols for > sending and receiving Usenet news. -- ?? using a layered protocol model, here uucp protocols run on top onf some lower level; if it is a simple byte stream link, then uucp does a lot of the link layer tasks, else it uses a simpler protocol ("t" ?) if it is on top of TCP/IP which handles those details. of course SLIP and PPP are just implementations of a TCP/IP link. News and mail then traverse this link, with their own protocols (SMTP, NNTP), so I am not sure what it means for uucp to be more efficient; than what? > People have found that it's more efficient to ship >all Usenet news over UUCP over TCP/IP than it is to try to negotiate >out with your next-door neighbors about which news articles you have, >and which they have -- the latter is part of a news protocol called NNTP. -- isn't this issue independent of which lower level protocol is used to transact the news? with any underlying protocol one can subscribe to batched news delivery. Again NNTP is a higher level protocol. > If you're at the edge of the Internet, UUCP is more efficient. -- I don't see why. Isn't the issue that at the edge of the news routing web, that selective batched news transfers are the most efficient? Noting that almost any news reader today can do this, regardless of the protocol used to connect. PS: I am an admitted non-expert in news protocols... ------------------------------------------ Gregory Guthrie ------------------------------------------ ------------------------------------------ Gregory Guthrie ------------------------------------------ ------------------------------ End of UUPC-Info-Request Digest ******************************