Date: Sun, 23 Nov 97 14:06:09 EST From: Snuffles@kew.com Subject: UUPC-Info-Request Digest 1997 #19 To: uupc-info-digest@kew.com Message-ID: Reply-To: UUPC-Info-Request@kew.com UUPC-Info-Request Digest Sun, 23 Nov 97 Volume 1997: Issue 19 Today's Topics: Continuation Lines and INEWS (3 msgs) Forwarding (yes again...) uucp feeds drying up? UUPC security problem? (4 msgs) UUPC security problem? (was: Re: uucp feeds drying up?) (4 msgs) 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: Fri, 21 Nov 1997 19:26:30 -0500 From: MarchHare@momeraths.org (Sysop) Subject: Continuation Lines and INEWS To: uupc-info@kew.com -----BEGIN PGP SIGNED MESSAGE----- I got an email not long ago, where I had posted in one of those long, ongoing threads. The emailer pointed out that I was posting broken headers. Specifically, when I checked, more than a couple of my posts were breaking up halfway through the References: line. It appears that what is happening is this: I use Yarn to read news and email. When Yarn posts to a newsgroup, and the References: line is over 80 characters, it puts in a newline, then a space at the next line and continues the information. For example: References: <345553c5.4602220@news.vianet.on.ca>\n <34640122.4243835@news.vianet.on.ca> <34734579.52112730@news.vianet.on.ca> Now, I looked through inews.c, and it sees the first whitespace in the line, and it assumes that it has reached the end of the headers: while (fgets(buf, sizeof(buf), input) != NULL) { if (buf[0] == ' ' || buf[0] == '\n' || buf[0] == '\r') { OK = KWTrue; break; } ... So, then you wind up with something like this: References: <345553c5.4602220@news.vianet.on.ca> Lines: 21 <34640122.4243835@news.vianet.on.ca> <34734579.52112730@news.vianet.on.ca> Lines: 18 Rest of message ... OK, I was able to then patch up inews.c to force it to ignore the beginning whitespace. Then, I got pretty curious as to which one is the correct behavior. So, I whipped out my trusty RFC822, and here are two quotes: [Quote 1.] Each header field can be viewed as a single, logical line of ASCII characters, comprising a field-name and a field-body. For convenience, the field-body portion of this conceptual entity can be split into a multiple-line representation; this is called "folding". The general rule is that wherever there may be linear-white-space (NOT simply LWSP-chars), a CRLF immediately followed by AT LEAST one LWSP-char may instead be inserted. Thus, the single line To: "Joe & J. Harvey" , JJV @ BBN can be represented as: To: "Joe & J. Harvey" , JJV@BBN and To: "Joe & J. Harvey" , JJV @BBN and To: "Joe & J. Harvey" , JJV @ BBN The process of moving from this folded multiple-line representation of a header field to its single line represen- tation is called "unfolding". Unfolding is accomplished by regarding CRLF immediately followed by a LWSP-char as equivalent to the LWSP-char. Note: While the standard permits folding wherever linear- white-space is permitted, it is recommended that struc- tured fields, such as those containing addresses, limit folding to higher-level syntactic breaks. For address fields, it is recommended that such folding occur August 13, 1982 - 5 - RFC #822 Standard for ARPA Internet Text Messages between addresses, after the separating comma. [End Quote 1.] [Quote 2.] 3.4.2. WHITE SPACE Note: In structured field bodies, multiple linear space ASCII characters (namely HTABs and SPACEs) are treated as single spaces and may freely surround any symbol. In all header fields, the only place in which at least one LWSP-char is REQUIRED is at the beginning of continua- tion lines in a folded field. [End Quote 2.] Therefore, I submit that Yarn is following the above behavior, and it is inews.exe that is not. While I'm at it, I also noticed in inews.c that it rewrites the date-time stamp. Why? John - -- //------------------------------------------------------------------------ // momerath@apk.net sevot yhtils eht dna ,gillirb sawT` // MarchHare@momeraths.org ebaw eht ni elbmig dna eryg diD // ,sevogorob eht erew ysmim llA // .ebargtuo shtar emom eht dnA // In case of stupidity, break glass. -----BEGIN PGP SIGNATURE----- Version: 2.6.3a Charset: cp850 iQCVAwUBNHYqubMHJ4yl8MPNAQGDlgP/azT8RKBTVyIyCzRcYYmXD5xk6k75xGgi NgGjYg25jQgZyrJw+gZEz4z3QdVHDzAxNoNUoJ3d98uD4SGGS/deqJsn/gbEmpfP hYj30VRTbM6VzihH6mTwcnweiDGXIXPrNnnJWP0HQc0bHFmEt4JBaSsh3YfF6tA+ z8JjlNOGr7w= =nRNW -----END PGP SIGNATURE----- ------------------------------ Date: Sat, 22 Nov 1997 08:50:45 -0500 From: Drew Derbyshire Subject: Continuation Lines and INEWS To: Sysop Sysop wrote: > Now, I looked through inews.c, and it sees the first whitespace in the > line, and it assumes that it has reached the end of the headers: . . . > OK, I was able to then patch up inews.c to force it to ignore the > beginning whitespace. Then, I got pretty curious as to which one is the > correct behavior. So, I whipped out my trusty RFC822, and here are two > quotes: . . . > Therefore, I submit that Yarn is following the above behavior, and it is > inews.exe that is not. You're right, but for the wrong reason, I think. RFC-822 is for mail, not news. Unless the news RFC specifically references RFC-822, it's a completely different standard. In any case, however, yes, headers should be able to wrap. Added to the bug list. > > While I'm at it, I also noticed in inews.c that it rewrites the > date-time stamp. Why? Beats the hell of out of me. I think to insure its in standard format. That, and it was easier to eat the line than save it. -- Internet: ahd@kew.com Voice: 781-279-9812 "Don't blame me, I voted for HAL!" - Kendra Derbyshire ------------------------------ Date: Sun, 23 Nov 1997 07:48:52 -0600 From: cwinemil@keys.lonestar.org (Chris Winemiller) Subject: Continuation Lines and INEWS To: Drew Derbyshire , Sysop Drew Derbyshire writes: >Sysop wrote: > >> Therefore, I submit that Yarn is following the above behavior, and it is >> inews.exe that is not. > >You're right, but for the wrong reason, I think. RFC-822 is for mail, not >news. Unless the news RFC specifically references RFC-822, it's a >completely different standard. In any case, however, yes, headers should >be able to wrap. Added to the bug list. RFC 1036 (Standard for Interchange of USENET Messages) does reference RFC 822: 2. Message Format ...A standard format for mail messages has existed for many years on the Internet, and this format meets most of the needs of USENET. ...Therefore, the rule is adopted that all USENET news messages must be formatted as valid Internet mail messages, according to the Internet standard RFC-822. ...In any situation where this standard conflicts with the Internet standard, RFC-822 should be considered correct and this standard in error. Regards, Chris -- Chris Winemiller Internet: cwinemil@keys.lonestar.org ------------------------------------------------------------------------- "Hey Rocky! Watch me pull some intelligence out of the Internet!" ------------------------------ Date: Wed, 19 Nov 1997 07:39:06 -0500 From: Drew Derbyshire Subject: Forwarding (yes again...) To: Ken Wiens , UUPC-Info Mailing List Of general interest, so copied to the list ... Ken Wiens wrote: > I have a UUCP site, with multiple sites below me. Every so often, one > of these sites will decide to discontinue receiving mail through me, and > move to a commercial ISP. > > It's easy to forward mail addressed to an individual on my site, but I > need to forward an entire sites mail to another address(s). > > For example - my site is kgwcal.cuug.ab.ca > A site I provide for is swancal.kgwcal.cuug.ab.ca > > I need to take (for example), mail addressed to jeff@swancal.cuug.ab.ca > and forward it to his new address of jeff@cadvision.com We don't support it unless the site is directly connected (in your systems file), in which case you forward under the old name and the site is than responsible for the aliasing. In general, you can arrange that the MX recorda for that subdomain point at the new MX forwarder, and that the MX forwarder understands the old name as well as the new. (The MX forwarder is still responisble for the aliasing). Or, you can write a gateway program driven from RMAIL to readdress the mail, in which case no external changes are needed. -- Internet: ahd@kew.com Voice: 781-279-9812 WNTC radio - We Never Talk Coherently ------------------------------ Date: Wed, 19 Nov 1997 11:41:04 -0800 From: "Vance, Steve" Subject: uucp feeds drying up? To: > My ISP just got bought out by another, larger company, and I've been > informed that they will no longer be offering uucp connections. > > I have called around, and so far only about 1 in ten ISPs even offer > uucp service at all. Prices range from $39 to $79 a month for it. > This seems expensive, especially since I have been paying $8.33 a > month before ($100 a year). > > It seems I am in danger of losing my domain! Does anyone have any > ideas on how to be "on the Internet" for less than the cost of a > full-time link (which is maybe a couple of hundred dollars a month)? > Help! > ------------------------------ Date: Fri, 21 Nov 1997 18:22:03 -0500 From: MarchHare@momeraths.org (Sysop) Subject: UUPC security problem? To: uupc-info@kew.com -----BEGIN PGP SIGNED MESSAGE----- On Thu, 20 Nov 97 20:58:03 +0100, hajo@quijote.Huellhorst.Zierke.com (Hans-Joachim Zierke) wrote: >I was in a similar situation over here. > >Solution: Now my second provider is xs4all in Amsterdam. I connect to >xs4all by means of a local throwaway provider. > >This works very well, with two problems: >1) UUCP over IP means cleartext passwords via Internet. This is a security >problem. >2) It is possible to crash any xs4all UUCP site just by sending mail, as >far as these sites use UUPC as their software. The trick is, that the >xs4all SMTP server understands RFC 821 source routing, while the UUCP >gateway does not know what to do with it, and writes just "C rmail" into >the commandfile, no address. I don't understand 2). How does this differ from normal operation? An SMTP server should not be trying to directly send a UUCP site email. It would have to be converted into a UUCP packet first. John - -- //------------------------------------------------------------------------ // momerath@apk.net sevot yhtils eht dna ,gillirb sawT` // MarchHare@momeraths.org ebaw eht ni elbmig dna eryg diD // ,sevogorob eht erew ysmim llA // .ebargtuo shtar emom eht dnA // In case of stupidity, break glass. -----BEGIN PGP SIGNATURE----- Version: 2.6.3a Charset: cp850 iQCVAwUBNHYYVrMHJ4yl8MPNAQE8WgP+MOVuxq2361onXRFPLjLJgWSvkeqfSUCy BOJuSJ3MDUGJaLl23J0l507Ihtbn2nl1/XsWKK9mWGX4DSVFi9aJX+GPSUnwJwge oLXR6b875PARCPJrQaiBc+A1hqGBEI1u33/eWuF1zh703NOaR6xm3/tJoQCjTKG6 HTsCglnlUE0= =yW36 -----END PGP SIGNATURE----- ------------------------------ Date: Sat, 22 Nov 1997 11:49:35 -0500 From: MarchHare@momeraths.org (Sysop) Subject: UUPC security problem? To: uupc-info@kew.com -----BEGIN PGP SIGNED MESSAGE----- On Fri, 21 Nov 97 16:22:22 +0100, hajo@quijote.Huellhorst.Zierke.com (Hans-Joachim Zierke) wrote: > >General Protection Fault exception occurred at EIP = 1F16BEAB on thread 0001. >Exception occurred in C Library routine called from EIP = 00021AF4. >Register Dump at point of exception: >EAX = 00000000 EBX = 00000000 ECX = FFFFFFFF EDX = 0003125A >EBP = 000824E4 EDI = 00000000 ESI = 00000000 ESP = 000824E0 > CS = 005B CSLIM = 1FFFFFFF DS = 0053 DSLIM = 1FFFFFFF > ES = 0053 ESLIM = 1FFFFFFF FS = 150B FSLIM = 00000030 > GS = 0000 GSLIM = 00000000 SS = 0053 SSLIM = 1FFFFFFF >Process terminating. >SYS1808: >Der Proze‹ wurde gestoppt. Der Softwarediagnosecode >(Fehlercode der Ausnahmebedingung) ist 0005. An OS/2 system, correct? ... >If I delete the X-File and feed the D-File into rmail, rmail eats and >delivers it happily. > ... > >The Zierke.com domain has two computers, one of them 400 kilometers away >from here. It is checked once a day - it sends a "alive" message. But it's >not that funny to travel 800 kilometers to fix a hanging computer program. You _could_ have it start up TELNETD.EXE on startup and then telnet into it to work on the problem. OK, I'm going to duck now, because I know what a security mess IBM's TELNETD is. OTOH, you could also have a look at OS2YOU, which allows remote operation of an OS/2 machine through one chosen "device", which can include IPX or TCP/IP. John - -- //------------------------------------------------------------------------ // momerath@apk.net sevot yhtils eht dna ,gillirb sawT` // MarchHare@momeraths.org ebaw eht ni elbmig dna eryg diD // ,sevogorob eht erew ysmim llA // .ebargtuo shtar emom eht dnA // In case of stupidity, break glass. -----BEGIN PGP SIGNATURE----- Version: 2.6.3a Charset: cp850 iQCVAwUBNHcOPbMHJ4yl8MPNAQHtYwP+P2vQZ+ooW0wGgl3CH3s55GJ5uplITWUc we1st9wzYjzhXMcFXFVLMtMkLmib94vd+ZzYR2bKYWG9mf2TqeSL6vdjPH2M8Rok nnJBJVZiFKsR+ZEPbL7nhQieYcGkChhBUtzsvcUlReJfY0bFV/8xPwgecFP6NlPe XfBKwRvaalo= =oaGF -----END PGP SIGNATURE----- ------------------------------ Date: Sun, 23 Nov 97 00:11:47 +0100 From: hajo@quijote.Huellhorst.Zierke.com (Hans-Joachim Zierke) Subject: UUPC security problem? To: uupc-info@kew.com MarchHare@momeraths.org (Sysop) schrieb am 21. 11. 97: >>2) It is possible to crash any xs4all UUCP site just by sending mail, as >>far as these sites use UUPC as their software. The trick is, that the >>xs4all SMTP server understands RFC 821 source routing, while the UUCP >>gateway does not know what to do with it, and writes just "C rmail" into >>the commandfile, no address. >I don't understand 2). How does this differ from normal operation? An >SMTP server should not be trying to directly send a UUCP site email. It >would have to be converted into a UUCP packet first. If I send mail to @terra.xs4all.nl:eric.I.crash.your.site.now the smtp server will correctly route it to the uucp login, the smtp/uucp gate will correctly route it to Eric, it will write a correct D-xyexample file and a nearly fine X-xyexample file. But the X-xyexample file, that is delivered to Eric and resolved by his UUPC uuxqt, will have a line C rmail instead of C rmail
If the UUPC uuxqt receives such a commandfile C line, it crashes. So if I have fun to do that, I can crash Erics site several times a day just by sending mail. (Hi, Eric ;-) ) hajo -- [Hans-Joachim Zierke]=======================[Telefon H³llhorst:] [ Hauptstra‹e 21 ] hajo@Zierke.com [ 05744 / 920678 ] [ D-32609 H³llhorst ] [ Telefon Berlin: ] [Fax: 05744 / 920679]=======================[ 030 / 394 8447 ] ------------------------------ Date: Sun, 23 Nov 1997 13:05:30 +0000 From: Eric Veldhuyzen Subject: UUPC security problem? To: hajo@Zierke.com Hans-Joachim Zierke writes: > MarchHare@momeraths.org (Sysop) schrieb am 21. 11. 97: >>> 2) It is possible to crash any xs4all UUCP site just by sending >>> mail, as far as these sites use UUPC as their software. The trick >>> is, that the xs4all SMTP server understands RFC 821 source >>> routing, while the UUCP gateway does not know what to do with it, >>> and writes just "C rmail" into the commandfile, no address. >> I don't understand 2). How does this differ from normal operation? >> An SMTP server should not be trying to directly send a UUCP site >> email. It would have to be converted into a UUCP packet first. > If I send mail to @terra.xs4all.nl:eric.I.crash.your.site.now Hey! I don't think I'll like that! > the smtp server will correctly route it to the uucp login, the > smtp/uucp gate will correctly route it to Eric, it will write a > correct D-xyexample file and a nearly fine X-xyexample file. > But the X-xyexample file, that is delivered to Eric and resolved by > his UUPC uuxqt, will have a line > C rmail > instead of > C rmail
> If the UUPC uuxqt receives such a commandfile C line, it crashes. So > if I have fun to do that, I can crash Erics site several times a day > just by sending mail. (Hi, Eric ;-) ) Nice to have friends on the net.... For the record, is this a UUPC specific bug or does Taylor UUCP also dump core after getting such messages? -- #!perl5 # Eric Veldhuyzen eric@terra.xs4all.nl # Moderator of Comp.os.os2.moderated PGP-KeyID: 0xFB64FCB3 $!=$;=$_+(++$_);($:,$~,$/,$^,$*,$@)=$!=~ # Perl is Fun! /.(.)...(.)(.)....(.)..(.)..(.)/;`$^$~$/$: $^$*$@$~ $_>&$;` # TEAM OS/2 ------------------------------ Date: Thu, 20 Nov 97 20:58:03 +0100 From: hajo@quijote.Huellhorst.Zierke.com (Hans-Joachim Zierke) Subject: UUPC security problem? (was: Re: uucp feeds drying up?) To: uupc-info@kew.com "Vance, Steve" schrieb am 19. 11. 97: >> I have called around, and so far only about 1 in ten ISPs even offer >> uucp service at all. Prices range from $39 to $79 a month for it. >> This seems expensive, especially since I have been paying $8.33 a >> month before ($100 a year). I was in a similar situation over here. Solution: Now my second provider is xs4all in Amsterdam. I connect to xs4all by means of a local throwaway provider. This works very well, with two problems: 1) UUCP over IP means cleartext passwords via Internet. This is a security problem. 2) It is possible to crash any xs4all UUCP site just by sending mail, as far as these sites use UUPC as their software. The trick is, that the xs4all SMTP server understands RFC 821 source routing, while the UUCP gateway does not know what to do with it, and writes just "C rmail" into the commandfile, no address. With this type of X-commandfile, you can remote-crash any UUPC installation, regardless of version. UUPC admins with anonymous logins might want to check, wether this imposes a security problem for their site. hajo ------------------------------ Date: Fri, 21 Nov 1997 08:44:44 -0500 From: Drew Derbyshire Subject: UUPC security problem? (was: Re: uucp feeds drying up?) To: hajo@Zierke.com Hans-Joachim Zierke wrote: > Solution: Now my second provider is xs4all in Amsterdam. I connect to > xs4all by means of a local throwaway provider. > > This works very well, with two problems: > 1) UUCP over IP means cleartext passwords via Internet. This is a security > problem. This is, as they say, a known restriction. Intercepting such transmissions requires an untrusted listening to your wire -- very easy in a corporate LAN environment, harder in a dial-up ISP environment. In our case, we use our regular firewall running on FreeBSD (http://www.freebsd.org) to restrict telnet and UUCPD access to known sites, specific IP addresses when possible and networks when not. This restricts inbound snooping being useful to only our "friends", not people we have ever heard of. Running two sites which support sequence numbering (UUPC/extended doesn't, it could be added), you can also at least be advised of forged connections because if the sequence numbers get out of sync > 2) It is possible to crash any xs4all UUCP site just by sending mail, as > far as these sites use UUPC as their software. The trick is, that the > xs4all SMTP server understands RFC 821 source routing, while the UUCP > gateway does not know what to do with it, and writes just "C rmail" into > the commandfile, no address. Are you saing their (UNIX) UUCP gateway does not? UUPC/extended has always supported source routing, and bug reports on it will be actively examined. > With this type of X-commandfile, you can remote-crash any UUPC > installation, regardless of version. UUPC admins with anonymous logins > might want to check, wether this imposes a security problem for their > site. I don't think it crashes, it would abort after UUPC/extended RMAIL complaints. Big difference. If it is an explicit abort on this condition, a patch to UUPC/extended's UUXQT would be possible to bounce the mail (since UUXQT knows RMAIL should have an argument). At the same time, malicous attacks can do an Denial of Service (DoS) attack on any mail system, if by nothing else by filling the disk. That's why you have GOT to monitor your running systems -- people who say "I didn't get mail for three weeks, and when I checked ..." lose my interest. -- Internet: ahd@kew.com Voice: 781-279-9812 "During emergency landing, replace dinner tray and bring seat to upright position. Extinguish all smoking materials . . . including the spacecraft, if possible." - Spaceman Spiff (aka Calvin) ------------------------------ Date: Fri, 21 Nov 97 16:22:22 +0100 From: hajo@quijote.Huellhorst.Zierke.com (Hans-Joachim Zierke) Subject: UUPC security problem? (was: Re: uucp feeds drying up?) To: uupc-info@kew.com Drew Derbyshire schrieb am 21. 11. 97: >This is, as they say, a known restriction. Intercepting such >transmissions requires an untrusted listening to your wire -- very easy in >a corporate LAN environment, harder in a dial-up ISP environment. That's why I risk it. >Are you saing their (UNIX) UUCP gateway does not? I don't know wether it is Unix... ;-) >I don't think it crashes, General Protection Fault exception occurred at EIP = 1F16BEAB on thread 0001. Exception occurred in C Library routine called from EIP = 00021AF4. Register Dump at point of exception: EAX = 00000000 EBX = 00000000 ECX = FFFFFFFF EDX = 0003125A EBP = 000824E4 EDI = 00000000 ESI = 00000000 ESP = 000824E0 áCS = 005B CSLIM = 1FFFFFFF DS = 0053 DSLIM = 1FFFFFFF áES = 0053 ESLIM = 1FFFFFFF FS = 150B FSLIM = 00000030 áGS = 0000 GSLIM = 00000000 SS = 0053 SSLIM = 1FFFFFFF Process terminating. SYS1808: Der Proze‹ wurde gestoppt. Der Softwarediagnosecode (Fehlercode der Ausnahmebedingung) ist 0005. --------------------------------------------------------------------- 11-21-1997 14:30:11 SYS3175 PID 282b TID 0001 Slot 0081 D:\UUCP\BIN\UUXQT.EXE c0000005 1f16beab P1=00000001 P2=00000000 P3=XXXXXXXX P4=XXXXXXXX EAX=00000000 EBX=00000000 ECX=ffffffff EDX=0003125a ESI=00000000 EDI=00000000 DS=0053 DSACC=d0f3 DSLIM=1fffffff ES=0053 ESACC=d0f3 ESLIM=1fffffff FS=150b FSACC=00f3 FSLIM=00000030 GS=0000 GSACC=**** GSLIM=******** CS:EIP=005b:1f16beab CSACC=d0df CSLIM=1fffffff SS:ESP=0053:000824e0 SSACC=d0f3 SSLIM=1fffffff EBP=000824e4 FLG=00012286 UPCR12S.DLL 0001:0002beab -------------------------------------------------------------------- >it would abort after UUPC/extended RMAIL >complaints. If I delete the X-File and feed the D-File into rmail, rmail eats and delivers it happily. >At the same time, >malicous attacks can do an Denial of Service (DoS) attack on any mail >system, if by nothing else by filling the disk. Filling a modern hard disk with 3000 cps needs quite a lot of time. >That's why you have GOT to >monitor your running systems -- people who say "I didn't get mail for three >weeks, and when I checked ..." lose my interest. The Zierke.com domain has two computers, one of them 400 kilometers away from here. It is checked once a day - it sends a "alive" message. But it's not that funny to travel 800 kilometers to fix a hanging computer program. hajo -- [Hans-Joachim Zierke]=======================[Telefon H³llhorst:] [ Hauptstra‹e 21 ] hajo@Zierke.com [ 05744 / 920678 ] [ D-32609 H³llhorst ] [ Telefon Berlin: ] [Fax: 05744 / 920679]=======================[ 030 / 394 8447 ] ------------------------------ Date: Fri, 21 Nov 1997 14:25:28 -0500 From: Drew Derbyshire Subject: UUPC security problem? (was: Re: uucp feeds drying up?) To: hajo@Zierke.com, UUPC-Info Mailing List Hans-Joachim Zierke wrote: > > Drew Derbyshire schrieb am 21. 11. 97: > > >This is, as they say, a known restriction. Intercepting such > >transmissions requires an untrusted listening to your wire -- very easy in > >a corporate LAN environment, harder in a dial-up ISP environment. > > That's why I risk it. > > >Are you saing their (UNIX) UUCP gateway does not? > > I don't know wether it is Unix... ;-) The 'standard' UNIX UUCP gateway is sendmail, which of course does understand source routing. Generating invalid data for the remote site is not really our problem, the abort (not the crash you show below) and hang is specifically programmed that way because we expect the system to be monitored. > >I don't think it crashes, > > General Protection Fault exception occurred at EIP = 1F16BEAB on thread 0001. Now THAT's a crash. I'll look at it. > >it would abort after UUPC/extended RMAIL > >complaints. > > If I delete the X-File and feed the D-File into rmail, rmail eats and > delivers it happily. > > >At the same time, > >malicous attacks can do an Denial of Service (DoS) attack on any mail > >system, if by nothing else by filling the disk. > > Filling a modern hard disk with 3000 cps needs quite a lot of time. > > >That's why you have GOT to > >monitor your running systems -- people who say "I didn't get mail for three > >weeks, and when I checked ..." lose my interest. > > The Zierke.com domain has two computers, one of them 400 kilometers away > from here. It is checked once a day - it sends a "alive" message. But it's > not that funny to travel 800 kilometers to fix a hanging computer program. I didn't say it was. I am saying someone _at_ _the_ _site_ needs to be monitoring it, NO WHERE WHATSOEVER do we claim UUPC/extended is suitable for locked room operation, it runs SOLELY on machines which really need someone to look at the console (NT, OS/2, DOS). If you running a machine 800 miles away, it should be a system which properly allows remote admin, i.e. UNIX (commerical, FreeBSD, Linux, whatever). I'd really love to say, "yes, you are doing exactly what we expect and unattended operation is our highest priority", but its just not true with all the robust free UNIX server software out there. -ahd- -- Internet: ahd@kew.com Voice: 781-279-9812 "Concrete Shoes, Cyanide, TNT! | Contracts, neck ties, HIGH VOLTAGE! Dirty deeds, done dirt cheap | Dirty deeds, done dirt cheap . . ." - AC/DC ------------------------------ End of UUPC-Info-Request Digest ******************************