View Issue Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0002386 | unreal | ircd | public | 2005-03-04 15:25 | 2005-03-04 16:30 |
Reporter | SolutechUK | Assigned To | syzop | ||
Priority | normal | Severity | major | Reproducibility | always |
Status | resolved | Resolution | fixed | ||
Platform | Win32 | OS | Win XP pro SP2 | OS Version | SP2 |
Product Version | 3.2.3 | ||||
Summary | 0002386: Squits Ircservices on join | ||||
Description | When launching the services package it joins and immediately squits with Software caused connection abort . And SQUIT Unknown error | ||||
Steps To Reproduce | Start servcices as normal and it just wont stay in there . | ||||
Additional Information | Tested the config on a 3.2.2b build and it sits pretty in there no probs . Go back to pre-1 and it squits immediately | ||||
Tags | No tags attached. | ||||
3rd party modules | None | ||||
|
What version of IRCServices? And are you sure it's configured for Unreal 3.2? Also, any messages in the services log would be helpful. |
|
Yup its configured fine . Set it up on 3.2.2b and runs without a hitch .I ported the config from my live box where its been running aok. Move the same config onto pre-1 and bam it squits on connect .I even chucked up an old 3.2.2b test install and it connected fine to that and was stable. Tested with irc services 45 and 48 (latest)Both were clean installs to test the beta with. Nothing in services log at all . It seems to squit it the second after it connects so ircservices doesnt even have time to process the log b4 it gets shutdown . |
|
hm, nothing in log and it squits... perhaps it crashes? or would something differently happen then? (I'm not familiar with w32.. eventlog? :P). |
|
The following quote is taken from the ircservice mailing list. I hope this helps you. "Hello all seem like Services does not start with new Unreal build, that support NICKCHARS feature. here is services log: ------- [Mar 04 10:58:09.431373 2005] debug: Sent: PROTOCTL SJOIN SJOIN2 SJ3 NICKv2 VHP VL NOQUIT UMODE2 TOKEN NICKIP [Mar 04 10:58:09.433700 2005] debug: Sent: PASS :********** ... more sends, SERVER, NICK other ... [Mar 04 10:58:09.458010 2005] debug: Received: :irc.nwg-nv.ru NOTICE AUTH :*** Looking up your hostname... [Mar 04 10:58:09.460022 2005] debug: Received: :irc.nwg-nv.ru NOTICE AUTH :*** Found your hostname (cached) [Mar 04 10:58:09.531932 2005] debug: Received: PROTOCTL NICKCHARS= [Mar 04 10:58:09.540767 2005] debug: Sent: ERROR :Need NICKv2 protocol [Mar 04 10:58:09.543939 2005] Remote server doesn't support NICKv2" |
|
[oh you got a useful log? scratch my previous bugnote then] Oh... yeah Unreal now sends 2 PROTOCTL's to a server.. because I think we were aproaching (or hitting) the MAXPARA limit and I wanted to be safe with that AND the "512" buffer limit. But... that wouldn't be illegal, eh codemastr? It's nowhere stated at least, and it seems logical to allow multiple PROTOCTLS :) Anyway, what I could do is just reverse the order: send most protoctls first, and then the NICKCHARS= one after that :).. Although that might not fix the ircservices issue, I don't know. |
|
[Fri Mar 04 19:29:15 2005] - SERVER services.ourchat.tk [Fri Mar 04 19:29:15 2005] - Lost connection to services.ourchat.tk[127.0.0.1]:Software caused connection abort Thats from ircd.log services.log is blank . No mention of the error on the ircservices list which they say as its based on a cvs and not an official build they wont even look at yet . |
|
hmmm mayby should have gotten this into also :) Not me that is testing this I saw the bug and Im on the ircservice mailing list so I thought I share the info. This was also in the mail: "and code from Unreal sources (send_proto procedure, s_serv file): ------- void send_proto(aClient *cptr, ConfigItem_link *aconf) { char buf[512]; sendto_one(cptr, "PROTOCTL NICKCHARS=%s", langsinuse); sprintf(buf, "CHANMODES=%s%s,%s%s,%s%s,%s%s", CHPAR1, EXPAR1, CHPAR2, EXPAR2, CHPAR3, EXPAR3, CHPAR4, EXPAR4); #ifdef ZIP_LINKS if (aconf->options & CONNECT_ZIP) { sendto_one(cptr, "PROTOCTL %s ZIP %s", PROTOCTL_SERVER, buf); } else { #endif sendto_one(cptr, "PROTOCTL %s %s", PROTOCTL_SERVER, buf); #ifdef ZIP_LINKS } #endif } ... more code ... ------- Server always send first PROTOCTL NICKCHARS, therefor services crashed to start. I just tested, moved NICKCHARS after "normal" PROTOCTL... and services doesn`t start again, it crashed on PROTOCTL NICKCHARS, because no NICKv2 in this PROTOCTL directive." |
|
heh too bad :p. While, when I carefully look, I think merging them in PROTOCTL would not cause any problem (enough room), one might argue that merging them to fix this problem is only moving the problem ahead (or however you say that in english).. there will be a day, probably not too far from now, that we might need multiple PROTOCTL lines (eg: if just 1 or 2 tokens are added, or if we add more languages that can be combined), so I'm not entirely sure what to do :p. |
|
Talk nicly to Andrew Church? |
|
Just merged PROTOCTLS for now, I'll see if we can have a word with our services friends another time when we really need it -- I didn't think they would be processing PROTOCTLs in such an 'old' way :). Thanks for the report. |
Date Modified | Username | Field | Change |
---|---|---|---|
2005-03-04 15:25 | SolutechUK | New Issue | |
2005-03-04 15:25 | SolutechUK | 3rd party modules | => None |
2005-03-04 15:32 |
|
Note Added: 0009403 | |
2005-03-04 15:40 | SolutechUK | Note Added: 0009404 | |
2005-03-04 16:00 | syzop | Note Added: 0009407 | |
2005-03-04 16:00 | psadi | Note Added: 0009408 | |
2005-03-04 16:04 | syzop | Note Added: 0009409 | |
2005-03-04 16:08 | SolutechUK | Note Added: 0009410 | |
2005-03-04 16:10 | psadi | Note Added: 0009411 | |
2005-03-04 16:23 | syzop | Note Added: 0009412 | |
2005-03-04 16:26 | psadi | Note Added: 0009413 | |
2005-03-04 16:30 | syzop | Status | new => resolved |
2005-03-04 16:30 | syzop | Resolution | open => fixed |
2005-03-04 16:30 | syzop | Assigned To | => syzop |
2005-03-04 16:30 | syzop | Note Added: 0009414 | |
2005-03-04 16:32 | syzop | Note Edited: 0009414 |