Unrealircd problems with last patch
Hi, i've upgraded c-ares and Unreal from ports but Unreal won't work. It start without a problem but when someone try to connect to the server it crash with a core dump error: Jun 16 09:03:33 hazard kernel: pid 57652 (ircd), uid 0: exited on signal 11 (core dumped) I've tried to recompile unreal and c-ares whitout any result (make install finish without problems on both ports). If there's something that i could try, please tell me... I'm on a FreeBSD 7.2-RELEASE-p1 #0. Thanks ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Unrealircd problems with last patch
On Wed, Jun 17, 2009 at 08:24:38AM +0200, Andrea 'simplex' Zulato wrote: Hi, i've upgraded c-ares and Unreal from ports but Unreal won't work. It start without a problem but when someone try to connect to the server it crash with a core dump error: Jun 16 09:03:33 hazard kernel: pid 57652 (ircd), uid 0: exited on signal 11 (core dumped) I've tried to recompile unreal and c-ares whitout any result (make install finish without problems on both ports). If there's something that i could try, please tell me... I'm on a FreeBSD 7.2-RELEASE-p1 #0. Hi, I've CC'd Gerrit Beine (the actual maintainer of the irc/unreal port :) and Ilya Andreev, who reported the same problem to me yesterday. Some time ago, I sent my proposed c-ares update for testing to all the maintainers of ports that depend on c-ares directly. My patches made the ports compile, but I didn't have the proper setup to actually test them working, since I'm not quite familiar with the programs themselves. Thus, I asked the maintainers for help with testing, and after nobody replied for a week or so, I went ahead and commited the update. Now Ilya Andreev and you have both hit a problem with UnrealIRCd, and Ilya seems to have found a solution. Could you try putting the attached patch-res.c into the irc/unreal/files/ directory and rebuilding UnrealIRCd? If this patch helps, I could commit it if Gerrit Beine does not mind. G'luck, Peter -- Peter Pentchev r...@ringlet.netr...@space.bgr...@freebsd.org PGP key:http://people.FreeBSD.org/~roam/roam.key.asc Key fingerprint FDBA FD79 C26F 3C51 C95E DF9E ED18 B68D 1619 4553 This inert sentence is my body, but my soul is alive, dancing in the sparks of your brain. pgp9sY6QwwSdg.pgp Description: PGP signature
Re: Unrealircd problems with last patch
On Wed, Jun 17, 2009 at 11:52:02AM +0300, Peter Pentchev wrote:
On Wed, Jun 17, 2009 at 08:24:38AM +0200, Andrea 'simplex' Zulato wrote:
Hi, i've upgraded c-ares and Unreal from ports but Unreal won't work.
It start without a problem but when someone try to connect to the server
it crash with a core dump error:
Jun 16 09:03:33 hazard kernel: pid 57652 (ircd), uid 0: exited on signal
11 (core dumped)
I've tried to recompile unreal and c-ares whitout any result (make
install finish without problems on both ports).
If there's something that i could try, please tell me...
I'm on a FreeBSD 7.2-RELEASE-p1 #0.
Hi,
I've CC'd Gerrit Beine (the actual maintainer of the irc/unreal port :)
and Ilya Andreev, who reported the same problem to me yesterday.
Some time ago, I sent my proposed c-ares update for testing to all
the maintainers of ports that depend on c-ares directly. My patches
made the ports compile, but I didn't have the proper setup to actually
test them working, since I'm not quite familiar with the programs
themselves. Thus, I asked the maintainers for help with testing, and
after nobody replied for a week or so, I went ahead and commited the update.
Now Ilya Andreev and you have both hit a problem with UnrealIRCd,
and Ilya seems to have found a solution. Could you try putting
the attached patch-res.c into the irc/unreal/files/ directory and
rebuilding UnrealIRCd? If this patch helps, I could commit it if
Gerrit Beine does not mind.
Hmm, that's funny. I keep forgetting that sometimes mailing lists
may strip attachments :) Here's the patch (sorry if you're receiving
it twice)
G'luck,
Peter
diff -u -r1.1.1.1.6.1.2.71.2.26 -r1.1.1.1.6.1.2.71.2.27
--- src/res.c 2009/02/01 16:43:33 1.1.1.1.6.1.2.71.2.26
+++ src/res.c 2009/05/13 10:28:06 1.1.1.1.6.1.2.71.2.27
@@ -722,21 +722,34 @@
} else
if (*param == 'i') /* INFORMATION */
{
- struct ares_config_info inf;
+ struct ares_options inf;
int i;
+ int optmask;
- ares_get_config(inf, resolver_channel);
+ ares_save_options(resolver_channel, inf, optmask);
sendtxtnumeric(sptr, ** DNS Configuration Information
**);
sendtxtnumeric(sptr, c-ares version: %s,ares_version(NULL));
- sendtxtnumeric(sptr, timeout: %d, inf.timeout);
- sendtxtnumeric(sptr, tries: %d, inf.tries);
- sendtxtnumeric(sptr,# of servers: %d, inf.numservers);
- for (i = 0; i inf.numservers; i++)
- sendtxtnumeric(sptr, server #%d: %s, i+1,
inf.servers[i] ? inf.servers[i] : [???]);
-
- /* TODO: free or get memleak ! */
+
+ if(optmask ARES_OPT_TIMEOUTMS)
+ sendtxtnumeric(sptr, timeout: %d,
inf.timeout);
+ if(optmask ARES_OPT_TRIES)
+ sendtxtnumeric(sptr, tries: %d, inf.tries);
+ if(optmask ARES_OPT_SERVERS)
+ {
+ sendtxtnumeric(sptr,# of servers: %d,
inf.nservers);
+ for (i = 0; i inf.nservers; i++)
+ sendtxtnumeric(sptr, server #%d: %s,
i+1, inet_ntoa(inf.servers[i]));
+ }
+ if(optmask ARES_OPT_DOMAINS)
+ {
+ sendtxtnumeric(sptr,# of search domains: %d,
inf.ndomains);
+ for (i = 0; i inf.ndomains; i++)
+ sendtxtnumeric(sptr, domain #%d: %s,
i+1, inf.domains[i]);
+ }
sendtxtnumeric(sptr, ** End of DNS Configuration Info
**);
+
+ ares_destroy_options(inf);
} else /* STATISTICS */
{
sendtxtnumeric(sptr, DNS CACHE Stats:);
--
Peter Pentchev r...@ringlet.netr...@space.bgr...@freebsd.org
PGP key:http://people.FreeBSD.org/~roam/roam.key.asc
Key fingerprint FDBA FD79 C26F 3C51 C95E DF9E ED18 B68D 1619 4553
I am the thought you are now thinking.
pgpJSC2VMwUOx.pgp
Description: PGP signature
Re: [CFT] GTK frontend for VirtualBox
On Tue, June 16, 2009 4:45 pm, Alexander Logvinov wrote: Hello! For GTK lovers I've created a port for vboxgtk 0.4.1: http://www.logvinov.ru/files/dist/vboxgtk.tar . Sometimes it works. :) 2VBox Team: The latest 0.5.0-beta1 version of vboxgtk wants python bindings for the VirtualBox Python API support. Any chance to get it? Sure. I'll look into it but it can take some days. -- Bernhard Fröhlich http://www.bluelife.at/ ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
portupgrade/ruby issue? (Stale lock file was found. Removed.)
After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Sporadic - because they appear when there are no parallel execution of any portupgrade tools (or any other package/port related tools). They appear in situations wheere they never appeared before. I see this on two stable/7 systems. -- Andriy Gapon ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: pkg_libchk: a missing library is not detected
Dominic Fandrey kamik...@bsdforen.de writes: Luck. The app linking to the old library will have a dependency on the old version. pkg_add will find the origin, issue a warning about app-1.0 needing lib-0.1 but lib-0.2 is installed and proceed. app will not start, because of the missing library. I've never had this case. I've got the impression that pointyhead rebuilds all dependencies. Let me remind you the case with my original question: 1. Graphics/eog was broken (as you said) by indirect dependency. 2. Pkg_libchk didn't catch the case. Did it happen by design? WBR -- bsam ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
On Wed, 17 Jun 2009 12:16:46 +0300 Andriy Gapon a...@icyb.net.ua wrote: After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Sporadic - because they appear when there are no parallel execution of any portupgrade tools (or any other package/port related tools). They appear in situations wheere they never appeared before. I see this on two stable/7 systems. Got the same on five 7-Stable Machines. -- Tobias Lott ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Safe to run squid_user=root ?
On Tue, 16 Jun 2009 22:48:17 -0500
Jeffrey Goldberg jeff...@goldmark.org wrote:
www/squid30 sets up an rc.d startup script that includes
squid_user=${squid_user:-squid}
This makes it impossible to get squid to listen on a port lower than
1024.
If I specify
squid_user=root
in my rc.conf will I be doing something stupid? Does squid
appropriately drop privileges after binding to a socket?
I don't use squid 3.x but it does in squid 2.x, look
for cache_effective_user and cache_effective_group in the default
configuration file.
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
Tobias lott writes: After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Sporadic - because they appear when there are no parallel execution of any portupgrade tools (or any other package/port related tools). They appear in situations wheere they never appeared before. I see this on two stable/7 systems. Got the same on five 7-Stable Machines. And on -Current from April. So, not a function of the OS version. Mine happens every time I run portupgrade/portversion, even when prior runs completed successfully. It seems harmless, but it would be Really Nice if it were fixed. Robert Huff ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
В Wed, 17 Jun 2009 13:32:31 +0200 Tobias lott tl...@gamesnet.de пишет: TL TL TL On Wed, 17 Jun 2009 12:16:46 +0300 TL Andriy Gapon a...@icyb.net.ua wrote: TL TL TL After recent massive ports update (I think ruby was touch and TL probably portupgrade too) I started getting seemingly sporadic TL Stale lock file was found. Removed. messages. What's TL interesting is that those messages are produced on stdout, not TL stderr. TL TL Sporadic - because they appear when there are no parallel TL execution of any portupgrade tools (or any other package/port TL related tools). They appear in situations wheere they never TL appeared before. TL TL I see this on two stable/7 systems. TL TL TL Got the same on five 7-Stable Machines. TL +2 -CURRENT -- wbr, tiger ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: pkg_libchk: a missing library is not detected
Boris Samorodov wrote: Dominic Fandrey kamik...@bsdforen.de writes: Luck. The app linking to the old library will have a dependency on the old version. pkg_add will find the origin, issue a warning about app-1.0 needing lib-0.1 but lib-0.2 is installed and proceed. app will not start, because of the missing library. I've never had this case. I've got the impression that pointyhead rebuilds all dependencies. Let me remind you the case with my original question: 1. Graphics/eog was broken (as you said) by indirect dependency. 2. Pkg_libchk didn't catch the case. Did it happen by design? WBR Are you certain that this was a direct dependency? Maybe just a dependency of graphics/eog was broken. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Put your creativity and passion to work at Art Institute Online
3DA [1]Train for a= creative career at The Art Institute of Pittsburgh - Online Division. Find An Online Program Now! [2]View in Browser |= [3]Unsubscribe style= font-family:Arial, Helvetica, sans-serif; f= ont-size:10px; color:#000; text-decoration:noneRemove My Email 3D= [4]= 3DNow [5]D= ear John, You have the talent, the imagination, and the passion to excel in a creative career. All you need is the right place to start. Get the creative skills you need to turn your talent into an exciting career from the leader in online creative arts education. Learn m= ore about The Art Institute of Pittsburgh - Online Division now. The Art Institute of Pittsburgh - Online Division offers online associate's degrees, bachelor's degrees, and diplomas in a variety of creative fields, including: [6]o Video Game Design= [7]o Animation [8]o Graphic Design [9]o Fashion [10]o Web Design[11]o And More! [12] Learn more about The Art Institute of Pittsburgh - Online Division now! [13][v=] [14]3DLearn [15]3DPrepare [AI_=] *Program availability varies by subje= ct The Art Institute of Pittsburgh Main Campus: 420 Boulevard o= f the Allies, Pittsburgh, PA 15219 This email was sent by: College.us.com 2711 N. Sepulveda Blvd. # 148, Manhattan Beach, CA 90266 You have subscr= ibed to receive information from us with po...@freebsd.org This email was sent to you because you requested we send you periodical inf= ormation about schools and colleges. To stop receiving email from Colle= ge.us.com, [16]Remove m= e or write us at the address listed above Please view our [17]Privacy Policy [jtdpwgwvg_sflhcbjclyl.gifo=] References 1. file://localhost/tmp/3Dht 2. 3Dhttp://www.new=/ 3. 3Dhttp://www.news.college.us.com/dbzgcqczc_sflhcbjclyl_lrkwdw 4. 3Dhttp://www.news.college.us.com/lrkdswskt_ 5. 3Dhttp://www.news.college.us.com/jt 6. 3Dhttp://www.news.college.us.com/jtdpwgwdj_ 7. 3Dhttp://www.news.college.us.com/gbpjqsqpj_ 8. 3Dhttp://www.news.college.us.com/pwkzqmqkm_ 9. 3Dhttp://www.news.college.us.com/aqvltktvq_ 10. 3Dhttp://www.news.college.us.com/yhdvjwjdb_ 11. 3Dhttp://www.news.college.us.com/yhdvjwjdz_ 12. 3Dhttp://www.news.college.us.com/tdchkjklk_ 13. 3Dhttp://www.news.college.us.com/dbzgcq 14. 3Dhttp://www.news.college.us.com/ohngvf 15. 3Dhttp://www.news.college.=/ 16. 3Dhttp://www.news.college.us.com/sfnhysydb_ 17. =http://www.news.college.us.com/eswmrtrnm_sflhcbjclyl_ueqwbwjf.html___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
Robert Huff roberth...@rcn.com writes: Tobias lott writes: After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Sporadic - because they appear when there are no parallel execution of any portupgrade tools (or any other package/port related tools). They appear in situations wheere they never appeared before. I see this on two stable/7 systems. Got the same on five 7-Stable Machines. And on -Current from April. So, not a function of the OS version. Mine happens every time I run portupgrade/portversion, even when prior runs completed successfully. It seems harmless, but it would be Really Nice if it were fixed. Not *completely* harmless; running separate portupgrade processes in parallel is pretty risky at the moment... ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Firefox 3.5 RC1 ready for testing!
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Firefox 3.5 RC1 ready for testing! Howdy, Yesterday FF3.5 RC1 was released! Unfortunately RC1 need a newer version of devel/nspr, we already updated devel/nspr and it seems to be work all fine. If you like to play with both updates you can try following patch [1]. I need to make a exp-run request for this update to make sure nothing will be failed. [1] http://miwi.homeunix.com/patches/ff3.5rc1.diff Please let us know if you see any problems. - - Martin - -- +---+---+ | PGP: 0xB1E6FCE9 | Jabber : miwi(at)BSDCrew.de | | Skype : splash_111 | Mail : miwi(at)FreeBSD.org | +---+---+ | Mess with the Best, Die like the Rest! | +---+---+ -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.11 (FreeBSD) iEYEARECAAYFAko48kUACgkQdLJIhLHm/OkrLQCgqpmdaV63yNtulDpgUMZ/fqEh Yo8AoN6DY0Pa0cckqOkmfVWdCo8c0lwT =iJJn -END PGP SIGNATURE- ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
Lowell Gilbert writes: After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Got the same on five 7-Stable Machines. And on -Current from April. So, not a function of the OS version. Mine happens every time I run portupgrade/portversion, even when prior runs completed successfully. It seems harmless, but it would be Really Nice if it were fixed. Not *completely* harmless; running separate portupgrade processes in parallel is pretty risky at the moment... When was it officually cleared to do that? Last I knew it was do at your own risk. Robert Huff ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
Robert Huff roberth...@rcn.com writes: Lowell Gilbert writes: After recent massive ports update (I think ruby was touch and probably portupgrade too) I started getting seemingly sporadic Stale lock file was found. Removed. messages. What's interesting is that those messages are produced on stdout, not stderr. Got the same on five 7-Stable Machines. And on -Current from April. So, not a function of the OS version. Mine happens every time I run portupgrade/portversion, even when prior runs completed successfully. It seems harmless, but it would be Really Nice if it were fixed. Not *completely* harmless; running separate portupgrade processes in parallel is pretty risky at the moment... When was it officually cleared to do that? Last I knew it was do at your own risk. I don't think it was ever officially supported, but it was the purpose behind adding the lock files in the first place. I've never used it heavily, although I did put it through a fairly heavy wringer when it was under development. For real use, I've found it convenient when building an upgrade to a particularly large port (generally OpenOffice) while upgrading a large number of other ports as well. I don't worry about it being foolproof, because my build server has nothing fundamentally valuable on it. But at the moment, the locking seems to be completely broken, which raises it to a whole other level of risk that I can't be bothered to mess with. [Not that I generally care how long port builds take; that's computer time, not human time.] - Lowell ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: pkg_libchk: a missing library is not detected
Dominic Fandrey kamik...@bsdforen.de writes: Boris Samorodov wrote: Dominic Fandrey kamik...@bsdforen.de writes: Luck. The app linking to the old library will have a dependency on the old version. pkg_add will find the origin, issue a warning about app-1.0 needing lib-0.1 but lib-0.2 is installed and proceed. app will not start, because of the missing library. I've never had this case. I've got the impression that pointyhead rebuilds all dependencies. Let me remind you the case with my original question: 1. Graphics/eog was broken (as you said) by indirect dependency. 2. Pkg_libchk didn't catch the case. Did it happen by design? Are you certain that this was a direct dependency? Maybe just a dependency of graphics/eog was broken. I didn't say it was a direct dependency. And the dependecy may be broken. Well, thanks for your help. Next time I'll try to have more info about a case. WBR -- bsam ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Firefox 3.5 RC1 ready for testing!
On Wed, Jun 17, 2009 at 03:40:21PM +0200, Martin Wilke wrote: Firefox 3.5 RC1 ready for testing! Howdy, Yesterday FF3.5 RC1 was released! Unfortunately RC1 need a newer version of devel/nspr, we already updated devel/nspr and it seems to be work all fine. If you like to play with both updates you can try following patch [1]. I need to make a exp-run request for this update to make sure nothing will be failed. [1] http://miwi.homeunix.com/patches/ff3.5rc1.diff Built and installed fine. Was running fine until I viewed the demo link from http://hacks.mozilla.org/2009/06/3d-transforms-isocube/ at which point firefox crashed. I'll see if I can get any details about the crash. -- WXS ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: acroread9 crashes after maybe 10 seconds of operation.
Hiroki Sato h...@freebsd.org wrote in 20090617.141203.07900852@allbsd.org: hr Robert Huff roberth...@rcn.com wrote hr in 18997.22089.242834.29...@jerusalem.litteratus.org: hr ro Um. hr ro Are there plans to get it to work with something more recent? hr ro I was under the (uninformed) impression linux_base-fc-4 was, ah, hr ro workable but no longer favored. hr hr The ports collection still assumes fc4 as the default, so I think it hr is the primary target. hr hr Anyway, I will try other configurations including one in your report. hr I guess the issue is due to some incomplete (or not-fully-compatible) hr compat-layer implementations of features available in Linux 2.6.x. I could reproduce the symptom (RSException), but this has also been reported on Linux: http://forums.adobe.com/message/1931692. The release note of 9.1.2 says it is solved but it remains as far as I can check. BTW, could you give it a try to set sysctl compat.linux.osrelease=2.4.2 and let me know if it works or not? -- Hiroki pgpM15Xi34PTj.pgp Description: PGP signature
Updating HPLIP port
Currently, the port version of HPLIP is 2.8.2, released on 02-09-2008. The current version is: 3.9.4b, released on 04-29-2009. It supports many additional features and corrects several bug in the older versions. http://hplipopensource.com/hplip-web/release_notes.html This is the tenth version released since the version presently in the ports tree. I have tried contacting the port maintainer, amis...@am-productions.biz however, he is probably busy since I have not heard from him. Is it possible that someone might be interested in updating this port. http://hplipopensource.com/hplip-web/downloads.html I tried it myself; however, it complains about a missing file. I probably need automake or something; however, my skills are not that sophisticated. Thanks! -- Jerry ges...@yahoo.com Matrimony isn't a word, it's a sentence. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: portupgrade/ruby issue? (Stale lock file was found. Removed.)
2009/6/17 Andriy Gapon a...@icyb.net.ua:
After recent massive ports update (I think ruby was touch and probably
portupgrade
too) I started getting seemingly sporadic Stale lock file was found.
Removed.
messages.
It is probably caused by ruby-1.8.7's bug fogetting to call finalizers.
This is filed as Bug #1556 (http://redmine.ruby-lang.org/issues/show/1556
... japanese page; please translate it ;)
Workaround for the case may be using at_exit instead of finalizers.
My sample fix is:
--- pkgdb.rb.orig 2009-04-14 16:56:38.0 +0900
+++ pkgdb.rb2009-06-18 01:49:19.0 +0900
@@ -103,8 +103,8 @@
#end
end
- def PkgDB.finalizer
-Proc.new {
+ def PkgDB.define_at_exit
+at_exit {
PkgDBTools.remove_lock(LOCK_FILE)
}
end
@@ -113,7 +113,7 @@
@db = nil
@lock_file = Process.euid == 0 ? LOCK_FILE : nil
@db_version = DB_VERSION
-ObjectSpace.define_finalizer(self, PkgDB.finalizer)
+PkgDB.define_at_exit
setup(*args)
end
--- portsdb.rb.orig 2009-04-14 16:56:38.0 +0900
+++ portsdb.rb 2009-06-18 01:49:04.0 +0900
@@ -153,8 +153,8 @@
end
end
- def PortsDB.finalizer
-Proc.new {
+ def PortsDB.define_at_exit
+at_exit {
PkgDBTools.remove_lock(LOCK_FILE)
}
end
@@ -163,7 +163,7 @@
@db = nil
@lock_file = Process.euid == 0 ? LOCK_FILE : nil
@db_version = DB_VERSION
-ObjectSpace.define_finalizer(self, PortsDB.finalizer)
+PortsDB.define_at_exit
set_ports_dir(alt_ports_dir)
set_db_dir(alt_db_dir)
set_db_driver(alt_db_driver)
--
Hiroto Kagotani
hiroto.kagot...@gmail.com
___
freebsd-ports@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-ports
To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
avoiding build dependency on docbook, etc. in new port
I'm porting some software that has a build dependency on docbook2man in order to generate its manpages from .docbook files. Testing the port in tinderbox takes a long time, most of it because of the need to build the docbook infrastructure. It seems a shame to use all that CPU time and install all those packages just to get ready to convert some manpages. What's the preferred approach in cases like this? Should I keep the build dependency on docbook2man et al, or should I put pregenerated copies of the manpages in the files directory of the port? ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
2 Costa Rican Gems!
Just 119,000 for a new 3-bedroom bungalow! Visit: http://ucanaffordit.com/luzsol for details. Just 159,000 for a new 2 bedroom on beach alley! Furnished! Visit: http://sandysunnycr.com/luzsol for more details. Buy now,values are rising once again in Costa Rica. RCI SA Avenida 7 Calle 14 200 metros norte del parqueo Ronald San Jose, Costa Rica 877-900-1188 unsubscribe: http://ucanaffordit.com/unsubscribe ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: Firefox 3.5 RC1 ready for testing!
Wesley Shields wrote: On Wed, Jun 17, 2009 at 03:40:21PM +0200, Martin Wilke wrote: Firefox 3.5 RC1 ready for testing! Howdy, Yesterday FF3.5 RC1 was released! Unfortunately RC1 need a newer version of devel/nspr, we already updated devel/nspr and it seems to be work all fine. If you like to play with both updates you can try following patch [1]. I need to make a exp-run request for this update to make sure nothing will be failed. [1] http://miwi.homeunix.com/patches/ff3.5rc1.diff Built and installed fine. Was running fine until I viewed the demo link from http://hacks.mozilla.org/2009/06/3d-transforms-isocube/ at which point firefox crashed. I'll see if I can get any details about the crash. That happened here too. Casual browsing has caused no problems so far though. -- Christian Laursen ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Earn More with an Online Degree
3DCOLLEGE.US.COM You a= re receiving this message because you submitted the email address = po...@freebsd.org to receive information from http://www.directscholar.com on Jun 09 2009 07:25AM. [1]Go here to add us to your address= book [2]View in Browser | [3]Unsubscribe st= yle=color:#66; text-decoration:noneRemove My Email [4]= img src=http://college.us.com/e/feb_2008/images/generic_01.jpg; border =0alt=Can't see images? Use link above to view in browser. [5][generic_0=] [6]Dear John: Earn more money with a career-fo= cused online program from a top accredited online school. Find a top school with an online pro= gram that fits your career goals here. We've compiled a list of well-respected, fully accredited online schools fo= r you to choose from -- so you can study on your schedule without givi= ng up work or personal obligations. These schools offer a variety of online certificate, associate's, bachelor= 's, and master's degree programs in a variety of high-demand fields includi= ng: o Business Administration o Criminal Justice o Hospitality, Travel and Tourism o Information Technology o Human Resources o Health Care o Paralegal Studies o And many more! Take your career -- and your salary -- to the next level toda= y! Learn more about our top online scho= ols now! [7][generic_04.=] [8]= img src=http://college.us.com/e/feb_2008/images/generic_05.jpg; border =0 Thi= s email was sent by: College.us.com 2711 N. Sepulveda Blvd. # 148, Manhattan Beach, CA 90266 You have subscribed to receive information from us with po...@freebsd.org This email was sent to you because you requested we send you periodic= al information about schools and colleges. To stop receiving email from College.us.com, [9]Remove me or write us at the address li= sted above Please view our [10]Privacy Policy [mylsvkfqf_kwblkrwkbmb.gifo=] References 1. 3Dhttp://www.news.college.us.com/njzhrqnhn_kwblkrwkbmb_r 2. 3Dhttp://www.news.college.us.com/mylsvkfss_kwblkrwkbmb_ 3. 3Dhttp://www.n=/ 4. 3Dhttp://www.news.college.us.com/hgsbwdpbd_ 5. 3Dhttp://www.news.college.us.=/ 6. 3Dhttp://www.news.college.us.com/tdchkbzhd_ 7. 3Dhttp://www.news.college.us.co=/ 8. 3Dhttp://www.news.college.us.com/vptzcgfjt_ 9. 3Dhttp://www.ne=/ 10. 3Dhttp://www.news.college.us.com/mylsvkfqm freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: avoiding build dependency on docbook, etc. in new port
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: I'm porting some software that has a build dependency on docbook2man in order to generate its manpages from .docbook files. Testing the port in tinderbox takes a long time, most of it because of the need to build the docbook infrastructure. It seems a shame to use all that CPU time and install all those packages just to get ready to convert some manpages. What's the preferred approach in cases like this? Should I keep the build dependency on docbook2man et al, or should I put pregenerated copies of the manpages in the files directory of the port? Hi Charlie, I feel your pain! I don't think there's any problem pre-generating the man page and keeping it in the port's files/ directory. On the other hand, I recently went through this experience and went a different route. YMMV! I took over maintainership for security/logcheck a while back, and it used docbook2man to create its one (!) man page. After a while, I got some requests to strip out that dependency. A user was kind enough to point me to docbook2X: http://docbook2x.sourceforge.net/ This is a much lighter-weight tool for converting docbook to manual pages, among other functions. I eventually added it to the ports tree and updated the logcheck port to use it. This worked well for me, but I started getting reports of mysterious build failures related to the docbook-man conversion process: http://unix.derkeiler.com/Mailing-Lists/FreeBSD/questions/2009-05/msg01466.html Since I couldn't figure out how to reproduce the problem here, a kind user granted me remote access to a machine that exhibited the problem. I finally figured out that docbook2x conflicts with docbook-4.1 and refuses to work correctly. The logcheck port now does some checks to determine which docbook-man conversion tool to use, instead of blindly installing docbook2x: http://www.freebsd.org/cgi/cvsweb.cgi/ports/security/logcheck/Makefile Cheers, Greg - -- Greg Larkin http://www.FreeBSD.org/ - The Power To Serve http://www.sourcehosting.net/ - Ready. Set. Code. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFKOVEA0sRouByUApARAoqdAKCtVoPrd4HYZ+7i0R9IpDwZT/h+XgCfdK0w J19UPAyWPAutPRQfrCEaiUA= =SJ89 -END PGP SIGNATURE- ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
FreeBSD Port: py25-fail2ban-0.8.3_1
Hello. As you know, many people are now using PF on their FreeBSD servers. I've made some changes to make it work with PF, and I think they are worth including in the port by default. I'm including the configuration needed to use Fail2Ban with PF on FreeBSD. Basically, you make a table in PF, and add to the table as the 'actionban' and remove the entry from the table as the 'actionunban'. Pretty simple. You can easily see a list of banned addresses using: sudo pfctl -t fail2ban -T show Thanks for the port of this great utility! - Chris -- Chris Jones CCNP, JNCIA-M Senior Systems Manager Pittsburg State University E-mail: cjo...@pittstate.edu Phone: 1.620.235.4158 -- The production of too many useful things results in too many useless people. -Karl Marx FAIL2BAN EDITS FOR PF ON FREEBSD Chris Jones - 2009.06.17 ./jail.conf: # PF jail [ssh-pf] enabled = true filter = sshd action = pf sendmail-whois[name=SSH, dest=em...@domain.com] logpath = /var/log/auth.log ./action.d/pf.conf: [Definition] actionstart = actionstop = actioncheck = actionban = pfctl -t fail2ban -T add ip actionunban = pfctl -t fail2ban -T delete `pfctl -t fail2ban -T show 2/dev/null | grep ip` [Init] port = ssh localhost = 127.0.0.1 /etc/pf.conf: table fail2ban persist block in on $ext_if from fail2ban ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: avoiding build dependency on docbook, etc. in new port
Greg Larkin escribió: This is a much lighter-weight tool for converting docbook to manual pages, among other functions. I eventually added it to the ports tree and updated the logcheck port to use it. Another idea: generating man pages directly with XSLT stylesheets. Actually, these docbook2man and docbook2X do the same I think but they hide the details. To use the XSL stylesheets you need docbook-xsl and libxslt and calling xsltproc with the proper parameters. It's pretty easy. Cheers, -- Gabor Kovesdan FreeBSD Volunteer EMAIL: ga...@freebsd.org .:|:. ga...@kovesdan.org WEB: http://people.FreeBSD.org/~gabor .:|:. http://kovesdan.org ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: avoiding build dependency on docbook, etc. in new port
On Wed 17 Jun 2009 at 13:24:32 PDT Greg Larkin wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: I'm porting some software that has a build dependency on docbook2man in order to generate its manpages from .docbook files. Testing the port in tinderbox takes a long time, most of it because of the need to build the docbook infrastructure. It seems a shame to use all that CPU time and install all those packages just to get ready to convert some manpages. What's the preferred approach in cases like this? Should I keep the build dependency on docbook2man et al, or should I put pregenerated copies of the manpages in the files directory of the port? Hi Charlie, I feel your pain! I don't think there's any problem pre-generating the man page and keeping it in the port's files/ directory. Actually, this is a suite of tools and there are eleven manpages. On the other hand, I recently went through this experience and went a different route. YMMV! I took over maintainership for security/logcheck a while back, and it used docbook2man to create its one (!) man page. After a while, I got some requests to strip out that dependency. A user was kind enough to point me to docbook2X: http://docbook2x.sourceforge.net/ This is a much lighter-weight tool for converting docbook to manual pages, among other functions. I eventually added it to the ports tree and updated the logcheck port to use it. Now I'm confused. When I used pkg_info to find out what package had installed docbook2man on my system, it said it was part of docbook2X-0.8.8_2. It was after I added textproc/docbook2X as a build dependency that I started seeing a bunch of docbook stuff getting installed in my tinderbox. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: avoiding build dependency on docbook, etc. in new port
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: On Wed 17 Jun 2009 at 13:24:32 PDT Greg Larkin wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: I'm porting some software that has a build dependency on docbook2man in order to generate its manpages from .docbook files. Testing the port in tinderbox takes a long time, most of it because of the need to build the docbook infrastructure. It seems a shame to use all that CPU time and install all those packages just to get ready to convert some manpages. What's the preferred approach in cases like this? Should I keep the build dependency on docbook2man et al, or should I put pregenerated copies of the manpages in the files directory of the port? Hi Charlie, I feel your pain! I don't think there's any problem pre-generating the man page and keeping it in the port's files/ directory. Actually, this is a suite of tools and there are eleven manpages. On the other hand, I recently went through this experience and went a different route. YMMV! I took over maintainership for security/logcheck a while back, and it used docbook2man to create its one (!) man page. After a while, I got some requests to strip out that dependency. A user was kind enough to point me to docbook2X: http://docbook2x.sourceforge.net/ This is a much lighter-weight tool for converting docbook to manual pages, among other functions. I eventually added it to the ports tree and updated the logcheck port to use it. Now I'm confused. When I used pkg_info to find out what package had installed docbook2man on my system, it said it was part of docbook2X-0.8.8_2. It was after I added textproc/docbook2X as a build dependency that I started seeing a bunch of docbook stuff getting installed in my tinderbox. Hi Charlie, My apologies - when you said that TB had to build a lot of packages to support docbook2man, I assumed that you were talking about the textproc/docbook-to-man port, not the actual docbook2man executable provided by textproc/docbook2X. Originally, textproc/docbook-to-man required 100+ dependencies to build. It's a much smaller number now, due to some recent changes in the xorg infrastructure: http://www.freebsd.org/cgi/cvsweb.cgi/ports/textproc/docbook-to-man/Makefile#rev1.15 You can see the difference here: fbsd70# cd /usr/ports/textproc/docbook-to-man make USE_XLIB=yes all-depends-list | wc -l 114 fbsd70# cd /usr/ports/textproc/docbook-to-man make all-depends-list | wc -l 28 fbsd70# USE_XLIB=yes used to be implied by USE_IMAKE=yes, but not any longer. Anyway, docbook2X has even fewer deps: fbsd70# cd /usr/ports/textproc/docbook2X make all-depends-list | wc -l 20 fbsd70# To me, that doesn't seem like a lot compared to the 100+ I was dealing with before. But if it's still too many to deal with, Gabor's suggestion might be better. Cheers, Greg - -- Greg Larkin http://www.FreeBSD.org/ - The Power To Serve http://www.sourcehosting.net/ - Ready. Set. Code. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFKOYrY0sRouByUApARAlAUAKCY3JcCYDLnBeQAqtp9EbqVBmwf6gCgiiqH /SYoQjGRUr8fqrruW+incQg= =nPVF -END PGP SIGNATURE- ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
k3b seems to have aquired a problem.
It has probably been a month since I have used k3b. k3b is a program that has worked for me for maybe 5 years or more without any problems. I just tried it on my laptop running up to date current (cvsup and build today) all ports up to date and using the kde menu I just got a the splash k3b box and nothing so I tried the command line and only get the following: /root # k3b kdecore (KAction): WARNING: KActionCollection::KActionCollection( QObject *parent, const char *name, KInstance *instance ) and then a new prompt. With ps -ax|grep k3b I see k3b running but nothing happens. A killall k3b kills the process and the splash graphic. Does anyone have any idea what this could be? Thanks, ed P.S. I did a portmaster sysutils/k3b and rebuilt but with the same results. My kde3 versions are all 3.10xx. KDE doesn't show any problems what so ever. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
Re: avoiding build dependency on docbook, etc. in new port
Greg Larkin wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: On Wed 17 Jun 2009 at 13:24:32 PDT Greg Larkin wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Charlie Kester wrote: I'm porting some software that has a build dependency on docbook2man in order to generate its manpages from .docbook files. Testing the port in tinderbox takes a long time, most of it because of the need to build the docbook infrastructure. It seems a shame to use all that CPU time and install all those packages just to get ready to convert some manpages. What's the preferred approach in cases like this? Should I keep the build dependency on docbook2man et al, or should I put pregenerated copies of the manpages in the files directory of the port? If they are too bulky for files/, you could always compress the manpages and add them via something like: .if !defined(NO_INSTALL_MANPAGES) DISTFILES+= thisportsmanpages.tar.gz .endif ... adding the appropriate target to install them, after placing thisportsmanpages.tar.gz on the FreeBSD or project servers. snip To me, that doesn't seem like a lot compared to the 100+ I was dealing with before. But if it's still too many to deal with, Gabor's suggestion might be better. As an example of the more economical approach that I think Gabor was suggesting, and which also respects NO_INSTALL_MANPAGES, see how pgj@ handled the manpage for devel/cppcheck (incl. files/patch-Makefile ). b. ___ freebsd-ports@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-ports To unsubscribe, send any mail to freebsd-ports-unsubscr...@freebsd.org
