Spec URL: http://dl.atrpms.net/all/italc.spec SRPM URL: http://dl.atrpms.net/all/italc-1.0.9-4.src.rpm Description: iTALC is a use- and powerful didactical tool which lets you view and control computers in your labs and interact with students in a modern way. It supports Linux and Windows NT/2000/XP and it even can be used transparently in mixed environments.
[rpmbuild@rocky italc]$ rp -i italc-1.0.9-4.src.rpm italc.src: W: summary-not-capitalized intelligent teaching and learning with computers Summary doesn't begin with a capital letter. italc.src: E: unknown-key GPG#66534c2b The package was signed, but with an unknown key. See the rpm --import option for more information. 1 packages and 0 specfiles checked; 1 errors, 1 warnings. if I remember, will update as soon as koji is up
Thanks for spotting this. I'll fix the capitalization, but since it's minor I won't be repackaging yet, just consider it done :) On the GPG key: Since it is not yet reviewed it canot carry any official Fedora GPG key. Please ignore this warning, it will automagically go away once the package goes through bodhi etc.
$ rpmbuild --rebuild italc-1.0.9-4.src.rpm ... checking QTDIR... configure: error: *** QTDIR must be defined, or --with-qtdir option given error: Bad exit status from /var/tmp/rpm-tmp.uuZdGM (%build) RPM build errors: Bad exit status from /var/tmp/rpm-tmp.uuZdGM (%build)
I've no idea what I'm doing, so likely this is unfair, but I installed italc and italc-master rpms then: $ italc Loading socket Config module ... Creating backend ... Loading x11 FrontEnd module ... Failed to load x11 FrontEnd module. Loading simple Config module ... Creating backend ... desc table has been created. Loading socket FrontEnd module ... Starting SCIM as daemon ... desc table has been created. QInotifyFileSystemWatcherEngine::addPaths: inotify_add_watch failed: No such file or directory Mon Aug 25 21:56:53 2008: [warning] QFileSystemWatcher: failed to add paths: /tmp/ibus-ellson/ The splash screen was up, but the application was hung and had to be aborted with <ctrl>C Shouldn't it provide a configuration screen?
(In reply to comment #3) > $ rpmbuild --rebuild italc-1.0.9-4.src.rpm > ... > checking QTDIR... configure: error: *** QTDIR must be defined, or --with-qtdir > option given > error: Bad exit status from /var/tmp/rpm-tmp.uuZdGM (%build) This is due to presence of qt3-devel. Please uninstall qt3-devel or use a chroot builder like mock to build w/o qt3-devel. This should probably be addressed with a BuildConflicts: qt3-devel or patching the sources to properly use pkg-config. For the latter there is an upstream bug report here: ([ 2040452 ] should use pkg-config to locate qt4) https://sourceforge.net/tracker/index.php?func=detail&aid=2040452&group_id=132465&atid=724375 But in Fedora's build concept this is not surfacing as it will install only the packages mentioned as BuildRequires: and not qt3-devel.
(In reply to comment #4) > I've no idea what I'm doing, so likely this is unfair, but > I installed italc and italc-master rpms then: > > $ italc > Loading socket Config module ... > Creating backend ... > Loading x11 FrontEnd module ... > Failed to load x11 FrontEnd module. > Loading simple Config module ... > Creating backend ... > desc table has been created. > Loading socket FrontEnd module ... > Starting SCIM as daemon ... > desc table has been created. > QInotifyFileSystemWatcherEngine::addPaths: inotify_add_watch failed: No > such file or directory > Mon Aug 25 21:56:53 2008: [warning] QFileSystemWatcher: failed to add > paths: /tmp/ibus-ellson/ > > > The splash screen was up, but the application was hung and had to be aborted > with <ctrl>C > > Shouldn't it provide a configuration screen? Yes, that's what it does. What did you test it under, rawhide? I tested this on F9 and it worked OK. Note that ica must be running for italc (the master app) to be able to work. But usually you get a note after the splash screen that this needs to happen and the program bails out. Also I've never seen this verbose output even if ica is not running, usually there is no output on the console at all. Maybe it is firing up some QT components for the first time and you see a QT issue? If you didn't try this under F9, could you please do so? And we can dig into any rawhide problems in the next step.
Re #3: I removed qt3-devel, and a bunch of stuff that depended on it. I notice that even after a fresh login I still have QTDIR=/usr/lib64/qt-3.3 in my environment. Not sure where this is getting set? Its not getting set in my .bash* scripts. Anyway, same problem. I would have tried removing qt3, but this would have resulted in the removal of most of KDE. I tried a subshell and "unset QTDIR" before running rpmbuild, but this resulted in: checking QTDIR... /usr checking Qt includes... /usr/include checking for moc-qt4... no checking for moc... no configure: error: *** not found! Make sure you have Qt-devel-tools installed! error: Bad exit status from /var/tmp/rpm-tmp.TQpaPw (%build) RPM build errors: Bad exit status from /var/tmp/rpm-tmp.TQpaPw (%build) I can't find any "moc-qt4" rpms. I tried on a reasonably vanilla fedora-9 system, same problem.
/usr/bin/moc-qt4 is part of qt-devel, which is a build requirement. Does it exist on your system (ls /usr/bin/moc-qt4)? Is the qt-devel package damaged (rpm -V qt-devel)
QTDIR is being set by qt3, not qt3-devel (as I assumed), so indeed removing it for test building outside of chroots is not going to work. I patched up the specfile a bit to allow building even if qt3 bits are around, please use the following patch: @@ -1,7 +1,7 @@ -Summary: intelligent teaching and learning with computers +Summary: Intelligent teaching and learning with computers Name: italc Version: 1.0.9 -Release: 4%{?dist} +Release: 5%{?dist} License: GPLv2+ Group: System Environment/Daemons URL: http://italc.sourceforge.net/ @@ -46,6 +46,10 @@ %patch3 -p1 %build +# QTDIR may be set by a qt3 instance, remove it +# (italc shouldn't be checking for QTDIR in the first place, it's +# obsoleted in QT4) +unset QTDIR # docdir handling is broken in configure.in, but for good measure, # let's keep it where it should be %configure --docdir=%{_defaultdocdir}/%{name}-%{version}
Re #8: on rawhide, a fresh install of qt-devel installs a broken link: lrwxrwxrwx 1 root root 20 2008-08-26 06:44 /usr/bin/moc-qt4 -> ../../../bin/moc-qt4 I'll report this against qt...
Re #9: Tested that the patch worked, and that builds failed without it, on fc9.i386 and fc9.x86_64. Thanks. Re #6: OK, on fc9.i386 and fc9.x86_64 running italc from the command line immediately after a fresh install produces much more reasonable set of warning popups.
Re #10: The Rawhide bug was cause by bad qt,qt-devel rpms sometime in the past.
... oops hit enter too soon... The issue was solved in bug #460123
Re #6: But the problem with the italc executable from the freshly built rpms is still there on Rawhide: $ italc Loading socket Config module ... Creating backend ... Loading x11 FrontEnd module ... Failed to load x11 FrontEnd module. Loading simple Config module ... Creating backend ... desc table has been created. Loading socket FrontEnd module ... Starting SCIM as daemon ... desc table has been created. QInotifyFileSystemWatcherEngine::addPaths: inotify_add_watch failed: No such file or directory Tue Aug 26 16:08:00 2008: [warning] QFileSystemWatcher: failed to add paths: /tmp/ibus-ellson/
(In reply to comment #14) > Re #6: But the problem with the italc executable from the freshly built rpms is > still there on Rawhide: What about F9, does it work as advertised? One needs to have ica on all systems running, which is probably the largest issue on F9 and not completely solved (if you need remote access during gdm, kdm, ldm, ..., one needs to add several different autostart mechanisms for ica). If it works on F9, I'd suggest to to allow for rawhide to kickstart again. Since the infrastructure incident two weeks ago there was no update and rawhide was in the middle of going alpha to beta.
Yes, fc9 is fine, as far as I've progressed with it. See #11. Ok with allowing rawhide to catch up. I was just trying to record that the issue is still open.
Is this still on the agenda? I don't mind taking the review if that helps.
Sure it is, but maybe there are newer spec/srpms, I just looked it up and there are the latest: Spec URL: http://dl.atrpms.net/all/italc.spec SRPM URL: http://dl.atrpms.net/all/italc-1.0.9-6.src.rpm Note that I found some issues with multiple users using X (with fastuserswitch).
Again, is this still on the agenda? If so, I'll do the review. Otherwise, I'm pretty interested in getting iTALC packaged for Fedora.
It still is on the agenda, but I'm aging with it and someday either me or this review will die of age :( There are probably not that many educators or other people with needs that this package addresses. Having poured my pessimism into this report, you are welcome to review it!
Axel, I'll review this over the next few days if you're still happy to maintain it. We're looking at using it in the school I work for.
Although I'm cutting in my own flesh: A review point against italc as packaged is that by now libvncserver and x11vnc (almost) passed their reviews and should then be used instead of italc's own embedded code. You can go on and review all other parts and we should then concentrate on getting italc working against the other packages once they have stabilized and entered the repo. Thanks!
Hi there. I looked into the bundled libs issue and talked with upstream (Tobias Doerffel) about it. He's pretty confident that we won't manage to separate the bundled libraries from the code without a major code overhaul because many the libs are modified and tighly integrated with the code. Unfortunately he doesn't seem to be actively developing iTALC anymore. But he did confirm the project is not dead, it's just the lowest priority project for him right now.
Well, the school I'm working with ended up not wanting italc, so my motivation for an in-depth review has waned.
Sorry, that should be "motivation for *taking on* an in-depth review has waned".
Axel, it seems that your old src.rpm does not build on the current distros ( EL-6 and F14 scratch builds both failed as here: http://koji.fedoraproject.org/koji/getfile?taskID=2328766&name=build.log ) If you are still motivated to push italc forward, please fix it and I'll do the review.
Please clear the whiteboard if providing a package which builds.
Created attachment 473626 [details] italc spec file I have used an old spec file of Alex's with a few changes to successfully build italc on F14 x86_64. If someone would like to try to build and install the application, please do so.
*** Bug 717595 has been marked as a duplicate of this bug. ***
What about try it again? I think this kind of software it's a key factor in schools. If you don't like Italc, there is also another interesting (and similar) applications called Epoptes: http://www.epoptes.org
(In reply to q2dg from comment #30) > What about try it again? I think this kind of software it's a key factor in > schools. > > If you don't like Italc, there is also another interesting (and similar) > applications called Epoptes: http://www.epoptes.org I will package this one later per your request.
Hello. There's some news about it? Thanks!