On Thursday 30 December 2004 01:25, D. Hugh Redelmeier wrote: > Context: Fedora Core 3 dropped LICQ. I'd like to have it on FC3. So > I'm exploring how to get it. Up until now, I know nothing about LICQ > development. I'm even going to report trivial problems since you are > only ignorant once. > > Although all I'm doing is complaining, I'm very grateful for all the > work you have put into this. It's just that there is no need to talk > about what works. > > 1) http://licq.wibble.net/faqs.php displays as raw HTML on my firefox. > This is true of other pages on this site. > http://www.licq.org/faqs.php displays properly. > I don't know why and have not investigated.
Hm, a broken mirror :( > 2) http://www.licq.org/contact.php does not list the mailing lists. > I think that it should. Good idea, i fixed it already. > 3) [Jon has fixed this one.] > http://www.licq.org/help.php has a broken link to the "Licq Mailing > List": http://www.licq.org/info.html#lml > > I found this list via google. > > 4) Fedora Core 3 on x86-64 can also run FC3 binaries for regular x86. I > tried licq-1.3.0-1.fc2.i386.rpm but it would not install -- missing > dependencies. That is reasonable (not a bug). > > 5) I downloaded licq-1.3.0-1.fc2.src.rpm in the hope that I could rebuild > from source. This file appears to be corrupt: > > $ rpm --checksig licq-1.3.0-1.fc2.src.rpm > licq-1.3.0-1.fc2.src.rpm: sha1 MD5 NOT OK > > [Jon says that RPMs are not really supported -- user submissions > are posted. In that case, I suggest that this RPM should be > removed because it is broken. That probably means that the > binary RPM should be removed too since it would be then hard > to conform with the GPL for the binary RPM.] > > 6) I'm building 1.3.0 from the tarball. doc/README says to do > configure and make as root. It would be better if one could > do that as an ordinary user. I would bet that that would work. > Only "make install" is likely to need root. That's true. I just fixed it in CVS. > 7) The "make" yielded several warnings. I think that the code should > be fixed. Warning: I'm not a C++ programmer. > > icqd-tcp.cpp:429: warning: int format, different type arg (arg 3) > users.size() probably yields a size_t, not an int. > One easy fix: cast users.size() to int. > Better fix: replace %d with something more appropriate. > > utility.cpp:241: warning: int format, different type arg (arg 4) > message.cpp:762: warning: int format, different type arg (arg 3) > icqd-chat.cpp:1098: warning: int format, different type arg (arg 4) > These are probably all like the previous one > > > countrycodes.c:287: warning: implicit declaration of function > `strcasecmp' Missing #include <string.h> > > rtf.ll:181: warning: unused variable 'c' > rtf.ll:191: warning: unused variable 'c' > Uses of these variables are commented out. > Perhaps these definitions should be commented out as well > (the lines are adjacent). Hm, we have to look at this. > 8) In building plugins/qt-gui, I got the following warning: > > In file included from catdlg.cpp:147: > catdlg.moc:135:22: warning: no newline at end of file > > Actually, catdlg.moc does end in a newline. > catdlg.cpp does not end in a newline. This should be fixed. > And then so should GCC -- its message is confused. Ok, this is now fixed. > I seem to have a working LICQ! To actually test it, I'll need to > await my user. > > It isn't clear how to migrate an old .licq to 1.30. Certainly, a pure > copy does not work (hard to copy a FIFO). Also, the plugin names are > in a configuration file, and they've changed. Hm, no, they did not change. Licq does update the configuration files automatically when it detects old config-files. Well, have a happy new year, i'm off now ;-) Thanks for all your reports! Bye, Thomas -- email: [EMAIL PROTECTED] IRC: DreamForce on #licq UIN: 75450207 (urgent messages only) -- Gegen Software-Patente! http://webshop.ffii.org/ Bitte informieren Sie sich!
pgpqL14dimySQ.pgp
Description: PGP signature
