V5.28.00.patches with Qt backend dies in font / fontconfig

ROOT
21

VIVE L’AMOUR!
on your demand of “Wed Feb 16, 2011 1:58” in this thread, I “complemented” my existing ROOT installation (built with “–enable-qt --enable-qtgsi”, installed into “/…/root/v5-28-00-patches/”) with your qtFullRoot (built according to point “3.” of your “How2Install4Unix.html”, installed into a dedicated subdirectory “/…/qtRoot/”).
Afterwards, I built your “qtRoot/qtExamples” (according to the point “4. Test” of your “How2Install4Unix.html”) and ran the newly built tests. This is described in my post of “Wed Feb 16, 2011 14:52” in this thread.
As shown there, the “CustomCanvasMenu” example dies right after start, while the “HelloWord” example exhibits the same problem with “new TBrowser()” as the “standard root”. The latter has been described in my post of “Tue Jan 11, 2011 18:03” in this thread (note that I explicitly say there that “new TCanvas()” is NOT sufficient to demonstrate the existing problem).
The presented “output” of my trials always shows what I run, how I run it and what I get (and the included “stack traces” show full paths to the involves shared libraries, in case somebody has any concerns about LD_LIBRARY_PATH).
Once again, there is no need to install qtFullRoot in order to see the problem. The “standard root” configured with “–enable-qt --enable-qtgsi” (and if you’re really hard try also to “–enable-builtin-freetype”), plus a “.rootrc” containing “Gui.Backend: qt” and “Gui.Factory: qt” is sufficient (see my first post in this thread of “Sun Jan 09, 2011 18:20” and then of "Tue Jan 11, 2011 14:16 " and finally of “Tue Jan 11, 2011 18:03”).
If nobody else sees this problem now, I consider this thread closed. I can wait. :mrgreen:
I am stupid. No?
Pepe Le Pew.

P.S. Just to make it explicit … here’s the current “configuration” of my root (v5-28-00-patches, checked out revision 38007 on 2011.02.08 at around 17:44 UTC, on a Ubuntu 10.04.2 LTS i686 with gcc 4.4.3) …

Enabled support for asimage, astiff, builtin_afterimage, cintex, clarens, editline, exceptions, explicitlink, fftw3, gviz, gdml, genvector, gsl_shared, krb5, ldap, mathmore, memstat, minuit2, mysql, opengl, peac, pythia6, python, qt, qtgsi, reflex, roofit, ruby, rpath, shadowpw, shared, soversion, ssl, table, tmva, unuran, x11, xft, xml, xrootd.

22

[quote=“Pepe Le Pew”]Once again, there is no need to install qtFullRoot in order to see the problem[/quote]This is not what I said. I said I do not see the problem. I recommended you the full installation to save the time and eliminate the guess job . So far you opted to ignore it. [quote=“Pepe Le Pew”]If nobody else sees this problem now, [/quote]I do agree that the installation of ROOT with option you originally applied should be sufficient and something has to be fixed . However, I have no your env. to reproduce your “use case” and to see your problem. [quote=“Pepe Le Pew”]I consider this thread closed. I can wait. [/quote]It is not clear what you are going to wait for? Do you need that functionality or not ? If you need the working version you should try to apply the recommended approach. It had helped many others. (just search the forum). It should help you too. [quote=“Pepe Le Pew”] I “complemented” my existing ROOT installation (built with “–enable-qt --enable-qtgsi”, installed into “/…/root/v5-28-00-patches/”) with your qtFullRoot (built according to point “3.” of your “How2Install4Unix.html”, installed into a dedicated subdirectory “/…/qtRoot/”).[/quote]I am sorry, that the page is not clear enough yet. What you did is NOT correct. Speaking “complement” I meant the ROOT installation that was done with Qt disabled. The “complement” means adding the Qt-layer to the “Qt-less” ROOT installation.[quote=“Pepe Le Pew”]I am stupid. No?[/quote]Try to apply the full installation. As I said it ought to eliminate a lot of guess job (See your own example above) . The script was written by QtROOT users, it have been polished by many other QtROOT users for the new generation of QtRoot users.