Community Packages

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#18374 - [qgis] and python support

Attached to Project: Community Packages
Opened by Claudio (kimera) - Wednesday, 17 February 2010, 16:27 GMT
Last edited by Dan Griffiths (Ghost1227) - Saturday, 13 March 2010, 19:38 GMT
Task Type Bug Report
Category Packages
Status Closed
Assigned To Ionut Biru (wonder)
Dan Griffiths (Ghost1227)
Architecture All
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 1
Private No

Details

Description: The previous update (qgis 1.3.0-2) and the latest one (qgis 1.4.0-1) lost phython bindings support. It means that a large amount of qgis plugins doesn't work, or better, is "lost in compilation" (f-tools and plugin-manager too). When I tried to recompile this package using PKGBUILD, cmake returns "Python bindings disabled due dependency problems!". Going in the src/qgis-1.4.0/cmake subdirectory, it turned out that Python cmake module can't find SIP package, recently updated to release 4.10.
So downgrading to SIP 4.9.3-1 and PyQt 4.6.2-1 I tried to recompile Qgis 1.4.0 again, and this time cmake enabled Phython binding support.
Sadly I have no clues about troubles with SIP.

This task depends upon

Closed by  Dan Griffiths (Ghost1227)
Saturday, 13 March 2010, 19:38 GMT
Reason for closing:  Fixed
Comment by Thomas Dziedzic (tomd123) - Tuesday, 09 March 2010, 07:40 GMT
PKGBUILD includes my changes.
sip.patch includes the workaround.

qgis-1.4.0-2.src.tar.gz fixes the problem (temporarily).

Read the patch on why it doesn't work originally. (hint: it's a base 10 vs 16 issue :S)

P.S. We need to contact upstream about this issue. (post link if you already contacted them)
Comment by Dan Griffiths (Ghost1227) - Tuesday, 09 March 2010, 09:03 GMT
I don't use it, so can't verify the patch. If someone can verify that this works (or better yet, post a link to an upstream report) I'll see about rebuilding it.
Comment by Thomas Dziedzic (tomd123) - Tuesday, 09 March 2010, 15:15 GMT
Ok, I made the patch when I was very tired. I think I should explain it a little more.

cmake compares 40700 to any version that is returned by the python function. Since the latest version is 4.10, the returned version is 40a00.
I think that cmake just thinks that 40a00 < 40700 and doesn't treat the number as hex.

If you do decide to use this patch, you should include sip>=4.7 in the depends and makedepends of the PKGBUILD. Since my patch completely ignores the version check.

Loading...