Bug 31858 - stable bump for pygtk-0.6.11 and pygtk-2.0.0
|
Bug#:
31858
|
Product: Gentoo Linux
|
Version: 1.4_rc4
|
Platform: All
|
|
OS/Version: Linux
|
Status: RESOLVED
|
Severity: enhancement
|
Priority: P2
|
|
Resolution: FIXED
|
Assigned To: python@gentoo.org
|
Reported By: liquidx@gentoo.org
|
|
Component: Library
|
|
|
URL:
|
|
Summary: stable bump for pygtk-0.6.11 and pygtk-2.0.0
|
|
Keywords:
|
|
Status Whiteboard:
|
|
Opened: 2003-10-23 16:21 0000
|
hi alpha/sparc/ppc/amd64 devs,
i'm hoping to clean up pygtk and mark the latest version stable as it has many
bugfixes over the 1.99.x series especially with regards to threading.
can you please take a look at pushing pygtk-2.0.0 to stable and let me know if
there are any bugs?
also i would like it if pygtk-0.6.11 (which is in a different slot for gtk+-1.2)
could be marked stable as well as it has been in the tree for quite a while now
but seems to be neglected.
thanks
Reproducible: Always
Steps to Reproduce:
What are some good ways to test these are ok?
the best way would be to merge and run something that depends on pygtk like
net-www/straw, dev-util/meld or new-p2p/nicotine.
i think dev-util/meld would be the easiest because it doesn't have many dependencies.
2.0 already was stable on ppc and 0.6.11 is now too
Marked 0.6.11 as stable for sparc.
In testing for 2.0.0, i built net-www/straw, however it keeps dying with
bus errors before the GUI is loaded. I don't know if it's related to this
or not, but any debugging help would be appreciated :)
hrmm .. any sort of debugging output with the bus error? straw does use some
rather aggressive asynchronous networking modules, but i've never seen this
before.
Nothing else shows up when starting it via an xterm, Just the bus error.
Does python have an equivalent to sh -x or bash -x?
try changing the first line in /usr/bin/straw to:
#!/usr/bin/python2.2 -v
that should enable verbose mode
Sorry for the delay in response.
I've attached the output of running straw using python -v.
It looks like the problem is related to mxDateTime, which I don't really know anything about, but I'm guessing it's a module distributed with Python.
are you using dev-python/egenix-mx-base-2.0.5 ? basically, it should work with
all the egenix-mx-base (this provides mxDateTime). anyhow, the package is
rather sensitive to gcc optimisations. i've filtered it to ignore -O3 but there
maybe other flags it doesn't like.
although i haven't had any reports of problems since i added filter-flags in
the egenix-mx-base ebuilds. does :
python -c "import mxDateTime"
crash or abort?
i think an easier candidate to test out pygtk is dev-util/meld. it is a gui for
diff with some nice highlighting features, written in pure python.
Strangely enough, python doesn't think it can find it;
python -c "import mxDateTime"
Traceback (most recent call last):
File "<string>", line 1, in ?
ImportError: No module named mxDateTime
meld works fine, so i've stabilized pygtk-2.0.0 on sparc.
note to self: check before clicking "commit"
it should be python -c "import datetime"
thanks weeve for marking pygtk
No luck with python -c "import datetime" either, still can't find the module.
hrmm, that is odd, what are the files listed as installed for egenix-mx-base?
Attached a file list for egenix-ms-base
Both of these are marked stable on alpha
closing. thanks for all your help!