pyicq-t-0.8.1 fails with legacyjit xdb driver

[GoogleCodeExporter]

What steps will reproduce the problem?
- Try to use pyicq-t-0.8.1 with the legacyjit xdb driver backend

pyicq-t will silently fail to show up as online for any registrated client
and is as such unusuable.
A traceback gets lost somewhere in the dark, it does not end up in the log
and is not shown via Jabber either. strace reveals it:

[...]
exceptions.AttributeError: XDB instance has no attribute 'getCSettingList'
(full traceback is in attached file)

So the reason for it seems to be that some new code uses the functions
getCSettingList, getCSetting and setCSetting, which have been implemented
for at least mysql und xml backends, but not for legacyjit.
I've added dummy implementations (same as for the other *Setting functions
as well) and it works now. This only fixes legacyjit, other backends
probably fail the same way.
See attached patch.

Original issue reported on code.google.com by hoffi...@gmail.com on 4 Jan 2009 at 4:23

Attachments:

• pyicqt-0.8.1-legacyjit-traceback.log
• pyicqt-0.8.1-legacyjit-compat.patch

[GoogleCodeExporter]

Hi!
legacyJIT stores data in same format as JIT. It necessary only for migration. 
It not
supports all xmlfiles's features. Don't use it always when it's possible.
It's no kind of bug. This driver non-native for PyICQt and we don't should break
internal structure of JIT's spool

Original comment by r000ns...@gmail.com on 5 Jan 2009 at 2:57

• Changed state: WontFix

[GoogleCodeExporter]

I don't quite understand your decision here. What I do understand is that the 
legacy
JIT format cannot support all new features, and I'm not demanding that either. 
I just
wished that I could keep using the legacy jit format and that it does not break 
my
whole setup in a minor version update, that's why I came up with the patch.

Also, a similar case (getSetting, setSetting, setSettingList) is not supported 
by the
format either, but there are dummy functions in place as well to avoid 
breakage, so I
don't understand why you refuse to add these dummies for getCSetting* functions.

That's why I'd kindly ask you to rethink your decision.

Original comment by hoffi...@gmail.com on 11 Jan 2009 at 2:54

[GoogleCodeExporter]

Ok. I will add dummy functions to legacyjit backend and to other too

Original comment by r000ns...@gmail.com on 12 Jan 2009 at 8:17

• Changed state: Accepted

[GoogleCodeExporter]

This has been fixed in 0.8.1.1, right? At least the code looks like it and it 
started
working for me again without patching around. :)

So thanks for fixing that. ;)

Original comment by hoffi...@gmail.com on 20 Jan 2009 at 6:13

[GoogleCodeExporter]

Original comment by r000ns...@gmail.com on 12 Feb 2009 at 5:41

• Changed state: Fixed