[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Please consider patching upstream bug 172182 in kdelibs5

Hi again,

I noticed the bug was patched in a recent upload. Thanks!

Unfortunately, because of this I've discovered potentially trickier hiccup 
with the cache that affects anyone still running KDE 3 apps inside a KDE 4 
desktop, like I do (Amarok and K3B mostly, in my case). The two different 
cache cleaner versions (from 3 and 4) end up duking it out in a battle to 
rule the cache folder, i.e. they periodically nuke each other's cache files.

It seems fine while only KDE 4 apps are running, but I noticed that while I 
was running Amarok, kdeinit (3) was running, and my cache was getting 
cleared, so I'm pretty sure this is the reason.

I'm not sure how you would go about solving this issue. Perhaps block KDE 3 
cleaner from running while KDE 4 kdeinit is active, or maybe just use a 
different cache location.

Thought you should know.


Brendon Higgins wrote (2008-11-07 11:38 am):
> Hi,
> I'd like to ask (nicely, of course) that the maintainers of KDE 4.1 in
> experimental apply the patch to fix a versioning issue in the http cache
> cleaner. It affects KDE since at least 4.1.2, and somehow missed out on
> being fixed in 4.1.3. The upstream bug report is here:
> https://bugs.kde.org/show_bug.cgi?id=172182
> The problem is a version mismatch between the http cache creator and the
> version the cache cleaner expects, resulting in the cache cleaner removing
> *all* files, and thus making all http caching completely worthless (and, I
> suspect, also causing weird browsing hiccups when files sometimes disappear
> (due to cleaning) *while* they are being downloaded).
> The fix is simply to bump the version that the cache cleaner expects.
> I was patiently waiting for this fix to come through in 4.1.3, but now that
> version is tagged I see it was neglected somehow. So I ask you.
> Keep up the good work.
> Peace,
> Brendon

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply to: