Closed GoogleCodeExporter closed 9 years ago
Verified on both windows and linux, using current 0.2.3 builds.
Original comment by ak.hep...@gmail.com
on 23 Sep 2009 at 2:19
Here's a patch for this that seems to work.
Only tested against current svn 321 on linux with pidgin 2.6.1
If windows has a different method of defining/printing ULL, then it'll have to
be
fixed for that build.
Original comment by ak.hep...@gmail.com
on 23 Sep 2009 at 3:19
Attachments:
found the gaim-extprefs.sf.net winbuild pidgin environment package for cygwin,
so
installed that.
However, either cygwin is truncating the unsigned_long_long id, or something
else is
occurring on the platform which is keeping this same fix from working.
Could just be me, tho.
Original comment by ak.hep...@gmail.com
on 23 Sep 2009 at 5:07
might be better using gint64 and guint64 instead.
Original comment by eionrobb
on 23 Sep 2009 at 5:09
Approximately when is this going to be fixed?
I don't mean to be rude, but this makes mbp completely unusable. This should
definitely have critical priority.
Original comment by gfn...@gmail.com
on 23 Sep 2009 at 7:41
Reporting that this is also happening on Windows.
Is there anything I can do to help speed up the patch for this? It started
happening
sometime last night.
Original comment by patricia...@gmail.com
on 23 Sep 2009 at 8:44
Twitpocalypse good name for this problem :(
started on my windows xp yesterday... pidgin is running out of ram.
One big freeze after 30 min and not rely usable.
Hope u can fix this soon,
love and need microblog-purple much,
so btw THANKS for your great work!
Original comment by xdan...@gmail.com
on 23 Sep 2009 at 9:45
Same issue started for me about 14 hours ago on Windows. Just shutting it down
until
a fix is available.
Original comment by foyle...@gmail.com
on 23 Sep 2009 at 12:00
Terribly sorry, I'm out of office for the whole day so I missed this
"Twitpocalypse2" :).
Will integrate ak.hepcat patch very soon.
Original comment by somsaks
on 23 Sep 2009 at 1:25
Patch landed in SVN. Will release new version soon.
@ak.hepcat BTW, by checking the patch, it seems to me that you have fix the
timeline_id and twitter count. From my understanding, it seems both of this are
constant that woudn't change over time isn't it? Are there any other
side-effect of
these 2 variables?
The twitter id was changed to unsigned long long for a long time ago
(Twitpocalypse1).
Original comment by somsaks
on 23 Sep 2009 at 2:11
it looked like there were a couple places that got missed and were still
'unsigned
long' only, so this was just catching those up.
anything that looked like an ID from twitter i bumped up to guint64, just to
play it
safe.
what's odd is that this patch works under linux, but not under windows -- at
least
not from my cygwin build environment. I don't know if that's due to the limited
support cygwin has for guint64-equiv data sizes or not.
Hopefully somebody has a non-cygwin Win32 build environment and can test this
theory out.
Or figure out what i've done wrong... ;-)
Also, sed 's/unsigned long long/guint64/g' has no effect, as guint64 is just a
typedef for the first.
Original comment by ak.hep...@gmail.com
on 23 Sep 2009 at 3:48
I bet a number of Windows users would be happy to give the fixed version a shot
if
you'll compile it.
Original comment by edwardah...@gmail.com
on 23 Sep 2009 at 5:20
Edward-
Since i can't compile for windows outside of cygwin, and cygwin builds are
apparantly
broken, this doesn't help.
I can -assume- that the patch works for a non-cygwin Win32 build environment,
as it
does work for Linux build environments.
Original comment by ak.hep...@gmail.com
on 23 Sep 2009 at 6:46
Pardon the noobish question: How do I compile it for Windows then? ><
Original comment by patricia...@gmail.com
on 23 Sep 2009 at 11:42
I'm not a developer. just a user with a bit of tenacity. I don't know how
you'd
compile it for windows, outside of the cygwin/mingw environment, or if it's even
possible.
For the record, I downloaded the MS VisualStudio trial, and have had about 7
hells
worth of a time trying to make things compile before giving up.
Original comment by ak.hep...@gmail.com
on 24 Sep 2009 at 12:11
I've built it for Windows, but it doesn't fix the problem.
Original comment by eionrobb
on 24 Sep 2009 at 12:46
What's odd for me is that it doesn't happen on my Vista machine. Only on my XP
machine.
Original comment by tjahjono
on 24 Sep 2009 at 1:51
It seems to be the problem about unsigned long long size is not the same on all
platform. I will change
to use time instead of id.
Original comment by somsaks
on 24 Sep 2009 at 2:00
Issue 154 has been merged into this issue.
Original comment by somsaks
on 24 Sep 2009 at 6:09
You need to use %I64d instead of %llu.
Patch attached.
Original comment by Drah...@googlemail.com
on 24 Sep 2009 at 11:13
Attachments:
Revoked the patch: unstable.
Original comment by Drah...@googlemail.com
on 24 Sep 2009 at 1:27
The weird thing is that this is not happened on all people. At least for me it
works
ok (I'm using 0.2.3 on Windows right now, latest SVN on my Jaunty works fine
too).
Seems there's something weird with Mingw that prevent correct translation of
%llu.
Wonder which Linux are you guys using the cause the same problem (I'm using
Ubuntu
Jaunty + Windows Vista Home Premium).
Original comment by somsaks
on 24 Sep 2009 at 1:53
BTW, the windows build instruction is in README on Wiki
http://code.google.com/p/microblog-purple/wiki/README
Original comment by somsaks
on 24 Sep 2009 at 1:55
Might be a Windows XP only bug. 0.2.3 was working fine until a started
following the
turbine (lotro, doo) twitter yesterday.
My development platform is also jaunty.
Original comment by Drah...@googlemail.com
on 24 Sep 2009 at 2:20
Could be a Windows XP only problem. Friend and me both using XP and we're
having that
prob.
Original comment by patricia...@gmail.com
on 24 Sep 2009 at 2:29
That's weird, my Vista seems ok with snprintf+%llu.
BTW, I change to g_snprintf now. Please try the latest build and see if it's
work for
you (mine's ok).
Original comment by somsaks
on 24 Sep 2009 at 2:59
0.2.4 works on my XP install.
I think it was a MinGW bug, and the workaround you've included is functioning
correctly.
Original comment by ak.hep...@gmail.com
on 24 Sep 2009 at 4:20
Working great here too.
Original comment by foyle...@gmail.com
on 24 Sep 2009 at 4:36
Working great here on Win XP SP3 as well.
Original comment by gfn...@gmail.com
on 24 Sep 2009 at 6:28
XP SP2 is good to go.
Original comment by edwardah...@gmail.com
on 24 Sep 2009 at 6:36
XP SP2 works fine too. Thanks for the help guys!
Original comment by patricia...@gmail.com
on 24 Sep 2009 at 11:39
For some reason 2.4 doesn't seem to be working for me. Pidgin still crashes when
Twitter is activated. I don't know what's wrong with it.
Original comment by kamigoro...@gmail.com
on 25 Sep 2009 at 2:46
Scratch my last message. The new patch works fine. Apparently, if your Twitter
doesn't work after installing 0.2.4, then find your Twitter message logs and
delete
them all. That seemed to do the trick for me.
Original comment by kamigoro...@gmail.com
on 25 Sep 2009 at 3:44
Original comment by somsaks
on 26 Sep 2009 at 5:36
Installed the new 0.2.4 version and this seems to have fixed the issue. I'm no
longer receiving the 100s of tweets, but I also haven't received a tweet yet.
Will
post again if either happens.
Original comment by lerner.s...@gmail.com
on 29 Sep 2009 at 12:44
I'm Debian maintiner. I have the report of this problem from March 2010,
version of the microblog-purple 0.2.4, so I guess that it is not fixed after
all. The link to the bug is
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=573606.
Original comment by karolina...@gtempaccount.com
on 10 Jun 2010 at 10:49
Original issue reported on code.google.com by
eionrobb
on 22 Sep 2009 at 9:25