Open GoogleCodeExporter opened 9 years ago
Having the same issue. 0.3.0 is my first install, so I can't speak for any
past versions. Running Pidgin 2.7.1 on Windows 7.
Original comment by jpruit...@gmail.com
on 10 Jun 2010 at 5:12
The code is roughly the same on Windows (It's the same for Windows).
Are you guys in the country with DST?
Original comment by somsaks
on 10 Jun 2010 at 8:43
I'm in the US, so yes.
Original comment by jpruit...@gmail.com
on 10 Jun 2010 at 9:08
Also yes [Israel].
I remember that at first it wasn't working (0.2.something), then it was fixed
(I don't remember the version, but there was an issue for it), and now broken
again.
Original comment by mikeage
on 11 Jun 2010 at 4:04
I'am in Germany (with DST), having the same Problem. When starting Pidgin,
tweets from the past are displayed with full date/time timestamps, tweets from
the "future" are displayed with normal timestamps.
Original comment by sumeru...@googlemail.com
on 9 Jul 2010 at 11:08
Hi,
are you sure this is microblog-purple's fault?
Now the time where I am is 17:28, if I look at
https://twitter.com/#!/thecleancoders/status/80996231823888384 twitter tells me
this was "1 hour ago" on hovering over it using the mouse the page displays:
15:53 - twitter is off by half an hour...
pidgin displays this tweet (I started just now, so it is a fetch of an older
tweet) with timecode: 15:53:01
So: microblog-purple and twitter are showing me the same time for the tweet.
Original comment by andresen.nils@gmail.com
on 15 Jun 2011 at 3:34
https://twitter.com/#!/BUGabundo/status/81019677031141376
"12 Minutes ago"
(11:26:10 AM) BUGabundo: The noise of a thousand voices, overlapping
https://twitter.com/#!/EugeneKay/status/81022722473336832
"36 seconds ago"
(10:37:58 AM) EugeneKay: This is a test of the emergency tweeting system
Local timezone is Central Daylight Time(GMT-5), Windows clock shows at 10:38. I
suspect that the issue has to do with conversion of the timestamp as returned
by the Twitter API, but I don't know enough about it to be able to speak
authoritatively.
Original comment by eugene@kashpureff.org
on 15 Jun 2011 at 3:40
interesting.. Eugene - you're running Windows or GNU/Linux?
Original comment by andresen.nils@gmail.com
on 15 Jun 2011 at 4:24
Windows 7, Pidgin 2.7.11, mbpidgin 0.3.0, though this issue has persisted for
at least a year - I first noticed it at the previous DST switch-over.
I can spool up a Linux VM and test further if you like.
Original comment by eugene@kashpureff.org
on 15 Jun 2011 at 4:28
Twitter and the Pidgin history show the right time, mbpidgin not.
Twitter time zone is "Berlin +1" like the local time zone.
Windows 7, Pidgin 2.7.11, mbpidgin 0.3.0-svn384
Original comment by sumeru...@googlemail.com
on 15 Jun 2011 at 4:37
Attachments:
Sorry, I was away a few days..
@eugene: if you could that'd be nice.
I don't have that problem on the WinXP I am using for compiling
microblog-purple. I do however on my employers' Win7. Now I got to figure the
differences...
The Win7 is set to "UTC+1" and currently running "summer time". I need to check
the WinXP
Original comment by andresen.nils@gmail.com
on 23 Jun 2011 at 1:48
Ok, seems I am on to something.
Those of you who are willing to try take a look at
https://bitbucket.org/nils.a/timetest/
either compile yourself or try the bin's in the download-section (linux and
windows 64bit only - it'll be a day or two until I can get my hands on a 32bit
windows.
Using my test-app the result is now the same on windows and gnu/linux:
---- cut ----
=== ENVIRON ===
TZ = (null)
tzname[0] = CET
tzname[1] = CEST
daylight = 1
timezone = -3600
Local time : Sat, 24.05.111 16:12
utc : Sat, 24.05.111 14:12
=== END ENVIRON ===
final msg_time = 1308199980
fix test (16.6.2011 04:53) --> Thu Jun 16 06:53:00 2011
final msg_time = 1308199980
Thu Jun 16 04:53:00 +0000 2011 --> Thu Jun 16 06:53:00 2011
done.
---- cut ----
Test for yourself: the parsed "fix test" should be the same as as the other
test and both should show the same difference as local time to utc.
@somsaks: please have a look at
https://bitbucket.org/nils.a/timetest/src/tip/test.c and tell me what you think
of mb_timegm (result's the same as timegm but this one runs in windows, too)
Original comment by andresen.nils@gmail.com
on 24 Jun 2011 at 3:23
I think the code is ok as long as it yield the same result.
Original comment by somsaks
on 25 Jun 2011 at 3:44
I added a patch to fix the problem. Please try it and let me know the results.
Original comment by andresen.nils@gmail.com
on 26 Jun 2011 at 5:55
Attachments:
For those unwilling to through the trouble of building under windows I attached
the installer for testing - it is the current svn-build (commit#387) including
this patch. (built against pidgin 2.7.11)
Original comment by andresen.nils@gmail.com
on 26 Jun 2011 at 6:01
Attachments:
I have tested the prebuilt module, and it has solved this issue for me on
Windows 7 x64, Pidgin 2.7.11. I had to perform a complete uninstall of the
existing plugin before installing the provided binary, but I chalk that up to
Windows being Windows.
Original comment by eugene@kashpureff.org
on 26 Jun 2011 at 2:30
I am having a similar issue in Australia.
I believe that the timezone isn't being used correctly when in daylight savings
mode.
Versions:
pidgin 2.10.1
twintgin 0.3.0
Original comment by Christop...@gmail.com
on 22 Feb 2012 at 11:00
@Christop: Did you try the corrected version
pidgin-microblog-0.3.0-svn387-Issue189.exe from above? It solves this issue
pretty good (and has some other nice new features as well).
@andresen.nils@gmail.com: Have you considered putting the svn387 version as
0.3.1 online?
Original comment by sumeru...@googlemail.com
on 23 Feb 2012 at 8:40
Original issue reported on code.google.com by
mikeage
on 1 Jun 2010 at 12:38