[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: meinproc4 on Mac, KLocale (was Re: What to test for 4.13?)
From:       Ian Wadham <iandw.au () gmail ! com>
Date:       2014-03-19 1:58:03
Message-ID: A8DF687B-4EC7-4993-AFC6-B43B38FF55B6 () gmail ! com
[Download RAW message or body]

Hi Luigi, Marko and Thomas,

It's my birthday today, so it was a nice present to wake up here
in Australia and find that you guys have been busy while I have
been sleeping =85 :-)

Today is my busy day when I present a weekly course on science
for lay people, so I cannot join in on the meinproc4 hunt today, but
maybe tomorrow.

FWIW I am a KDE Games developer working on Apple OS X.  Marko
is an old friend I have met in the MacPorts environment.  I think I may
have met you too, Luigi, at Akademy 2008 in Belgium.  Do you have
a brother called Pino?

On 19/03/2014, at 9:26 AM, mk-lists@email.de wrote:
> On 18 Mar 2014, at 23:15 , Luigi Toscano <luigi.toscano@tiscali.it> wrote:
>> Ok: I've seen the other message, but then I would start from this and ma=
ke
>> sure that the stacktrace is the correct one.
> Hmm=85
> =

>> Could you (or any other Mac user/developer) please try to
>> - recompile kdelibs with debug symbols
> Right now I am not able to do so (and I knew you=92d ask for it=85), beca=
use I haven=92t set up a parallel install for debugging.
> That=92s something I long planned to do since I migrated slowly to my new=
 machine, but I haven=92t achieved it YET.
> =

>> - run meinproc4 on any document using the debugger, to get a complete ba=
cktrace
> Ha, I tell you, that will be impossible, I am afraid, because those crash=
es come unpredictably...
> :-(

This smells to me like a concurrency problem, e.g. a multi-threading error
or, more likely perhaps, a multi-processing error, where MacPorts is
running more than one meinproc4 at once during an installation run
and they somehow start and finish at such times as to get in each others'
way --- concurrent access to temp files or Apple OS X resources, that sort
of thing.

I say this from the point of view NOT of a KDE Games programmer, but
of a guy who cut his teeth on O/S and real-time programming nearly
50 years ago.  Back then it was conventional wisdom that intermittent
faults were always in the hardware and software problems were always
reproducible.  I found out the hard way that that is not always the case.

When MacPorts does a big install run, with dependencies, it can go for
hours and it uses 2-4 cores on my quad-core machine all the time.

OTOH I can use meinproc4 solo scores of times a day on Apple OS X, when
writing doco for the KDE Games Handbooks --- and it never fails for me.

Marko, if you run a long test and it does not fail, try running two tests a=
t once
from the same command-line window (i.e. add & [ampersand] to the end of
each command, maybe sending stdout and stderr to separate files).

Re KDE libraries with debug: that might or might not help.

It can change the timing and make the problem go away, if it is as
I suspect it is.  I even experience that when testing KGoldrunner =85 :-)
That's a game where lots of little men run around the screen
concurrently ...

Cheers, Ian W.


>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscrib=
e <<
[prev in list] [next in list] [prev in thread] [next in thread]