mailing-list for TeXmacs Users

Text archives Help


Re: [TeXmacs] TeXmacs 1.0.7.8 refresh


Chronological Thread 
  • From: Gubinelli Massimiliano <address@hidden>
  • To: address@hidden
  • Subject: Re: [TeXmacs] TeXmacs 1.0.7.8 refresh
  • Date: Thu, 25 Nov 2010 11:54:36 +0100
  • Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=message-id:from:to:in-reply-to:content-type :content-transfer-encoding:mime-version:subject:date:references :x-mailer; b=Z/W51SNbLFnie1vT0xYjQorwOnSsaCHnW8UqrRXJLxE/W0KiqVBHhQklxaqlkp07jg 7Yn7tV8aSmTvqc+5dyJJv/bKjN/57fUHittPz95MSXyLLLHJo8Ax+npCgsXJopINqRTu mDIHbdvRclRu33sUGT7mUl0mIVbfxcPBJ6xfo=

Thanks for all these reports.
As pointed out by Sam, this bug is mostly probably linked with texmacs behavior under repainting. The current Qt solution (which is different from the X11 one due to limitations in the API) is to stop typesetting and repainting after a fixed amount of time and retry later to finish the work. This to allow user input and to avoid aving the GUI blocked while typesetting and repainting. Unfortunately I cannot reproduce the bug on my Mac and this make less easy for me to provide a reliable fix. I hope to be able to fix this issue soon.

best
max



On 25 nov. 10, at 11:46, Sam Liddicott wrote:


On 24/11/10 15:26, Alvaro Tejero Cantero wrote:
TeXmacs 1.0.7.8 on Ubuntu 10.04 64 bit compiled with QT support enters
a refresh loop in my system whereby about 3 lines are constantly
visible, another 1-3 keep appearing-disappearing and the rest is
blank.

This also happens when visiting the documentation, so anybody can
check (it is not document-dependent).


I noticed today that a paint loop leaves the CPU load very high.

Observation also leads me to say that I think the processing of a paint message will quit after certain time has been taken and paint the rest later.

If painting takes too long, then texmacs will get into a loop (perhaps qt doesn't easily report that a smaller region needs painting, or perhaps texmacs can't take advantage of this).

I had a case today where the problem did not go away by switching focus or covering the window, and even other texmacs windows which had painted fine (and hadn't been scrolled) couldn't paint.

The solution was to minimize them till the load went low, and then when I restored the stuck window it could paint.

I wonder if the paint used up CPU share to the degree that it could never get enough CPU to finish painting and so got stuck in a loop of always trying to paint.

Just a guess.

Sam

-á.


--
[FSF Associate Member #2325] <http://www.fsf.org/register_form?referrer=2325 >

<http://www.openrightsgroup.org/>




Archive powered by MHonArc 2.6.19.

Top of page