On Fri, Dec 08, 2006 at 08:02:26PM +0100, Enrico Forestieri wrote:
> Please find attached a patch revised along your observations. However,
> now I feel quite uncomfortable with applying a patch which I don't
> fully understand...
I could not discover any strange side effect, so I committed it.
On Thu, Dec 07, 2006 at 05:13:08PM +0100, Jean-Marc Lasgouttes wrote:
> > "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>
> Enrico> I did that, even if I still cannot see any improvement on
> Enrico> *nix. Anyway, the memory penalty is less than about 200 Kbytes
> Enrico> for any
On Thu, Dec 07, 2006 at 09:42:24PM +0100, Enrico Forestieri wrote:
> On Thu, Dec 07, 2006 at 06:20:07PM +0100, Andre Poenitz wrote:
> > On Thu, Dec 07, 2006 at 04:03:27PM +0100, Enrico Forestieri wrote:
> > > + QRect const & r = metrics.boundingRect(c);
> > > + // Qt/Win 3.2.1nc (at least) corrects
On Fri, Dec 08, 2006 at 04:23:23PM +0100, Jean-Marc Lasgouttes wrote:
> > "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>
> Enrico> Jean-Marc, may I apply the attached patch? It is unrelated to
> Enrico> bug 2900.
>
> It looks like a good idea, but wouldn't it be even better to r
> "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
Enrico> Jean-Marc, may I apply the attached patch? It is unrelated to
Enrico> bug 2900.
It looks like a good idea, but wouldn't it be even better to replace
- updated |= cursorUp(cur);
+
+ if (cur.inMathed())
+
Enrico Forestieri wrote:
I did that, even if I still cannot see any improvement on *nix.
Anyway, the memory penalty is less than about 200 Kbytes for any
document I test it with.
I routinely run linux-LyX over and ADSL line. This mode of display
need every optimization it can get. :-)
Helge Ha
> "Andre" == Andre Poenitz <[EMAIL PROTECTED]> writes:
Andre> On Thu, Dec 07, 2006 at 04:03:27PM +0100, Enrico Forestieri
Andre> wrote:
>> + QRect const & r = metrics.boundingRect(c); + // Qt/Win 3.2.1nc
>> (at least) corrects the GetGlyphOutlineA|W y + // value by the
>> height: (x, -y-height
On Dec 7, 2006, at 10:03 AM, Enrico Forestieri wrote:
On Thu, Dec 07, 2006 at 03:18:44PM +0100, Jean-Marc Lasgouttes wrote:
"Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
I'd still like this patch to be done in a slightly cleaner way.
Currently, one can see that the special code
On Thu, Dec 07, 2006 at 06:20:07PM +0100, Andre Poenitz wrote:
> On Thu, Dec 07, 2006 at 04:03:27PM +0100, Enrico Forestieri wrote:
> > + QRect const & r = metrics.boundingRect(c);
> > + // Qt/Win 3.2.1nc (at least) corrects the GetGlyphOutlineA|W y
> > + // value by the height: (x, -y-height
On Thu, Dec 07, 2006 at 04:03:27PM +0100, Enrico Forestieri wrote:
> + QRect const & r = metrics.boundingRect(c);
> + // Qt/Win 3.2.1nc (at least) corrects the GetGlyphOutlineA|W y
> + // value by the height: (x, -y-height, width, height).
> + // Other versions return: (x, -y, width
On Thu, Dec 07, 2006 at 09:47:56AM +0100, Abdelrazak Younes wrote:
> Enrico Forestieri wrote:
...
> >BTW, I noticed that LyX crashes if you press page down right after
> >loading this document:
> >http://bugzilla.lyx.org/attachment.cgi?id=1296&action=view
>
> I cannot access this link and 1296 doe
> "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
Enrico> I did that, even if I still cannot see any improvement on
Enrico> *nix. Anyway, the memory penalty is less than about 200 Kbytes
Enrico> for any document I test it with.
Enrico> With this patch LyX/Win 1.4 is again on a par w
On Thu, Dec 07, 2006 at 03:18:44PM +0100, Jean-Marc Lasgouttes wrote:
> > "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>
> >> I'd still like this patch to be done in a slightly cleaner way.
> >> Currently, one can see that the special code for qt 3.2.1nc is
> >> copied in two dif
> "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>> I'd still like this patch to be done in a slightly cleaner way.
>> Currently, one can see that the special code for qt 3.2.1nc is
>> copied in two different file. Enrico, you should model the
>> ascent/descent caching code on the
On Thu, Dec 07, 2006 at 10:09:00AM +0100, Jean-Marc Lasgouttes wrote:
> > "Georg" == Georg Baum <[EMAIL PROTECTED]> writes:
>
> Georg> Am Mittwoch, 6. Dezember 2006 23:32 schrieb Enrico Forestieri:
> >> LyX 1.4 is now usable at least as 1.3.7 was on Windows ;-)
>
> Georg> If that is the case
On Thu, Dec 07, 2006 at 09:47:56AM +0100, Abdelrazak Younes wrote:
> Enrico Forestieri wrote:
...
> >Index: src/frontends/qt2/qfont_loader.h
> >===
> >--- src/frontends/qt2/qfont_loader.h (revision 16192)
> >+++ src/frontends/qt2/qfont
On Thu, Dec 07, 2006 at 09:47:56AM +0100, Abdelrazak Younes wrote:
> Enrico Forestieri wrote:
...
> >Please, find attached a patch that caches ascent and descent values
> >on Windows and Mac. I don't activate the cache on *nix because I
> >cannot see a noticeable improvement (I could have sworn I w
> "Georg" == Georg Baum <[EMAIL PROTECTED]> writes:
Georg> Am Mittwoch, 6. Dezember 2006 23:32 schrieb Enrico Forestieri:
>> LyX 1.4 is now usable at least as 1.3.7 was on Windows ;-)
Georg> If that is the case I suggest to immediately release 1.4.4. A
Georg> lot of other fixes have accumulat
On Thu, Dec 07, 2006 at 09:47:56AM +0100, Abdelrazak Younes wrote:
...
> >Of course, after I got my coffee it occurred to me that I should had
> >not tried to backport the patch but rather its philosopy...
>
> Right!
> Good job ;-)
+1
> >Please, find attached a patch that caches ascent and d
Am Mittwoch, 6. Dezember 2006 23:32 schrieb Enrico Forestieri:
> LyX 1.4 is now usable at least as 1.3.7 was on Windows ;-)
If that is the case I suggest to immediately release 1.4.4. A lot of other
fixes have accumulated, and IMHO this one alone would justify a release.
We can then still see w
Enrico Forestieri wrote:
On Wed, Dec 06, 2006 at 04:30:09PM +0100, Jean-Marc Lasgouttes wrote:
"Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
I would be very interested, but I will need a windows user to do
that for me, since I cannot really test. Anyway caching ascent and
descent l
On Wed, Dec 06, 2006 at 04:30:09PM +0100, Jean-Marc Lasgouttes wrote:
> > "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
> >> I would be very interested, but I will need a windows user to do
> >> that for me, since I cannot really test. Anyway caching ascent and
> >> descent looks l
> "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>> I would be very interested, but I will need a windows user to do
>> that for me, since I cannot really test. Anyway caching ascent and
>> descent looks like a good idea.
Enrico> Jean-Marc are you looking at this? I had a quick loo
On Fri, Dec 01, 2006 at 10:40:59PM +0100, Jean-Marc Lasgouttes wrote:
> > "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
>
> Abdelrazak> Good news for Windows and Mac users, this bug is now
> Abdelrazak> fixed: http://bugzilla.lyx.org/show_bug.cgi?id=2900
>
> Abdelrazak> Jean-M
Georg Baum wrote:
Am Freitag, 1. Dezember 2006 21:08 schrieb Andre Poenitz:
On Fri, Dec 01, 2006 at 05:26:43PM +0100, Georg Baum wrote:
Yes. Then you make it clear that the qt4 frontend can only handle ucs2.
Which is wrong as far as I have been told. Qt 4 is supposed to be able
to render the f
Georg Baum wrote:
Am Samstag, 2. Dezember 2006 00:29 schrieb Abdelrazak Younes:
None that I see. I will recompile with Debug to see if there's any
assertion triggered but I am quite confident that this was the only
problem. Because the assertion happpened at the same line when scrolling
>from
Am Samstag, 2. Dezember 2006 00:29 schrieb Abdelrazak Younes:
> None that I see. I will recompile with Debug to see if there's any
> assertion triggered but I am quite confident that this was the only
> problem. Because the assertion happpened at the same line when scrolling
> >from top or from
Am Freitag, 1. Dezember 2006 21:08 schrieb Andre Poenitz:
> On Fri, Dec 01, 2006 at 05:26:43PM +0100, Georg Baum wrote:
> > Yes. Then you make it clear that the qt4 frontend can only handle ucs2.
>
> Which is wrong as far as I have been told. Qt 4 is supposed to be able
> to render the full unicod
On Fri, Dec 01, 2006 at 05:26:43PM +0100, Georg Baum wrote:
> Abdelrazak Younes wrote:
>
> > By the way, around section 6.2 of the UserGuide there are some accent
> > characters which are bigger than 65536.
>
> Which ones? I suspect that something is going wrong there.
>
> > Shall I add an asser
Abdelrazak Younes wrote:
Anyway caching ascent and descent
looks like a good idea.
Abdelrazak> The only drawback is that it requires about 20Mo
Abdelrazak> extra-memory when loading the UserGuide. If it turns out
Abdelrazak> to be too much, we can switch to a QHash based solution
Abdelrazak> inst
Georg Baum wrote:
Am Freitag, 1. Dezember 2006 18:27 schrieb Abdelrazak Younes:
It is the hungarian umlaut apparently that gives this weird value.
Looking at what's on screen, this is no surprise really...
Now fixed.
Indeed. Thanks.
Do you see any other strange character?
None that I se
Jean-Marc Lasgouttes wrote:
"Bo" == Bo Peng <[EMAIL PROTECTED]> writes:
What about using short ints instead of ints?
Bo> I do not know what exactly is going on here, but if there are a
Bo> lot of duplicates, std::map should help.
A std::map is what we had, but a vector is faster.
Of course t
Jean-Marc Lasgouttes wrote:
"Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Good news for Windows and Mac users, this bug is now
Abdelrazak> fixed: http://bugzilla.lyx.org/show_bug.cgi?id=2900
Abdelrazak> Jean-Marc, I think you can backport to 1.4 the ascent()
Abdelra
> "Bo" == Bo Peng <[EMAIL PROTECTED]> writes:
>> What about using short ints instead of ints?
Bo> I do not know what exactly is going on here, but if there are a
Bo> lot of duplicates, std::map should help.
A std::map is what we had, but a vector is faster.
Of course the downside is that it c
What about using short ints instead of ints?
I do not know what exactly is going on here, but if there are a lot of
duplicates, std::map should help.
Bo
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Good news for Windows and Mac users, this bug is now
Abdelrazak> fixed: http://bugzilla.lyx.org/show_bug.cgi?id=2900
Abdelrazak> Jean-Marc, I think you can backport to 1.4 the ascent()
Abdelrazak> and descent() part o
Am Freitag, 1. Dezember 2006 18:27 schrieb Abdelrazak Younes:
> It is the hungarian umlaut apparently that gives this weird value.
> Looking at what's on screen, this is no surprise really...
Now fixed. Do you see any other strange character?
Georg
Abdelrazak Younes wrote:
Georg Baum wrote:
Abdelrazak Younes wrote:
By the way, around section 6.2 of the UserGuide there are some accent
characters which are bigger than 65536.
Which ones? I suspect that something is going wrong there.
It happens when displaying the line after the macron
Georg Baum wrote:
Abdelrazak Younes wrote:
By the way, around section 6.2 of the UserGuide there are some accent
characters which are bigger than 65536.
Which ones? I suspect that something is going wrong there.
It happens when displaying the line after the macron accent related
line. The
Abdelrazak Younes wrote:
> Good news for Windows and Mac users, this bug is now fixed:
Another good news is that I am now at 13 seconds for the UserGuide
scrolling test :-)
Congratulations and many thanks. Now you need to improve the
performance of multiple windows. I have been working on two d
Georg Baum wrote:
Abdelrazak Younes wrote:
Georg Baum wrote:
Abdelrazak Younes wrote:
By the way, around section 6.2 of the UserGuide there are some accent
characters which are bigger than 65536.
Which ones? I suspect that something is going wrong there.
I'll investigate.
Thanks.
Shal
Abdelrazak Younes wrote:
> Georg Baum wrote:
>> Abdelrazak Younes wrote:
>>
>>> By the way, around section 6.2 of the UserGuide there are some accent
>>> characters which are bigger than 65536.
>>
>> Which ones? I suspect that something is going wrong there.
>
> I'll investigate.
Thanks.
>>>
Abdelrazak Younes wrote:
Good news for Windows and Mac users, this bug is now fixed:
Another good news is that I am now at 13 seconds for the UserGuide
scrolling test :-)
Abdel.
Georg Baum wrote:
Abdelrazak Younes wrote:
By the way, around section 6.2 of the UserGuide there are some accent
characters which are bigger than 65536.
Which ones? I suspect that something is going wrong there.
I'll investigate.
Shall I add an assertion for
those cases?
Yes.
But th
Abdelrazak Younes wrote:
> By the way, around section 6.2 of the UserGuide there are some accent
> characters which are bigger than 65536.
Which ones? I suspect that something is going wrong there.
> Shall I add an assertion for
> those cases?
Yes. Then you make it clear that the qt4 frontend
Good news for Windows and Mac users, this bug is now fixed:
http://bugzilla.lyx.org/show_bug.cgi?id=2900
Jean-Marc, I think you can backport to 1.4 the ascent() and descent()
part of the patch. I am quite confident that this alone will solve most
speed problem in 1.4 also.
By the way, around
46 matches
Mail list logo
