On Tue, Mar 28, 2006 at 08:30:29PM +0200, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
> >Attached a slightly improved patch that will go into trunk unless
> >objections come up.
> I have a comment with regard to your use of ParagraphList. Sorry I just
> noticed that now...
> In your new fu
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Before the ParagraphList rewrite, the slowdown was shared
Abdelrazak> between this and paragraph insertion/deletion. If we
Abdelrazak> manage to speed this up I am sure that the user experience
Abdelrazak> under window
Jean-Marc Lasgouttes a écrit :
"Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> UserGuide.lyx and Extended.lyx. Remember, this is on
Abdelrazak> windows and the mingw compiler is known to generate much
Abdelrazak> slower code than linux.
Sure, but I am sur
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> UserGuide.lyx and Extended.lyx. Remember, this is on
Abdelrazak> windows and the mingw compiler is known to generate much
Abdelrazak> slower code than linux.
Sure, but I am surprised that this particular code stands o
Jean-Marc Lasgouttes a écrit :
"Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> It seems to work the same. The speed gain will come with
Abdelrazak> use of splice. And we definitely must optimize
Abdelrazak> "updateCounter".
What document shows that updatecounters is a
Martin Vermeer a écrit :
Pfiouu... I just managed to compile the qt2 frontend... The up and down
operations are absolutely not fast under windows Martin. We absolutely
must disable the outline buttons when an outline operation is being
done. I managed to crash LyX by multiple clicking on the Up
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> It seems to work the same. The speed gain will come with
Abdelrazak> use of splice. And we definitely must optimize
Abdelrazak> "updateCounter".
What document shows that updatecounters is a problem? I fail to see a
sl
On Wed, Mar 29, 2006 at 04:29:49PM +0200, Abdelrazak Younes wrote:
> Jean-Marc Lasgouttes a écrit :
> >It is great, and I want it for 1.4.something.
> >
> >Would it be possible to remember the scrollbar position and reset it
> >afterwards (if entry is still visible)? Currently the resetting of the
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Just for information, isn't the policy to have bug-fixes
Abdelrazak> only for 1.4.x?
The policy is:
- no format changes
- only fixes that are correctly understood and simple (not the case
for 1.4.1)
- not too muc
Jean-Marc Lasgouttes a écrit :
It is great, and I want it for 1.4.something.
Would it be possible to remember the scrollbar position and reset it
afterwards (if entry is still visible)? Currently the resetting of the
scroll bar is annoying.
Pfiouu... I just managed to compile the qt2 fronte
Jean-Marc Lasgouttes a écrit :
"Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> Attached a new patch which does it this way. This patch also
Martin> positions the current header visibly in the window. In other
Martin> words, it does pretty much everything outlining is
On Wed, Mar 29, 2006 at 03:30:41PM +0200, Jean-Marc Lasgouttes wrote:
> > "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> Martin> On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
> >> Martin Vermeer a écrit :
...
> Martin> Attached a new patch which does it this way. T
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
>> Martin Vermeer a écrit :
Martin> ...
>> We should not define slot connections inside the ui file. If
>> possible, please do that in the code, this is simple enough:
On Tue, 2006-03-28 at 20:50 +0300, Martin Vermeer wrote:
> On Mon, Mar 27, 2006 at 06:17:11PM +0200, Abdelrazak Younes wrote:
> > Martin Vermeer a écrit :
> > >On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
> > >>Martin Vermeer a écrit :
>
> ...
...
> Give it a try.
Committed to t
Martin Vermeer a écrit :
Not very pretty... I see why you had your helper methods.
I suppose I'll just commit as is (it's not speed critical), perhaps with
FIXME's. This can always be modified later.
The splice method proposed in the patch I just sent would be very
efficient for the "up" an
Martin Vermeer a écrit :
Ah! No iterators at all.
None ;-)
s = pit;
pit_type t = s + len;
// and also this one:
pars.erase(s, t);
And there...
pars.erase(boost::next(pars.begin(), s),
boost::next(pars.b
On Tue, Mar 28, 2006 at 09:13:41PM +0200, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
> >On Tue, Mar 28, 2006 at 08:30:29PM +0200, Abdelrazak Younes wrote:
> >
> >>Martin Vermeer a écrit :
...
> >How? I don't quite see it.
> >
> Here is a modified version of your code (not tested and
Abdelrazak Younes a écrit :
// Lars has deleted all my convenience methods in
RandomAccessList
// especially one that offered this one:
pars.insert(q, s, p);
...
pars.erase(s, t);
Actually, we just need a swap function in RandomAccessList.
Martin Vermeer a écrit :
On Tue, Mar 28, 2006 at 08:30:29PM +0200, Abdelrazak Younes wrote:
Martin Vermeer a écrit :
Attached a slightly improved patch that will go into trunk unless
objections come up.
I have a comment with regard to your use of ParagraphList. Sorry I just
noti
On Tue, Mar 28, 2006 at 08:30:29PM +0200, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
> >Attached a slightly improved patch that will go into trunk unless
> >objections come up.
> I have a comment with regard to your use of ParagraphList. Sorry I just
> noticed that now...
> In your new fu
Martin Vermeer a écrit :
Attached a slightly improved patch that will go into trunk unless
objections come up.
I have a comment with regard to your use of ParagraphList. Sorry I just
noticed that now...
In your new function.
+void Outline(OutlineOp mode, Buffer * buf, pit_type & pit)
you use
On Mon, Mar 27, 2006 at 06:17:11PM +0200, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
> >On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
> >>Martin Vermeer a écrit :
...
> >This patch also positions
> >the current header visibly in the window. In other words, it does pretty
>
Martin Vermeer a écrit :
On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
Martin Vermeer a écrit :
...
We should not define slot connections inside the ui file. If possible,
please do that in the code, this is simple enough:
connect(moveinPB, SIGNAL(clicked()), this, SLOT(movein
On Mon, 2006-03-27 at 11:16 +0200, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
...
> We should not define slot connections inside the ui file. If possible,
> please do that in the code, this is simple enough:
>
> connect(moveinPB, SIGNAL(clicked()), this, SLOT(movein_adaptor()));
>
> A
Martin Vermeer a écrit :
On Sat, 2006-03-25 at 18:39 +0200, Martin Vermeer wrote:
See attached.
The Deeper and Shallower ("In" and "Out") functionalities are not yet
implemented.
You can test the moving of a sectioning header plus belonging text up
and down by entering into the minibuffer: "ou
On Sat, 2006-03-25 at 18:39 +0200, Martin Vermeer wrote:
> See attached.
>
> The Deeper and Shallower ("In" and "Out") functionalities are not yet
> implemented.
>
> You can test the moving of a sectioning header plus belonging text up
> and down by entering into the minibuffer: "outline 0" or "o
Martin Vermeer a écrit :
On Sat, Mar 25, 2006 at 06:44:42PM +0100, Abdelrazak Younes wrote:
Martin Vermeer a écrit :
See attached.
The Deeper and Shallower ("In" and "Out") functionalities are not yet
implemented.
You can test the moving of a sectioning header plus belonging text up
and down
On Sat, Mar 25, 2006 at 06:44:42PM +0100, Abdelrazak Younes wrote:
> Martin Vermeer a écrit :
> >See attached.
> >
> >The Deeper and Shallower ("In" and "Out") functionalities are not yet
> >implemented.
> >
> >You can test the moving of a sectioning header plus belonging text up
> >and down by ent
Martin Vermeer a écrit :
See attached.
The Deeper and Shallower ("In" and "Out") functionalities are not yet
implemented.
You can test the moving of a sectioning header plus belonging text up
and down by entering into the minibuffer: "outline 0" or "outline 1".
I have this stuff also working f
Martin Vermeer wrote:
> See attached.
I'm impressed how you workload scales.
Jürgen
30 matches
Mail list logo
