Braunstein Alfredo wrote:
>> Either ideas seem reasonable to me.
>>
>> Would it work for 1.3.x too?
>
> No idea (but it would seem so), I'll check tonight.
It doesn't. Still two redraws are invoked per page-down. Honestly don't know
why (and the old update scheme is a bit harder to investigate)
Ok to apply?
(makes qt behave like xforms and do not fire an update on setting the
scrollbar value)
Alfredo
Index: QContentPane.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/frontends/qt2/QContentPane.C,v
retrieving revision 1
Hello everybody,
a new release of our favourite debugging tool is available:
http://valgrind.kde.org/downloads.html
Michael
On Fri, Nov 21, 2003 at 06:02:31PM +0100, Michael Schmitt wrote:
> Hello,
>
> this patch removes the bv parameter from the inset methods
> open/close/status + removes a redundant update() call.
>
> Concerning the markDirty() calls, I noticed that there is a general
> mechanisms which is based o
Not nice:
vspace.C: In member function `const std::string
VSpace::asLatexCommand(const
BufferParams&) const':
vspace.C:478: Warnung: control reaches end of non-void function
vspace.C: In member function `int VSpace::inPixels(const BufferView&)
const':
vspace.C:514: Warnung: control reaches en
Hello,
this patch removes the bv parameter from the inset methods
open/close/status + removes a redundant update() call.
Concerning the markDirty() calls, I noticed that there is a general
mechanisms which is based on the action type. I.e. we do not check
whether a change has actually taken pl
small stuff
functionally equivalent
? 1.diff
? 1.diff.gz
? 2.diff
? 3.diff
? ?t
? fullredraw.diff
? par-row.diff
? tabular-il.diff
? textcache.diff
? insets/1.diff
? mathed/cursor.diff
? support/1.diff
Index: rowpainter.C
===
RCS file
The removed inset functions return the same value as the base functions
The rest are simple transformations of the code. No visible difference.
--
Those who desire to give up Freedom in order to gain Security, will not have,
nor do they deserve, either one. (T. Jefferson or B. Franklin or b
More simplification.
This waste a few cycles but only shows up if you have a text inset
covering a couple of dozen pages. Which is an unlikely event.
Andre'
--
Those who desire to give up Freedom in order to gain Security, will not have,
nor do they deserve, either one. (T. Jefferson or
small simplification
--
Those who desire to give up Freedom in order to gain Security, will not have,
nor do they deserve, either one. (T. Jefferson or B. Franklin or both...)
? .text.C.swp
? 1.diff
? 1.diff.gz
? 2.diff
? 3.diff
? ?t
? fullredraw.diff
? par-row.diff
? tabular-il.diff
? textc
On Friday 21 November 2003 3:46 pm, Andre Poenitz wrote:
> On Fri, Nov 21, 2003 at 03:41:49PM +, Angus Leeming wrote:
> > I'll have a play next week. Off to Snowdonia for a weekend of survival
> > training.
>
> I think you don't need to go. The situation in London is likely to
> normalize soon.
On Fri, Nov 21, 2003 at 03:41:49PM +, Angus Leeming wrote:
> I'll have a play next week. Off to Snowdonia for a weekend of survival
> training.
I think you don't need to go. The situation in London is likely to
normalize soon...
Andre'
On Friday 21 November 2003 3:34 pm, John Levon wrote:
> On Fri, Nov 21, 2003 at 08:36:53AM +, Angus Leeming wrote:
> > > You missed out *why* the core needs such fake move events. I
> > > understand the answer, but I do not know the question. "42" isn't
> > > good enoough for me :)
> >
> > The
On Fri, Nov 21, 2003 at 08:36:53AM +, Angus Leeming wrote:
> > You missed out *why* the core needs such fake move events. I
> > understand the answer, but I do not know the question. "42" isn't
> > good enoough for me :)
>
> The primary example is "select some text with the mouse and, with th
On Fri, Nov 21, 2003 at 07:44:19AM +0100, Alfredo Braunstein wrote:
> This seems to be the one:
>
> 2002-09-11 John Levon <[EMAIL PROTECTED]>
>
> * qscreen.C: use repaint() not update() for immediate change
>
> is it?
That looks like the one I'm talking about.
> AFAICT is not related
On Fri, Nov 21, 2003 at 03:56:16PM +0100, Andre' Poenitz wrote:
>
> still sluggish...
Wrong patch...
Andre'
Index: rowpainter.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/rowpainter.C,v
retrieving revision 1.91
diff -u -p -r
still sluggish...
Andre'
--
Those who desire to give up Freedom in order to gain Security, will not have,
nor do they deserve, either one. (T. Jefferson or B. Franklin or both...)
? .paragraph_funcs.h.swp
? 1.diff
? 1.diff.gz
? 2.diff
? 3.diff
? ?t
? fullredraw.diff
? par-row.diff
? tabular
Andre Poenitz wrote:
However, I think we can completely remove the parameters as they are
never used (in all functions with the same name).
Well spotted.
As it stands the cursor will get a BufferView * argument on construction
and keep this all the time as member. So the patch should be safe.
So
On Fri, Nov 21, 2003 at 02:57:28PM +0100, Lars Gullik Bjønnes wrote:
> Andre Poenitz <[EMAIL PROTECTED]> writes:
>
> | This moves us a step forward to using the same code for inset and main
> | text drawing.
> >
> | Towards the end of the Userguide things get a bit sluggish, though ;-)
>
> And yo
On Fri, Nov 21, 2003 at 02:45:35PM +0100, Michael Schmitt wrote:
> Hi Angus,
>
> you fixed a couple of compiler warnings ("unused parameters") by simply
> removing the names of corresponding function parameters.
>
> However, I think we can completely remove the parameters as they are
> never us
Andre Poenitz <[EMAIL PROTECTED]> writes:
| This moves us a step forward to using the same code for inset and main
| text drawing.
>
| Towards the end of the Userguide things get a bit sluggish, though ;-)
And you have of course the plan ready to get rid of the sluggishness?
--
Lgb
This moves us a step forward to using the same code for inset and main
text drawing.
Towards the end of the Userguide things get a bit sluggish, though ;-)
Andre'
--
Those who desire to give up Freedom in order to gain Security, will not have,
nor do they deserve, either one. (T. Jefferson
Hi Angus,
you fixed a couple of compiler warnings ("unused parameters") by simply
removing the names of corresponding function parameters.
However, I think we can completely remove the parameters as they are
never used (in all functions with the same name).
I prepared a small patch for "getCur
Angus Leeming wrote:
> Index: BufferView_pimpl.C
> Index: lyxfunc.C
> Index: lyxtext.h
> Revert the addition of spurious whitespace.
Sorry will do.
> Index: lyxtext.h
> + /// absolute document pixel coordinates of this LyXText
> + int x0_;
> + int y0_;
> I thought these were s
On Fri, Nov 21, 2003 at 11:54:28AM +0100, Alfredo Braunstein wrote:
> Alfredo Braunstein wrote:
>
> > dummy patch to allow extending a word/row selection with the mouse.
> >
> > Ok?
>
> Completely wrong patch ;-). Try this one instead.
Better ;-)
Andre'
Alfredo Braunstein wrote:
> dummy patch to allow extending a word/row selection with the mouse.
>
> Ok?
Completely wrong patch ;-). Try this one instead.
Alfredo
Index: text3.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/tex
Alfredo Braunstein wrote:
> dummy patch to allow extending a word/row selection with the mouse.
>
> Ok?
>
Index: BufferView_pimpl.C
Index: lyxfunc.C
Index: lyxtext.h
Revert the addition of spurious whitespace.
Index: lyxtext.h
+ /// absolute document pixel coordinates of this LyXText
+
On Fri, Nov 21, 2003 at 11:51:06AM +0100, Alfredo Braunstein wrote:
> dummy patch to allow extending a word/row selection with the mouse.
>
> Ok?
THis is yesterdays' patch, isn't it?
Andre'
dummy patch to allow extending a word/row selection with the mouse.
Ok?
Alfredo
Index: BufferView_pimpl.C
===
RCS file: /usr/local/lyx/cvsroot/lyx-devel/src/BufferView_pimpl.C,v
retrieving revision 1.465
diff -u -p -u -r1.465 BufferV
On Fri, Nov 21, 2003 at 10:39:26AM +0100, Lars Gullik Bjønnes spake thusly:
> Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> | (Surely paragraph numbering could be of use on the LaTeX side too? Are
> | there any documentclasses with numbered paragraphs?)
>
> All... and that is controlled by the
Lars Gullik Bjønnes wrote:
> Right.
>
> With debug mode turned on it uses safe iterators (checked
> iterators), that check if iterators compared belong to the same
> container etc. checks some bounds for containers, check that
> algorithm requirements are fullfilled etc.
>
> So if we are able to
Martin Vermeer <[EMAIL PROTECTED]> writes:
| (Surely paragraph numbering could be of use on the LaTeX side too? Are
| there any documentclasses with numbered paragraphs?)
All... and that is controlled by the "numbering depth"
>
| Lars, can this go in
>
| [ ] as such
| [ ] with or
| [ ] withou
Andre Poenitz <[EMAIL PROTECTED]> writes:
| On Fri, Nov 21, 2003 at 08:51:36AM +, Angus Leeming wrote:
>> In turn, buffer.h has this:
>> #include "ParagraphList_fwd.h"
>> class Buffer {
>> public:
>> class inset_iterator {
>> private:
>> ParagraphList::iterator
On Fri, Nov 21, 2003 at 08:51:36AM +, Angus Leeming wrote:
> In turn, buffer.h has this:
> #include "ParagraphList_fwd.h"
> class Buffer {
> public:
> class inset_iterator {
> private:
> ParagraphList::iterator pit;
> ParagraphList::iterator pend;
Lars Gullik Bjønnes wrote:
> PreviewLoader.C:594: instantiated from here
>
/opt/gcc34/lib/gcc/i686-pc-linux-gnu/3.4/../../../../include/c++/3.4/bits/stl_list.h:84:
> error: `__gnu_norm::_List_node<_Tp>::_M_data' has incomplete type
> ../../src/ParagraphList_fwd.h:18: error: forward declaration o
> "Felix" == Felix Kurth <[EMAIL PROTECTED]> writes:
Felix> did you apply it ? Or there are any further thinks to do for me
Felix> ?
I did not apply it yet, because I also need an entry for the
LaTeXConfig.lyx.in file describing what it is good for, where to find
it and (this is the part I ca
Angus Leeming wrote:
> The primary example is "select some text with the mouse and, with the
> mouse button still pressed, move the cursor beneath the LyX screen".
>
> The xforms frontend keeps scrolling (and selecting). No further
> movement of the cursor is required. (This is something copied f
John Levon wrote:
> On Thu, Nov 20, 2003 at 05:28:38PM +, Angus Leeming wrote:
>
>> John, you know the Qt stuff better than anyone else. Does my
>> proposal seem reasonable to you? To recap:
>>
>> xforms generates 'pseudo' XEvents in its main interation loop. See
>> below for a refresher ;-)
On Fri, Nov 21, 2003 at 09:20:34AM +0100, Alfredo Braunstein wrote:
> Alfredo Braunstein wrote:
>
> > I'll do the change.
>
> Ok to commit?
Looks good, so: Sure.
> + The above two fixes
> + x0_ -> xo_
> + some cursorPar()->getRow(cursor.pos()) -> cursorRow
Good idea ;-)
Andre'
Here's the evolved version. I added the feature of ID labels (counter
based) including paragraph ID labels, which are REQUIRED for AGU.
(Surely paragraph numbering could be of use on the LaTeX side too? Are
there any documentclasses with numbered paragraphs?)
...and general polish and improved co
Alfredo Braunstein wrote:
> I'll do the change.
Ok to commit?
+ The above two fixes
+ x0_ -> xo_
+ some cursorPar()->getRow(cursor.pos()) -> cursorRow
Alfredo
? bfri.C
Index: lyxtext.h
===
RCS file: /usr/local/lyx/cvsroot/lyx-de
Andre Poenitz wrote:
> We simply should draw full paragraphs. This way we could have a smooth
> drawing of 'really tall rows' and this would make the usage of
> RowPainter a bit simpler, too.
Fine.
> InsetText::draw
> LyXText::draw(to be implemented)
> Paragraph::draw (to be implem
On Fri, Nov 21, 2003 at 08:19:07AM +0100, Alfredo Braunstein wrote:
> First a fix to checkInsetHit. (some insets on screen cannot be
> reached by clicks).
> Fixing this remarks another problem: we draw at the row level
> (i.e. stop drawing when the row goes out of screen),
> but use the position
Andre Poenitz wrote:
> Isn't any comment related to 'update' slightly out of date?
Sorry, my bad. This in particular is from a file that somehow belongs to
other tree... (a discontinued class from trolltech distributed separately
that we have as-is inside qt/ or something).
Alfredo
On Thu, Nov 20, 2003 at 05:38:10PM +0100, Alfredo Braunstein wrote:
> I've grepped a little qt/*.C and found the following in qttableview.C:
>
> /*!
> Repaints the table view directly by calling paintEvent() directly
> unless updates are disabled.
>
> Erases the view area \a (x,y,w,h) if \a
First a fix to checkInsetHit. (some insets on screen cannot be
reached by clicks).
Fixing this remarks another problem: we draw at the row level
(i.e. stop drawing when the row goes out of screen),
but use the position information for all insets in the paragraph.
Thus the last rows in the last
46 matches
Mail list logo
