On Wed, 2006-02-15 at 07:52 +0100, Juergen Spitzmueller wrote:
...
> Obviously we are doing something illegal with the vector::iterator.
> It
> breaks at the marked operation:
>
> void QCitationDialog::up()
> {
> int const sel = selectedLB->currentItem();
>
> // Move the selected
Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
| > /usr/lib/gcc/i386-redhat-linux/3.4.4/../../../../include/c++/3.4.4/debug/sa
| >fe_iterator.h:127: error: attempt to copy-construct an iterator from a
| > singular iterator.
|
| Obviously we are doing something illegal with the vector::iterator.
Enrico Forestieri <[EMAIL PROTECTED]> writes:
> > > this is just to inform you that defining Q_CYGWIN_WIN in src/config.h
> > > is not sufficient to avoid X11 calls. I don't know why.
> >
> > That sounds as if some .C file doesn't #include
> Yes, that's it! All (or at least a large part) of the
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> Should we aim at replacing this stuff by an interface to
Martin> Enchant?
Martin> http://www.abisource.com/projects/enchant/
Sure, enchant looks very nice, but it is not going to replace
ControlSpellchecker, is it? I thought it
Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
> > /usr/lib/gcc/i386-redhat-linux/3.4.4/../../../../include/c++/3.4.4/debug/sa
> >fe_iterator.h:127: error: attempt to copy-construct an iterator from a
> > singular iterator.
>
> Obviously we are doing something illegal with the vector::iterator.
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> Is the iterator still valid after the erase operation?
I think it is not. Why not use
std::swap(form_->citekeys[sel -1], form_->citekeys[sel]);
JMarc
On Wed, 2006-02-15 at 10:10 +0100, Jean-Marc Lasgouttes wrote:
> > "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> Martin> Should we aim at replacing this stuff by an interface to
> Martin> Enchant?
>
> Martin> http://www.abisource.com/projects/enchant/
>
> Sure, enchant looks ver
Angus Leeming a écrit :
[...]
So, operator+ advances the iterator but operator- tells you how
far it would go but does nothing? That doesn't seem right.
Can you explain the rationale?
vector::iterator has these operators which does essentially that
(advance) and distance but list::iterator doe
Hello,
For those interested, here is a second version of the it_vector class.
With former one an assertion is triggered whenever you try to export to
latex. It seems that the at() method is used beyond the size() limit. I
am really not sure but there is maybe a hidden bug somewhere in the
"ex
Lots of good suggestions, thank you all!
Jean-Marc Lasgouttes wrote:
> Martin> Is the iterator still valid after the erase operation?
>
> I think it is not.
No, that's the problem indeed.
> Why not use
> std::swap(form_->citekeys[sel -1], form_->citekeys[sel]);
This one I like the most. How
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
> Angus Leeming a écrit :
> > So, operator+ advances the iterator but operator- tells you how
> > far it would go but does nothing? That doesn't seem right.
> > Can you explain the rationale?
> vector::iterator has these operators which does essential
Angus Leeming a écrit :
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
[...]
Having thought about it some more, I realise that I'm actually
perfectly happy with
it_vector::iterator it1 = ...;
int pos = ...;
it_vector::iterator it2 = it1 + pos;
However, I'm surprised that operator+ t
Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
| Lots of good suggestions, thank you all!
|
| Jean-Marc Lasgouttes wrote:
| > Martin> Is the iterator still valid after the erase operation?
| >
| > I think it is not.
|
| No, that's the problem indeed.
|
| > Why not use
| > std::swap(form_-
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| > Having said that, I don't suppose that I'd lose too much sleep if
| > the foreach declaration remained in it_vector.h but I'd be less happy
| > to see it remain as an it_vector member function.
|
| Let's wait if anybody else has an opinion on that
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
Juergen> This one I like the most. How about the attached patch? Fixes
Juergen> the crash in both frontends (John, I guess gtk needs
Juergen> something similar; btw with Bo's testcase the gtk citation
Juergen> dialog remains empty
Lars Gullik Bjønnes a écrit :
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| > Having said that, I don't suppose that I'd lose too much sleep if
| > the foreach declaration remained in it_vector.h but I'd be less happy
| > to see it remain as an it_vector member function.
|
| Let's wait if any
Jean-Marc Lasgouttes wrote:
> You should get rid of it completely and do
> string const tmp = form_->citekeys[sel];
> and maybe even replace the use of tmp by form_->citekeys[sel - 1]
>
> You even maybe add
> BOOST_ASSERT(sel > 0);
Sorry, I have just committed, but I can do those chang
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
Juergen> Jean-Marc Lasgouttes wrote:
>> You should get rid of it completely and do string const tmp
>> = form_->citekeys[sel]; and maybe even replace the use of tmp by
>> form_->citekeys[sel - 1]
>>
>> You even maybe add
(gdb) run
Starting program: /home/bartek/lyx-devel/src/lyx-gtk
/usr/lib/gcc/i686-pc-linux-gnu/3.4.5/include/g++-v3/debug/safe_iterator.h:127:
error: attempt to copy-construct an iterator from a singular iterator.
Objects involved in the operation:
iterator "this" @ 0x0xbfb6ddb0 {
type =
N11__g
Angus Leeming <[EMAIL PROTECTED]> writes:
> Well, given that these are generated files, it should be easy enough to fix
> them. Fancy putting together a patch?
The _moc.C files are generated from .h files that in turn are generated
form .ui files that I don't grok (is it XML?).
So, you will not
> "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
Enrico> Angus Leeming <[EMAIL PROTECTED]> writes:
>> Well, given that these are generated files, it should be easy
>> enough to fix them. Fancy putting together a patch?
Enrico> The _moc.C files are generated from .h files that in tu
Enrico Forestieri <[EMAIL PROTECTED]> writes:
>
> Angus Leeming ...> writes:
>
> > Well, given that these are generated files, it should be easy enough to fix
> > them. Fancy putting together a patch?
>
> The _moc.C files are generated from .h files that in turn are generated
> form .ui files
Hello,
Would it be possible to have a link to the diff and/or patch in the
messages sent to lyx-cvs?
Thanks in advance,
Abdel.
Bartek Kostrzewa wrote:
> (gdb) run
> Starting program: /home/bartek/lyx-devel/src/lyx-gtk
> /usr/lib/gcc/i686-pc-linux-gnu/3.4.5/include/g++-v3/debug/safe_iterator.h:1
>27: error: attempt to copy-construct an iterator from a singular iterator.
Do you have a recipe how to reproduce this crash?
Jü
Jean-Marc Lasgouttes <[EMAIL PROTECTED]> writes:
>
> > "Enrico" == Enrico Forestieri <[EMAIL PROTECTED]> writes:
>
> Enrico> Angus Leeming ...> writes:
> >> Well, given that these are generated files, it should be easy
> >> enough to fix them. Fancy putting together a patch?
>
> Enrico> Th
Jean-Marc Lasgouttes wrote:
> Juergen> Sorry, I have just committed, but I can do those changes, if
> Juergen> you like. Shall I prepare a second patch?
>
> I think it is better, because the code is a bit hybrid as it stands
> (iterators and direct access).
OK. I kept the tmp string for now, since
Abdelrazak Younes <[EMAIL PROTECTED]> writes:
| > IMHO let people use
| > std::for_each, at least you would have to change the api so that
| > functors are usable.
|
| True... It works for only one kind of functor (void func(T&)). I just
| wanted to hear opinion on that. The reason I have introdu
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
>> I thought it was not needed when modifying CPPFLAGS (or rather
>> CXXCPPFLAGS, I guess)...
Angus> Correct. But modifying these flags is a band aid that's needed
Angus> to fix broken code. All compiled files should #include
Angus> .
H
> "Juergen" == Juergen Spitzmueller <[EMAIL PROTECTED]> writes:
Juergen> OK. I kept the tmp string for now, since all this sel - 1 + 1
Juergen> - 2 confuses me ;-)
Juergen> Is this OK?
Yes.
JMarc
On Wed, 2006-02-15 at 14:22 +0100, Abdelrazak Younes wrote:
> Hello,
>
> Would it be possible to have a link to the diff and/or patch in the
> messages sent to lyx-cvs?
Lucky you. I wouldn't mind getting those messages in the first place.
- Martin
signature.asc
Description: This is a digital
> -Original Message-
> From: Martin Vermeer [mailto:[EMAIL PROTECTED]
> Sent: 15 February 2006 15:02
> To: YOUNES Abdelrazak (M3SYSTEM)
> Cc: lyx-devel@lists.lyx.org
> Subject: Re: SVN feature request
>
>
> On Wed, 2006-02-15 at 14:22 +0100, Abdelrazak Younes wrote:
> > Hello,
> >
> > W
Angus Leeming <[EMAIL PROTECTED]> writes:
>
> Enrico Forestieri ...> writes:
>
> >
> > Angus Leeming ...> writes:
> >
> > > Well, given that these are generated files, it should be easy enough to
> > > fix
> > > them. Fancy putting together a patch?
> >
> > The _moc.C files are generated fr
> "John" == John Spray <[EMAIL PROTECTED]> writes:
John> By the way, I wouldn't have been aware of this at all if I
John> hadn't been reading the list quite thoroughly lately. Perhaps it
John> would be a good idea to have some way of flagging any changes
John> which require work on the fronten
Martin Vermeer <[EMAIL PROTECTED]> writes:
| On Wed, 2006-02-15 at 14:22 +0100, Abdelrazak Younes wrote:
| > Hello,
| >
| > Would it be possible to have a link to the diff and/or patch in the
| > messages sent to lyx-cvs?
|
| Lucky you. I wouldn't mind getting those messages in the first place.
Enrico Forestieri <[EMAIL PROTECTED]> writes:
> > Make sense?
> Perfectly. But wait, are you saying that there is no way to achieve
> that by directly patching the .ui files?
I may be a "sed master" but I'm a "Qt hacker" ;-)
Having stated the disclaimer up front, I think that we are limited in wh
YOUNES Abdelrazak (M3SYSTEM <[EMAIL PROTECTED]> writes:
> > From: Martin Vermeer [mailto:[EMAIL PROTECTED]
> > Lucky you. I wouldn't mind getting those messages in the first place.
> news.gmane.org is your friend...
Especially if you grab the "sage" extenstion to firefox
(http://sage.mozdev.org/)
Juergen Spitzmueller wrote:
> Bartek Kostrzewa wrote:
>
>>(gdb) run
>>Starting program: /home/bartek/lyx-devel/src/lyx-gtk
>>/usr/lib/gcc/i686-pc-linux-gnu/3.4.5/include/g++-v3/debug/safe_iterator.h:1
>>27: error: attempt to copy-construct an iterator from a singular iterator.
>
>
> Do you have
On Wed, 2006-02-15 at 15:52 +0100, Lars Gullik Bjønnes wrote:
> Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> | On Wed, 2006-02-15 at 14:22 +0100, Abdelrazak Younes wrote:
> | > Hello,
> | >
> | > Would it be possible to have a link to the diff and/or patch in the
> | > messages sent to lyx-cvs
Martin Vermeer <[EMAIL PROTECTED]> writes:
| I am saying that they do not reach my mailbox martin.vermeer at hut.fi,
| where I asked them to be delivered.
And I said 'use procmail', and 'use fetchmail'. (or any other imap
client that can download)
--- file ~/.procmailrc ---
:0
! martin.vermeer a
Bartek Kostrzewa wrote:
> open new document -> insert -> graphics
>
> it does not matter wether the document is empty or not
I see, so it's probably a gtk specific problem. The gtk graphics dialog
crashes for me as well, amongst other, but I get another message:
glibmm-ERROR **: file objectbase.
I can't reproduce this, although I see you're using gcc 3.4.5, whereas
I'm using 4.0. This appears to be related to the signal handling in the
GtkLengthEntry class, although I can't see that it's doing anything
wrong.
What happens if you comment out lines 125 and 126 of GGraphics.C, thus
removin
On Wed, 2006-02-15 at 16:32 +0100, Juergen Spitzmueller wrote:
> glibmm-ERROR **: file objectbase.cc: line 77 (void
> Glib::ObjectBase::initialize(GObject*)): assertion failed: (gobject_ ==
> castitem)
> aborting...
> I'm not sure I have gtkmm and everything installed properly, though.
Problems
There are currently 74 bugs targeted at LyX 1.4.1. This is pure
madness.
Of these bugs, only 35 have a patch currently. I think I am going to
consider only those (plus the ones for which a patch magically
appears). And among these 35, I'll take the ones that are either very
low risk or really ne
Jean-Marc Lasgouttes wrote:
> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> madness.
What do you expect? ;-)
> Of these bugs, only 35 have a patch currently. I think I am going to
> consider only those (plus the ones for which a patch magically
> appears). And among these 35,
John C. Spray wrote:
> I can't reproduce this, although I see you're using gcc 3.4.5, whereas
> I'm using 4.0. This appears to be related to the signal handling in the
> GtkLengthEntry class, although I can't see that it's doing anything
> wrong.
>
> What happens if you comment out lines 125 and
Hello,
I have just discovered a potential bug with the keyboard selection:
Anywhere in the document, if you first press Shift+PageDown the effect
is the same as PageDown. But if you have already something selected
(with the keyboard or the mouse), Shift+PageDown will extend the
selection logic
Jean-Marc Lasgouttes <[EMAIL PROTECTED]> writes:
> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> madness.
I'm carefully using the plural "you" and "your" here rather than the more
friendly "we" and "our" because I'm completely underwhelmed by this 1.4.0/1.4.1
split.
Can someon
Jean-Marc Lasgouttes wrote:
> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> madness.
I get 84 open bugs for 1.4.1.
> The goal will be to try to get 1.4.1 out quickly, not to make it what
> 1.4.0 should have been in our dreams. In particular, I am tempted to
> keep the more in
On Wed, 2006-02-15 at 17:15 +0100, Jean-Marc Lasgouttes wrote:
> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> madness.
Could you post your URL for this?
> Of these bugs, only 35 have a patch currently. I think I am going to
> consider only those (plus the ones for which a pa
> "Angus" == Angus Leeming <[EMAIL PROTECTED]> writes:
Angus> Can someone explain to me what the purpose of the 1.4.0 release
Angus> is? You're proposing to ship what is known to be buggy
Angus> (crashing?) software so that you can then adopt the uniform of
Angus> the US Cavalry and ride to th
On 15/02/06, Martin Vermeer <[EMAIL PROTECTED]> wrote:
>
> On Wed, 2006-02-15 at 17:15 +0100, Jean-Marc Lasgouttes wrote:
> > There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> > madness.
>
> Could you post your URL for this?
>
> > Of these bugs, only 35 have a patch currently. I thi
> "Georg" == Georg Baum <[EMAIL PROTECTED]> writes:
Georg> Jean-Marc Lasgouttes wrote:
>> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
>> madness.
Georg> I get 84 open bugs for 1.4.1.
Yes, I forgot enhancements. I even get 85 :) And 40 with a patch.
JMarc
Martin Vermeer wrote:
> Crashes must be fixed.
Are we aware of any reproducible crashes?
Jürgen
Up arrow is working now. Thanks. (rev 13242).
Bo
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> On Wed, 2006-02-15 at 17:15 +0100, Jean-Marc Lasgouttes wrote:
>> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
>> madness.
Martin> Could you post your URL for this?
All bugs for 1.4.1:
http://bugzilla.lyx.org
Jean-Marc Lasgouttes wrote:
There are currently 74 bugs targeted at LyX 1.4.1. This is pure
madness.
Of these bugs, only 35 have a patch currently.
IMHO the real madness is that there are 35 known patches that won't go
into 1.4.0.
In particular, I am tempted to
keep the more involved spe
> "Michael" == Michael Gerz <[EMAIL PROTECTED]> writes:
Michael> Jean-Marc Lasgouttes wrote:
>> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
>> madness.
>>
>> Of these bugs, only 35 have a patch currently.
>>
Michael> IMHO the real madness is that there are 35 known patche
Am Mittwoch, 15. Februar 2006 21:04 schrieb Jean-Marc Lasgouttes:
> At some time the plan was to branch after 1.4.1.
I still think that this is a good idea.
Georg
On Wed, Feb 15, 2006 at 08:30:36PM +0100, Jean-Marc Lasgouttes wrote:
> > "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> Martin> On Wed, 2006-02-15 at 17:15 +0100, Jean-Marc Lasgouttes wrote:
> >> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> >> madness.
> Mar
Same documents 35 pages of math.
Two pre releases (pre3 and pre5).
Same arch: G4 MiniMac
Pre3 open the file without problem, scroll bottom, etc.
Pre5 CPU at %85, 3 minutes to scroll to the bottom. Everything is
sluggish. It is impossible to type.
I thought no big changes took place btw pre3 an
On Wed, Feb 15, 2006 at 05:15:20PM +0100, Jean-Marc Lasgouttes wrote:
>
> There are currently 74 bugs targeted at LyX 1.4.1. This is pure
> madness.
>
> Of these bugs, only 35 have a patch currently. I think I am going to
> consider only those (plus the ones for which a patch magically
> appears
I posted this bug some time ago on bugzilla:
http://bugzilla.lyx.org/show_bug.cgi?id=1955
Under Mac 10.4.2 and LyX 1.3.6, pasting from an external text editor
to LyX does not covert Mac style linebreaks to UNIX style
linebreaks. As a result, there are no linebreaks at all in anything
that
> "Matthew" == Matthew Cohen <[EMAIL PROTECTED]> writes:
Matthew> Under Mac 10.4.2 and LyX 1.3.6, pasting from an external text
Matthew> editor to LyX does not covert Mac style linebreaks to UNIX
Matthew> style linebreaks. As a result, there are no linebreaks at all
Matthew> in anything that i
Hello,
There are some DNS problems in our university, which results in reverse
lookup problems for the IP 141.225.11.87, and some of you might experience
problems with receiving emails from the lyx lists or seeing delayed
deliveries. We are working on the problem, and hopefully it will be
resolve
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> About bugs for 1.4.1 that have my name on it somewhere:
Martin> 580 (table messed up by some old LyX version): patch avail
Martin> Simple, risk-free
The nested ifs should only one, but the patch indeed looks good. Lars
concern
> "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
Martin> The conventional wisdom for this from the svn book is, that
Martin> you have an open trunk for this. There stuff gets tested, and
Martin> backported to branches/1.4 if found OK. And then released as
Martin> tags/1.4.1, tags/1.4.2
> "Michael" == Michael Gerz <[EMAIL PROTECTED]> writes:
Michael> Yes, we only end up with 50 patches that won't go into 1.4.0.
Michael> I am sorry to say so but the situation becomes more and more
Michael> absurd each day...
Something that would be nice before 1.4.0 is to update UPGRADING. An
> "Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Hello, I have just discovered a potential bug with the
Abdelrazak> keyboard selection: Anywhere in the document, if you first
Abdelrazak> press Shift+PageDown the effect is the same as PageDown.
Abdelrazak> But if you
On Tue, 14 Feb 2006, Enrico Forestieri wrote:
> > Now you can just use the file manager (username=lyx, password=LyXers) here
>
> Ach... I tried all possible combinations of lyx with lower and upper
> case letters but didn't think of LyXers :(
It's such a secure password... ;-)
> I uploaded them
On Wed, 15 Feb 2006, Angus Leeming wrote:
> YOUNES Abdelrazak (M3SYSTEM <[EMAIL PROTECTED]> writes:
> > > From: Martin Vermeer [mailto:[EMAIL PROTECTED]
> > > Lucky you. I wouldn't mind getting those messages in the first place.
> > news.gmane.org is your friend...
>
> Especially if you grab the
On Wed, 15 Feb 2006, Jean-Marc Lasgouttes wrote:
> > "Michael" == Michael Gerz <[EMAIL PROTECTED]> writes:
>
> Michael> Yes, we only end up with 50 patches that won't go into 1.4.0.
> Michael> I am sorry to say so but the situation becomes more and more
> Michael> absurd each day...
>
> Some
Personally I think it's madness to release something with
known critical bugs that have patches.
I think proposed patches should be in CVS. That way, when you update
you get the current patches & can apply the ones you want. It
would encourage more testing through everyday use, while allowing
b
Dear list,
I personally would like to see a better 1.4.0, rather than a quick
1.4.1 after it, especially when there are already many (maybe
untested) patches. I am using 1.4.0cvs everyday. While it has been
certainly usable, some edges need to be smoothed. For example,
1. 1.4.0cvs is significantl
On Wed, Feb 15, 2006 at 11:47:54PM +0100, Jean-Marc Lasgouttes wrote:
> > "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
> Martin> 2218
>
> Is the fact that the word at cursor is skipped due to your patch?
Don't know...didn't get that far yet.
The problem of skipping the first miss
On Wed, Feb 15, 2006 at 11:47:54PM +0100, Jean-Marc Lasgouttes wrote:
> > "Martin" == Martin Vermeer <[EMAIL PROTECTED]> writes:
>
> Martin> About bugs for 1.4.1 that have my name on it somewhere:
>
> Martin> 580 (table messed up by some old LyX version): patch avail
> Martin> Simple, risk-fr
Grag wrote:
> Pre5 CPU at %85, 3 minutes to scroll to the bottom. Everything is
> sluggish. It is impossible to type.
>
> I thought no big changes took place btw pre3 and pre4.
Can you provide a testcase?
Jürgen
Beck, Andrew Thomas - BECAT001 wrote:
> Personally I think it's madness to release something with
> known critical bugs that have patches.
which bugs are really "critical"?
Jürgen
Jean-Marc Lasgouttes a écrit :
"Abdelrazak" == Abdelrazak Younes <[EMAIL PROTECTED]> writes:
Abdelrazak> Hello, I have just discovered a potential bug with the
Abdelrazak> keyboard selection: Anywhere in the document, if you first
Abdelrazak> press Shift+PageDown the effect is the same as PageD
78 matches
Mail list logo
