Re: reliability of par->getInset()

2001-11-29 Thread John Levon
On Thu, Nov 29, 2001 at 02:49:18PM +0100, Juergen Vigna wrote: > BTW.: The bug was quite obvious and I found it after looking 10 minutes > into the source ;) and using a debugger to see where I crash (I use > ddd as frontend to gdb a really usefull tool!) sure, I can see the crash. B

Re: reliability of par->getInset()

2001-11-29 Thread Juergen Vigna
On 28-Nov-2001 John Levon wrote: > On Fri, Nov 23, 2001 at 10:25:20AM +0100, Juergen Vigna wrote: > >> I'm very nearly sure that the crashes are connected to access a LyXText >> which already freed and reallocated row stuff. There are some pointers >> in LyXText which hold pointerss to rows for

Re: reliability of par->getInset()

2001-11-28 Thread John Levon
On Fri, Nov 23, 2001 at 10:25:20AM +0100, Juergen Vigna wrote: > I'm very nearly sure that the crashes are connected to access a LyXText > which already freed and reallocated row stuff. There are some pointers > in LyXText which hold pointerss to rows for update stuff. This pointers > are not val

Re: reliability of par->getInset()

2001-11-23 Thread Juergen Vigna
On 23-Nov-2001 John Levon wrote: >> Well while you're inside another inset (the_locking_inset != 0) the outside >> cursor position does NOT changed! So cpos(bv) is always before the inset > > eep, so we have several different cursors like this, as well as the selection > cursor etc. ? Every Ly

Re: reliability of par->getInset()

2001-11-23 Thread John Levon
On Fri, Nov 23, 2001 at 10:34:57AM +0100, Juergen Vigna wrote: > > I don't see how the insettext function lockInsetinInset() can guarantee > > that cpos(bv) is at an inset ?? Is the correct fix to add an isInset check ? > > Well while you're inside another inset (the_locking_inset != 0) the outs

Re: reliability of par->getInset()

2001-11-23 Thread Juergen Vigna
On 23-Nov-2001 John Levon wrote: > I don't see how the insettext function lockInsetinInset() can guarantee > that cpos(bv) is at an inset ?? Is the correct fix to add an isInset check ? Well while you're inside another inset (the_locking_inset != 0) the outside cursor position does NOT changed!

Re: reliability of par->getInset()

2001-11-23 Thread Juergen Vigna
On 22-Nov-2001 John Levon wrote: > Unfortunately I've not yet been successful in dragging some info out of people for > the bugs I've looked at (breaking line with footnote etc., and crash when deleting > row in tabular) ... I'm very nearly sure that the crashes are connected to access a LyXTex

Re: reliability of par->getInset()

2001-11-23 Thread John Levon
On Fri, Nov 23, 2001 at 04:47:05PM +1000, Allan Rae wrote: > On Fri, 23 Nov 2001, Allan Rae wrote: > > > On Thu, 22 Nov 2001, John Levon wrote: > > > > > is getInset still failing to find insets after a par->isInset returns true ? > > > > Yes it would seem so. > > > > Open a new document, insert

Re: reliability of par->getInset()

2001-11-22 Thread Allan Rae
On Fri, 23 Nov 2001, Allan Rae wrote: > On Thu, 22 Nov 2001, John Levon wrote: > > > is getInset still failing to find insets after a par->isInset returns true ? > > Yes it would seem so. > > Open a new document, insert a figure float and put something inside it > (optional I think). Just so lon

Re: reliability of par->getInset()

2001-11-22 Thread Allan Rae
On Thu, 22 Nov 2001, John Levon wrote: > is getInset still failing to find insets after a par->isInset returns true ? Yes it would seem so. Open a new document, insert a figure float and put something inside it (optional I think). Open another document. Close the document you just opened. re

Re: reliability of par->getInset()

2001-11-22 Thread Allan Rae
On Thu, 22 Nov 2001, John Levon wrote: > Unfortunately I've not yet been successful in dragging some info out of people for > the bugs I've looked at (breaking line with footnote etc., and crash when deleting > row in tabular) ... breaking/joining line with any inset etc. should now be fixed. A

Re: reliability of par->getInset()

2001-11-22 Thread John Levon
On Thu, Nov 22, 2001 at 05:13:27PM +1000, Allan Rae wrote: > On Thu, 22 Nov 2001, John Levon wrote: > > > is getInset still failing to find insets after a par->isInset returns true ? > > > > There is still a lot of code that checks getInset return is non-null before > > using it, and it looks ug

Re: reliability of par->getInset()

2001-11-21 Thread Allan Rae
On Thu, 22 Nov 2001, John Levon wrote: > is getInset still failing to find insets after a par->isInset returns true ? > > There is still a lot of code that checks getInset return is non-null before > using it, and it looks ugly. > > How old is the code that could cause this to happen; is it still

reliability of par->getInset()

2001-11-21 Thread John Levon
is getInset still failing to find insets after a par->isInset returns true ? There is still a lot of code that checks getInset return is non-null before using it, and it looks ugly. How old is the code that could cause this to happen; is it still there ? thanks john -- "Your superior intelle