Re: row-breaking question

[Richard Heck]

On 01/26/2017 08:20 AM, Jean-Marc Lasgouttes wrote:
> Le 26/01/2017 à 10:54, Jean-Marc Lasgouttes a écrit :
>> I am not sure. The cause for 3 is that the code fails to account
>> precisely for the available space on next row (because when it evaluates
>> whether breaking would be a good idea. I just have to find the time to
>> write a patch.
>
> I pushed a fix at 8491962c6bc1. Scott, I would appreciate testing if
> you have time. This is eventually candidate for stable.

Let me know when you think eventually comes.

Richard

Re: row-breaking question

[Jean-Marc Lasgouttes]

Le 26/01/2017 à 10:54, Jean-Marc Lasgouttes a écrit :

I am not sure. The cause for 3 is that the code fails to account
precisely for the available space on next row (because when it evaluates
whether breaking would be a good idea. I just have to find the time to
write a patch.


I pushed a fix at 8491962c6bc1. Scott, I would appreciate testing if you 
have time. This is eventually candidate for stable.


JMarc

Re: row-breaking question

[Jean-Marc Lasgouttes]

Le 26/01/2017 à 00:48, Guillaume Munch a écrit :

They are mostly not desired.
* 3 is clearly a bug
* 2 and 5 should be avoidable, but it requires to have some knowledge
that breaking before the inset is more desirable than in the middle of
the text. It might be doable. I have code somewhere to allow each inset
to declare whether it allows breaking before/after.



This is the sort of behaviour which is fixed (IIRC) by the improvement
to row breaking described in
.


I am not sure. The cause for 3 is that the code fails to account 
precisely for the available space on next row (because when it evaluates 
whether breaking would be a good idea. I just have to find the time to 
write a patch.


Concerning my other remark, I do not think this will be fixed by this 
improvement. This is a separate question, I believe.


JMarc

Re: row-breaking question

[Guillaume Munch]

Le 21/01/2017 à 20:22, Jean-Marc Lasgouttes a écrit :

Le 21/01/2017 à 19:33, Scott Kostyshak a écrit :

The attached sequence of screenshots show the phases of row-breaking
that happens as I keep typing " abc" in the footnote. The word "OLS"
seems to jump around a lot and I find that (mildly) annoying.

Are all six phases of the row breaking desired?


They are mostly not desired.
* 3 is clearly a bug
* 2 and 5 should be avoidable, but it requires to have some knowledge
that breaking before the inset is more desirable than in the middle of
the text. It might be doable. I have code somewhere to allow each inset
to declare whether it allows breaking before/after.



This is the sort of behaviour which is fixed (IIRC) by the improvement
to row breaking described in
.

Re: row-breaking question

[Scott Kostyshak]

On Sat, Jan 21, 2017 at 08:22:52PM +0100, Jean-Marc Lasgouttes wrote:
> Le 21/01/2017 à 19:33, Scott Kostyshak a écrit :
> > The attached sequence of screenshots show the phases of row-breaking
> > that happens as I keep typing " abc" in the footnote. The word "OLS"
> > seems to jump around a lot and I find that (mildly) annoying.
> > 
> > Are all six phases of the row breaking desired?
> 
> They are mostly not desired.
> * 3 is clearly a bug
> * 2 and 5 should be avoidable, but it requires to have some knowledge that
> breaking before the inset is more desirable than in the middle of the text.
> It might be doable. I have code somewhere to allow each inset to declare
> whether it allows breaking before/after.

OK, let me know if I should make a bug report or test anything.

By the way, I know this stuff is complicated. If I report something,
it's just because I'm not sure if it is known.

Scott


signature.asc
Description: PGP signature

Re: row-breaking question

[Jean-Marc Lasgouttes]

Le 21/01/2017 à 19:33, Scott Kostyshak a écrit :

The attached sequence of screenshots show the phases of row-breaking
that happens as I keep typing " abc" in the footnote. The word "OLS"
seems to jump around a lot and I find that (mildly) annoying.

Are all six phases of the row breaking desired?


They are mostly not desired.
* 3 is clearly a bug
* 2 and 5 should be avoidable, but it requires to have some knowledge 
that breaking before the inset is more desirable than in the middle of 
the text. It might be doable. I have code somewhere to allow each inset 
to declare whether it allows breaking before/after.


JMarc