Am Samstag, 2. September 2006 20:21 schrieb Abdelrazak Younes:
> Georg Baum wrote:
> > That one will be fixed with one of my next patches, but I would be 
> > surprised if it was caused by my last commit.
> 
> Dunno, haven't checked really. I was giving you the full output, sorry.

Yes, I guessed that after trying to find out how my patch could have caused 
these errors ;-)

> >> ..\..\..\trunk\src\graphics\PreviewImage.C(59) : warning 
C4355: 'this' : 
> >> used in base member initializer list
> > 
> > That one is bogus.
> 
> Well, I think it is good practice to avoid that but yes it not harmful.

Yes, I also aim to avoid warnings in my code so that I easily can spot real 
problems. In this case the code really is OK. I don't know how this 
warning can be disabled here, maybe with a pragma?

> >> ..\..\..\..
\trunk\src\frontends\controllers\ControlSpellchecker.C(166) : 
> >> warning C4244: 'argument' : conversion from 'const 
> >> Paragraph::value_type' to 'unsigned char', possible loss of data
> > 
> > This one is not so easy (can digits be something else than ASCII 0-9?)
> 
> What is Paragraph::value_type?

Paragraph::value_type == lyx::char_type == boost::uint32_t.

> I guess it should be a char_type, we have  
> to verify that Aspell will understand that.

The spell checker is probably broken currently.

There are several more places where we assign a lyx::char_type to a char. 
These need to be fixed. I am currently not looking at these but 
concentrating on strings.


Georg

Reply via email to