On Sun, Mar 19, 2017 at 12:38:49PM +0000, José Abílio Matos wrote:
> On Wednesday, 15 March 2017 19.06.07 WET Scott Kostyshak wrote:
> > Attached is a minimal example from here:
> > 
> > http://tex.stackexchange.com/questions/358326/my-font-shows-up-in-the-font-l
> > ist-but-the-document-wont-compile
> > 
> > It was created with 2.0.8.1.
> > 
> > When I open it in 2.1.0 or 2.2.x I get:
> > 
> > Warning: Malformed LyX document: Can't find \use_mathdots.
> > Warning: Malformed LyX document: Missing \suppress_date.
> > Warning: Malformed LyX document: Can't find \use_package.
> > Warning: Malformed LyX document: Can't find \use_package amsmath.
> > Warning: Malformed LyX document: Can't find \use_package.
> > Warning: Malformed LyX document: Can't find \use_package.
> > Warning: Malformed LyX document: Can't find \use_package.
> > Warning: Malformed LyX document: No \font_sf_scale!
> > Warning: Malformed LyX document: No \font_tt_scale!
> > 
> > Is the document indeed malformed or is there possibly a bug in lyx2lyx?
> > 
> > Scott
> 
> As Enrico said usually lyx generated files have a larger header. So if this 
> file has been produced by lyx it is not in its pristine form.
> 
> With that said we could probably improve the warnings.
> 
> One example is when there is a new file format change that introduces new 
> header elements. The corresponding lyx2lyx change should add those headers 
> with the default values. I am not sure that we do this for all those cases.
> 
> What happens when lyx reads a file where those headers are not present is to 
> assume the default values. It happens sometimes that we change the default 
> values between major versions, a file that has the headers set is immune to 
> the change in the default values because those only apply to new files.
> 
> One other problem with labeling this files as malformed is that there are 
> several levels of malformed, from the the minor warning to the case where lyx 
> is not able to read the file.
> 
> OK, that is all for the little rambling that I had one this issue. Thanks for 
> reading it. :-)

Good thoughts. Once I realize that a file seems malformed, I just move
on. But you are right to think about how can we minimize the damage.

Scott

Attachment: signature.asc
Description: PGP signature

Reply via email to