mailing-list for TeXmacs Users

[Joseph Coffey <address@hidden>]

Joris van der Hoeven wrote:

> We all agree even though we should try to push publishers harder at
> supporting TeXmacs as a native format. We start to a have some contacts
> at Springer-Verlag, but it would really be helpful to always insist on
> sending papers in the native TeXmacs format.
>  
I think that getting the arxiv to accept texmacs files might be a first 
priority. They are tech savvy and submitting there is generally an early 
stage, before you want to got through a lot of latex tweaking. It is 
difficult to submit there now because, although they accept pdf/ps in 
theory; that generated by texmacs sets off their "generated by Latex" 
warning and requires you to send them emails begging them to accept them 
anyway. I'm not sure how often I (or anyone else) would want to do 
that.   (its a bit of a negotiation- the ps files generated are 
generally too large, and pdf is not text searchable.)

However I think it might be hard to get journals/the arxiv to accept .tm 
files until texmacs has a stable document format. The style files seem 
to change still.  I used two different versions of texmacs and the  
documents generated in the newer version upset the older one. (they were 
the newest version and an earlier in the 1.03 branch)

> > It complains that it doesn't know where macros.sty is. I assume this is
> > because either: The filter doesn't export the macro file.
> >    
>
> This is probably the right behaviour; I tend to interpret style files
> as hyperlinks which need to be exported separately.
>  
Why? why not just give the user a series of "export macro1.ts" queries 
etc... in the minbuffer? Im not sure why the user of TeXmacs should know 
much about Latex (I don't).  Some people use it just to avoid Latex. At 
any rate its looking for macros.sty not .cls (although I don't know the 
difference between the two)

>  
>
> > Then, I manually exported the macro file, and copied all the
> > newcommands inot my pre-amble.
> >    
>
> Yes, you can do that, or export to a separate .cls file. I would have to
> check whether this needs some special additional header information.
>
>  
>
> > The resulting document more or less compiled but it
> > generated many many  errors (35-50).  In addition some portions of my
> > document were lost:
> >
> > at least:
> > my abstract is gonet.
> >    
>
> That is strange.
>  
Stranger still I exported another document and the abstract was exported 
but not the author. (I think it might have been the date; its been a few 
days now.)

>  
>
> > a big coproduct sign is missing.
> >    
>
> I need to see this in order to believe that
>  
Ill try to find and send you the files seperately.

>  
>
> > my bibliography is gone.
> >    
>
> Bibliographies need to be regenerated with BiBTeX.
> One may argue whether this is the best behaviour possible,
> or whether we should generate bibitem statements and forget
> about the bib file.
>
>  
Why not a shell script that just runs bibtex along with the export? I as 
neophyte texmacs user just do what the help tells me.

> > I suspect there are other difficulties as well.
> >    
>
> Please let me know of any other difficulties. We try to do our best at
> making the LaTeX export filter as good as possible, but there are some
> tricky issues.
>
>
>  
I think a lot of the trickyness could be taken care of by exporting 
macros with ensuremath instaed of $$. Admittedly this is based on my 
experience with one document. In general the trouble seems to be that 
TeXmacs is more robust (less sensitive) to say reversing the order of 
strong and math modes.

The other difficulty I have had is that if I ask TeXmacs to number for a 
long article, It exports to asking for the .sty file 
number-long-article.sty but latex and I don't know where that is. I 
would guess other customizations yield similar difficulties.

Thank you so much for your help and excellent software,

Joe Coffey