[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Comments on 0.56

> Following the new release, I've just updated the links on the working
> group homepage at www.tug.org.

Thanks. 

> Now for some observations concerning the new release.
> 
> 1. The new markers: 
> 
> Although, the new markes are potentially nice idea if you insist on
> coveying the meaning unambiguosly, I personally find them a little 
> too distracting to work with, especially if there are several rows of
> missing glyph markers, e.g. if the lowercase upright Greek is missing.
> An additional problem is that you add a dependency of scalable fonts
> on a size-dependant MF font which provides the markers, while the
> black boxes were generated by glyph rules at the DVI level.

Yes this was also my concern, but I think I'm following Thierrys 
argumentation here: If they show up in print, they *should* be distracting,
because something is wrong. Also I'm only using one design size for the
markers: All fonts use markers10. For font (not encoding) tables, it 
would probably be best to leave the places empty, but I don't think there
is a way to achieve that from within TeX.
 
> I might be prepared to use the new markers for the EuroTeX paper for
> clarification, but I would prefer to return to black boxes otherwise
> (and preferably a bigger box for the few control glyphs and a smaller
> one for those many missing glyphs.)
> 
> 2. The new filenames: 
> 
> Using lowercase filenames for everything is apparently the politically
> correct solution for the current LaTeX, but now I'm finding myself
> getting lost in the etx and mtx directories, if the old msam, msbm 
> is sorted between the new MS1, MS2, and MSP.  I'd suggest introducing
> some subdirectoires below etx and mtx, along the lines
> 
>   etx/generic  (8r, ot1, t1, oml, oms, omx)
>   etx/ams      (msam, msbm, lasy)
>   etx/euler    (eufrak, euscr, euex)
>   etx/symbol   (psy, psyupright, psyitalic)
>   etx/mma      (math1, ..., math5)
>   etx/mathime  (my1mtmi, my2mtsy, my3mtex, ...)
>   etx/lucida   ...
>   etx/newmath  (mc, mcraw, mcin, msp, mspraw, mxp, mpxraw, mxpvar)
> 
>   mtx/sizes    (size1000, size0900, ...)
>   mtx/generic  (8rto8ritalic, ot1toot1upright omstomsam, ...)
>   mtx/mc
>   mtx/msp
>   mtx/mxp
>   ...
> 
> An alternative arrangement would be mtx/newmath, mtx/symbol, mtx/mma,
> mtx/mathime, mtx/lucida, etc.  It remains to be seen what is easier 
> to manage.

Yes, we should try something like that.
 
> 3. The distribution:  
> 
> - The lasy subdirectory contains files called blasy5, ... blasy10.  
>   Could these be renamed to lasyb5, ... lasyb10, so that they will
>   get installed under pk/<device>/public/latex rather than udner
>   pk/<device>/bitstream/unknown by MakeTeXPK?

That should be easy.

> - The bex and bams subdirectories both contain msamb10, msbmb10?
>   Which version is the correct one?  (bams contains sizes 5..10,
>   but bex is newer)

Don't know, I'll have to check.
 
> 4. The documentation:
> 
> There is a still a problem with the gray area in the font tables in
> bigdoc.dvi.  I've resent the patch for bigdoc.tex and fontchart.tex
> to Matthias which I've posted here shortly after the last release.
> Anyway, I've installed a modified version of bigdoc.tex at ww.tug.org.
> 
> Given the new layout-<whatever>.dvi files, I wonder if the font tables
> in bigdoc.dvi are needed at all, or whether each of the layout.dvi
> files now represents a chapter of bigdoc.dvi.

One line in my todo list actually reads: `Update bigdoc'. 
Currently we have more or less

- newmath.dtx/.fdd:  Dokumentation of the LaTeX interface
- layout-<whatever>: Font charts for each layout
- bigdoc:            Dokumentation on the Encodings and
                     Design goals, decisions and more 
                     font-related things.

While newmath.dtx/.fdd is OK (I guess), layout-<whatever> is just
a half-an-hour hack to get font tables for each layout in this release.
bigdoc has been neglected for some time and  will need an update.  

> Yet another problem is that the layout.dvi files are generated
> automatically for all font tables, including those that are
> non-existent in a given math layout.  I suppose a better solution
> would be to modify the LaTeX interface, so that the unavailable
> versions of e.g. MS1, MS2 and MX1 in mathptm or concrete/bold
> would be mapped to a dummy font rather than having the CM layout
> silently substituted.  Why not use the AMS "dummy.tfm" for this?

>From a usability standpoint, wouldn't it sometimes be better to have a symbol
from cm (even if it isn't a perfect match) than no symbol at all ? 
And would using dummy.tfm solve the table-generation problem ?
I guess you would just get empty tables, no ?

I'll have to think about that.

Matthias

PS I didn't look at the new Mathematica package yet. Does it have
an extensible font? If so, a comparison to xsb would be interesting. 

-- 
Matthias Clasen, 
Tel. 0761/203-5606
Email: clasen@mathematik.uni-freiburg.de
Institut fuer Mathematik, Albert-Ludwigs-Universitaet Freiburg