Some comments on CTAN:info/tdscd

Sebastian Rahtz TWG-TDS@SHSU.edu
Mon, 26 Feb 1996 12:58:34 GMT


 > the texi2html documentation to ensure that filenames remain uniqiue
 > when truncated to 8+3.  While I was at it I also regenerated all the
 > Info files for the same reason.  Expect my uploads at ftp.tex.ac.uk.
you are a hero.

 > 2. I had a look at the files you placed on CTAN:info/tdscd and noted
 > the following problems (some of this may be TDS releveant but most
i wont comment on all, but thank you very much for looking at this,
its really very helpful.
 > > /teTeX/texmf/bibtex/bib:
 > > README
 > > harvard
 > > 
 > Where's mrabbrev.bib (from AMS-LaTeX)?
 > Where's xampl.bib (from BibTeX)?
good point. i had forgotten both

 > Should be texmf/doc/generic/babel according to TDS.
 > 
 > > /teTeX/texmf/doc/dc:
 > 
 > Should be texmf/doc/fonts/dc.  In fact, that directory already exists,
 > but the contents is slightly different.  So what's messed up here?
me and Thomas at different times

 > > /teTeX/texmf/doc/latex/base:
 > > 00readme.txt
 > > Makefile.unx
 > > alltt.dtx
 > > alltt.ins
 > [...]
 > 
 > What are all the .dtx and .ins files doing here?  Doesn't the TDS 
 > say explicitly that thex belong into texmf/source/latex/base:
yes, my mistake. many of the same kind to undo
 > > /teTeX/texmf/doc/latex/revtex:
 > 
 > This looks incomplete.  There should be DVI files for all
 > man*.tex files, but there's probably no need for PS files.
yes, that was half-finished. 

 > > /teTeX/texmf/fonts/tfm:
 > [...]
 > > bitstrea
 > > bitstream
an interesting question....

 > > dctt2074.tfm
 > > dctt2488.tfm
 > > jknappen
 > > monotype
 > 
 > Something must have been screwed up here!
indeed...

 > TDS says "texmf/source" (without trailing `s')!  Here's where I would
 > expect to see a lot of stuff in texmf/source/{generic,latex}/<package>/...
 > In fact, you might put all the teTeX/extra-distrib stuff from CTAN here.
Thomas and I talked about whether to include .tar.gz packages of what
we build the TDS tree from, but it would cut down on the CD
space. that extra-distrib is my working area

 > What's this doing here?  "texmf/tex/initex" has been gone long since.
 > This stuff obviously belongs into texmf/tex/latex209/base/...
yup, all a mess, to reo

 > set up texmf.cnf, so that format files by default will also be searched 
 > in the corresponding directory on disk, no just on the CD.
thats true; it isnt set up in texmf.cnf as its stands; will add

 > 1.  I would find the output of "find /teTeX -print | sort" easier
 >     to check than the output of "ls -R /teTeX".  
ok

 > 2.  If time permits in the production cycle, it might be a good idea
 >     to have someone (us?) check the contents of a preliminary CD 
 >     for problems, mismatches, mix-ups, etc. as mentioned above before 
 >     releasing it to the general public.
oh yes, definitely! i have sent a preliminary CD to Thomas already (he
has done revised Make* scripts etc), and i propose in (say) a month to
make 3 or 4 and send out real CDs to people with some time and energy
to look at for a few weeks, as well as posting the list to CTAN. i
have got a supplier yet, but i would guess i'll need to get the final
product to them by the first week of May to be on the safe side.

 > What about including other related software such as Acrobat readers
i am planning to check Acrobat with Adobe.

 > or WWW brosers for those Unix platforms where they are available?
since we do not include GhostScript, isnt this a bit over the top?

but the short answer is that if someone puts together a tar.gz file,
claims responsibility for it, tests it, and it doesnt break anything
else, why not? but i'd be very reluctant to have only partial bin/
support - it would be sad to say "this CD is fully populated only for
linux and solaris". also, its  a slippery slope. why not add gzip,
unzip, netpbm, perl....

sebastian