[tex-live] Bug in updmap?!

Gerben Wierda Gerben.Wierda at rna.nl
Fri Oct 29 14:11:05 CEST 2004


On 29 Oct 2004, at 13:57, Fabrice Popineau wrote:

>
>> I agree completely. updmap on Windows is a port of updmap on unix. The
>> fact that you have not ported it as is now suddenly means that I have
>> to rewrite and add to my stuff and that is a lot of work. Unless
>                                               ^^^^^^^^^^^^^
> ??? Aren't you using Thomas script?

My system sits on top of TeX Live / teTeX and it shields the users from 
the command line. Thus, I have written packages which contain perl 
scripts. These packages are as robust as I can make them (given my 
time) and thus will for instance not call updmap --enable without 
having checked by running updmap --help and figuring out if --enable is 
available. If it is not, I install an updated version. So far so good, 
but some tools have changed between TL2003 and TL2004 in such a way 
(fmtutil for instance) that I cannot just use the TL2004 version in an 
(also supported) TL2003 setting. With new syntax added, I would have to 
do a lot of rewriting supporting even more variants. My package already 
has about 2000 lines of perl. And that is only the TeX i-Package.

>> Personally, I agree with Vladimir. You want a port of the updmap
>> too. Fine. You port the updmap tool. It is not that much work to parse
>> the options as updmap does it with perl, you just can't use getopt.
>
> As I already said: just do it.

Windows is not my area. I stay on Mac OS X and there I can use the unix 
stuff. That you can't on Windows is too bad for Windows (and there is 
more too bad for Windows, but I digress).

> Do you really know what you guys are doing?
>
> This year BachoTeX meeting ended up with a bunch of "recommendations"
> (call them the name you want) to make this project behave more
> correctly. That turned up into a major crisis. Unfortunately, it is not
> by ignoring problems that you solve them. They will just strike back.
>
> You talk about a lot of work. My guess is that you don't know what you
> are talking about.

Let's first get even: Given the fact that I have authored thousands and 
probably tens of thousands of lines of perl, your guess is wrong. I 
think you have no idea what my redistribution does.

Let's now get civilized again: these kind of comments do not help. 
Calling each other dumb is not going to help.

> Most  things  that   are  input into  TeXLive   are   without  any 
> prior
> discussion. Which means the day before releasing it, you see some 
> script
> or  some   new program  coming in  and  I should  spend   my time  to 
> be
> compatible with that?
>
> My position will be very clear: I give up unless someone can prove that
> this mess is going somewhere.

I don't care to much unless the syntax is going to change without 
backward compatibility in which case I will be very pissed off. And I 
can certify that parsing the command line options in your script 
without getopt is a lot less work than me having to contend with yet 
another variation that I have to shield from the user or pay the price 
in an additional load of support questions from people who's scripts 
suddenly have stopped working.

G



More information about the tex-live mailing list