[tex-live] script for LuaTeX font database updating

Khaled Hosny khaledhosny at eglug.org
Thu Feb 25 23:11:42 CET 2010


On Thu, Feb 25, 2010 at 10:40:45PM +0100, Reinhard Kotucha wrote:
> On 25 February 2010 Khaled Hosny wrote:
> 
>  > On Thu, Feb 25, 2010 at 07:15:30PM +0100, Reinhard Kotucha wrote:
>  > > On 25 February 2010 T T wrote:
>  > >  > Single common script is better. For Unix acting on argv[0] is
>  > >  > standard and allows the -sys version to be a symlink. For
>  > >  > Windows I need --sys switch for our universal wrapper, because
>  > >  > I can't alter argv[0] for scripts (well, for Lua script I
>  > >  > could, but switch will be less hackish :).
>  > > 
>  > > Since there are already some -sys scripts we can't get rid of, I
>  > > think it's worthwhile to consider whether it makes sense to
>  > > follow this scheme, just that such programs behave similar.
>  > 
>  > I'd say we shouldn't keep repeating the mistakes of the past just
>  > because we can't (or unwilling) to fix the old ones.
> 
> I don't think it's a mistake.  It was a design decision and the way it
> was done is reasonable:

I consider a "different behaviour based on executable name" a wrong
design decision and hence, IMHO, a mistake.

> A few years ago there were no distinct config files for users.  A user
> had to do everything manually if he had own fonts or own format
> files.  The -sys programs were introduced in TeTeX-2.0, if I recall
> correctly.
> 
> The only difference between a system-wide and a per-user configuration
> is that files are written to different directories.  Hence, the
> obvious way to add the new functionality without revising the scripts
> themselves was to add wrappers.  The scripts were thoroughly tested at
> this time and providing just wrappers was the safest solution.  Thus I
> think it was a good decision.

I don't understand this, how having "different behaviour based on
executable name" (which still requires editing the script to understand
this, right?) is a less invasive approach than, the more natural,
command line switches?
Andi, remind you, though this behaviour might be consistent through TeX
Live utilities, it differs from the rest of the world, which adds to the
frustration of newcomers (like me, that why I'm complaining, as all this
-sys naming did is confusing me like hell).

> Regarding other scripts, I think that it's not a good idea to invent
> other schemes because that would confuse people.  It's bad enough that
> there is no unique scheme how to process command-line options and it's
> annoying, for instance, that xpdf understands "-fullscreen" but not
> "--fullscreen".

That is because each application is using his own command line options
parsing routines instead of, say, using GNU's getopt. (This reminds me
with an issue I was concerned about; we should have a lua getopt like
library for lua scripts to use to avoid inconstancies like this, I guess
this should be done right now before we start making legacies in the
Never Changing[TM] TeX world).

Regards,
 Khaled

-- 
 Khaled Hosny
 Arabic localiser and member of Arabeyes.org team
 Free font developer


More information about the tex-live mailing list