On Mon, Jun 03, 2002 at 11:27:35AM +0200, Andre Poenitz wrote:
 
> On Mon, Jun 03, 2002 at 12:12:18PM +0300, Martin Vermeer wrote:
> > Um, "AMS user detection" goes wrong a lot more easily. Trust me, I know
> > users :-)
> > 
> > I think auto-detection is a definite must. That, or always include AMS
> > (the price is small), *and* don't allow the user to accidentally turn it 
> > off in either lyxrc or the document.
> 
> I still don't agree. We could have switched it on by default (which is, as
> you said, cheap), but allow the "power user" to disable it at his own risk.
> 
> We can't always use AMS since it is incompatible in some cases with
> "plain LaTeX" or other packages for that matter. So there must be a way for
> the user to scwitch it off.
> 
> > I like the validation mechanism and think we should keep it. The AMS 
> > dialog should only allow the user to turn AMS on when it is not
> > autodetected, never in the other direction.
> 
> That's plainly stupid. This would render LyX useless for a couple of
> people. And I for one hate a program that tries to tell me that it knows
> better then me and leaves me no chance to correct that behaviour. If that
> is consensus, we could as well drop LyX development and ask those nice
> people in Redmond whether they need some help.
> 
> > And as for the user defined \binom problem, this would only hit
> > knowledgable users that could be trusted to define \mybinom instead :-)
> 
> So you propose to let people to work around things that could be done
> properly (and even would require /less/ code)?

I would indeed rather inconvenience a few smart people to doing that, 
than have a lot of not-so-smart people curse LyX for mysteriously refusing 
to produce legal LaTeX.

So, by all means put in the option of suppressing the validation mechanism, 
*but* make it very separate from the current "Use AMS Math" button and clearly 
belonging to the guru department. Validation exists for a reason.

Trust me, I have worked with non-guru users. So have our Redmond friends,
extensively :-)

Martin

Attachment: msg39011/pgp00000.pgp
Description: PGP signature

Reply via email to