Re: Discrimination

[Peter Simons]

Reuben Thomas wrote:

 > Let us by all means have a regression test suite, but let us not wait
 > for it. We can do all sorts of hacking on the head and it won't
 > greatly inconvenience any of us or anyone else, nor will particular
 > orderings of the projects suggested so far save much work later, I'd
 > suggest.

yes, I agree. A good regression test suite would certainly help a lot to
improve the quality of our contents, but I wouldn't want to make that a
precondition for the cleanup effort. If we build a regression test suite
based on the cleaned-up contents, it will be just as good for all
practical purposes. After all, we are not aiming for an ISO 9000
certificate.


 >> Those one liners come handy in the latter case, even if they are
 >> apparently of scarce use for us.
 >
 > But they don't have to be separate macros, they can be in the docs
 > for the relevant macro being called.

After some consideration, I agree with Francesco and Filippo that those
one-liners are useful in their own right, even if those macros seem
trivial to experienced users like us. For a practical solution, I would
suggest that we obsolete macros only if they are outright broken or if
we feel that they're utterly pointless. For the time being, however, we
shouldn't remove macros just because they're trivial to implement on top
of something else. We have a lot of stuff on our plate already; let's
not attempt to change everything all at once.

Nonetheless, we need to improve the navigation, categorization, and
general structure of the archive. The current category index page is too
big and too unspecific to be helpful.


 > It's certainly worth providing simple macros. But it should
 > nonetheless be possible to obsolete some of the dumber ones by better
 > design. There's a difference between a complicated single line and a
 > simple one!

Yes, that is true.

Take care,
Peter