Re: fixincludes templates cleanup

[Zack Weinberg <zack at wolery dot cumb dot org>]

On Sat, Jan 22, 2000 at 10:04:39AM -0800, Bruce Korb wrote:
> 
> Hi Zack,
> 
> Thanks for removing some more of the outdated stuff.
> You removed a tiny bit more than I would like tho:
> 
> Zack Weinberg wrote:
> > ===================================================================
> > Index: fixinc/fixincl.tpl
> > --- fixinc/fixincl.tpl  1999/12/29 22:58:06     1.10
> > +++ fixinc/fixincl.tpl  2000/01/22 00:52:40
> 
> > - *  Description of [=hackname _Cap=] fix
> > + *  Description of [=hackname _cap=] fix
> >   */
> > -#define [=hackname _up #_FIXIDX + #%-32s _printf=] [=_eval _index=]
> > -tSCC z[=hackname _cap=]Name[] =
> > -     [=hackname _cap _krstr=];
> > +[= _IF FIXINC_DEBUG _exist
> > +=]#define [=hackname _up #_FIXIDX + #%-32s _printf=] [=_eval _index=]
> > +[= _ENDIF
> > +=]tSCC z[=hackname _cap=]Name[] =
> > +     [=hackname _krstr=];
> 
> These #defines exist for the theoretical future time
> when some piece of code needs to access the fix description
> for a particular fix.  Thus, for example, you could reference:
> 
>   fixDescList[ AIX_SYSWAIT_FIXIDX ].p_test_desc[0].p_test_regex
> 
> and get the compiled regular expression for the first
> application test for the aix_syswait fix.

I do not agree with this, for three reasons:

1) Such code would not work with the script version, so I doubt it
   will come into existence in the first place.
2) Should we need it, we can always put it back.
3) In the mean time, the only effect of having them there is to bulk
   up the diffs for fixincl.x.  That makes it impractical to mail
   people a patch with the generated files in, but we want to be able
   to do that, because most people don't have autogen lying around.

If you still feel they are necessary I will put them back, but please
consider this argument first.

zw