Re: check with SGML::Parser::OpenSP (and branches)

On Thu, 2005-08-18 at 09:07 +0200, Terje Bless wrote:
> Ville Skyttä <ville.skytta@iki.fi> wrote:
> 
> >I don't think there's much value in being able to pass arbitrary
> >parameters there per se, but the caching mode needs to be configurable.
> 
> Can we figure this out at runtime and Do The Right Thing?

Yes, I came to the conclusion that the above would be doable and the way
to go right after sending my last mail in the morning, on the way to
work.  No need to overengineer this.

See attached patch; the mod_perl test is there just for reference as we
don't support mod_perl in 0.7.x anyway.  But this patch would be
applicable to both 0_7 and HEAD (latter untested, but in principle),
whether we require mod_perl in the latter or not.  No config file
changes, and no _necessity_ to document anything, although it wouldn't
hurt.  Better, WDYT?

> >IMO it's more an issue whether anyone has a mod_perl 1 setup any more
> >to test with.
> 
> I can probably swing that — and give you access to it — if really needed.

Well, I have no personal interest towards getting the validator to run
with mod_perl 1, but if someone has, I can help out.

> Come to that we could probably paralell-install Apache1/mod_perl1 on qa-dev.

Yep... but let's get a decision whether we want to support running
validator 0.8+ _without_ some version of mod_perl, first, ok? :)
Support for running without mod_perl means keeping stuff so generic
anyway that I think the m_p 1 vs 2 vs both question is moot.

My .02€: I would _like_ to support mod_perl only, but I don't see a
technical reason why it would _have_ to be made so.