Re: CVS commit: pkgsrc/math/pari

[Amitai Schlair <schmonz%schmonz.com@localhost>]

On Tue, Feb 16, 2010 at 2:24 PM, Jens Rehsack 
<rehsack%googlemail.com@localhost> wrote:

> you didn't respond to my two review requests in pkgsrc-users@,
> neither to Aleksej's suggestions but ask for changes (which everyone would
> see) to a commit.

I hadn't known I needed to care about pari until your commit, which is
why I missed the start of the discussion and am only just jumping in
now. I agree, it'd have been more helpful if I'd been paying attention
sooner.

> Maybe I'm wrong (and in this case please accept my apology), but I didn't
> see any private question in your requests to modify math/pari.

No apology needed on my account (you're right, there was nothing
inherently private about anything I sent you), but others might not
care for it. Best to ask the sender's permission next time.

>> Your update, while needed, introduced at least one seriously
>> regression in usability.
>
> That's why it had 2 review requests on pkgsrc-users@ - where everyone could
> had respond. And that's why I want this usability regression being discussed
> in public - that everyone could read it and search for it in the archives.

Point well taken.

> I hope there will be some feedback to your patch on the list. The patch is
> fine for me except that x11 will be enabled by default. So you need to turn
> it off manually if you don't want x11 being a dependency.

Thanks for looking at the patch. I was just leaving the previous
behavior as the default, but a PKGREVISION bump is a sensible time to
change the default if we want to. Defaulting to no x11 is fine by me.

> Maybe someone wants tex dependency removed (for the price that no
> documentation will be generated etc.?)

Yes, I may look into this next (but anyone is welcome to beat me to it).

- Amitai