On Oct 4, 2012, at 17:27, Jeremy Lavergne <jer...@lavergne.gotdns.org> wrote:

>> I've agreed before that we should do this split, in favor of having 
>> variants. I agree it'll be a major pain to do.
>> 
>> The goal would be to have the quartz and x11 subports simultaneously 
>> installable.
>> 
>> But at that point, does anything speak against just making pango/cairo/gtk2 
>> always install both quartz and x11 parts and just get rid of the variants? 
>> Is that even possible?
> 
> Other than avoiding the x11 dependency tree...

Ah. Right. The desire not to have to install x11 bits. I suppose that's a valid 
request.


> I'm not sure if a package somewhere requires -quartz or -x11.

Right, I was wondering about that. But I'm hoping there isn't.


> Some ports catch my eye (their patchfile names):
> * py-enable: no-64-bit-quartz.diff
> * qjackctl: patch-configure-no-x11.diff
> * gecko-sharp2
> * aalib

I haven't looked at these.


> It might also defeat the purpose of some x11 subports:
> * freeciv/freeciv-x11

No, here this gives the user the choice of whether they want to run freeciv in 
x11 mode or not. That's still a valid choice, and one that has to be made at 
compile time for this port (and I suspect other ports).


> Also several ports have a default_variants for -x11:
> * giflib
> * openvrml

No, this is only part of the compatibility code assisting users in upgrading 
from the no_x11 variant. Users not specifying a variant will still get x11 by 
default, as we've so far standardized on with other ports.

> * pspp-devel

This port doesn't mention -x11.

> And of course, all the packages with their own x11 variants to help guide the 
> dependency tree to use x11.

Yes. It had thus far been an unwritten (?) rule that if a port can include x11 
support, then it should be on by default. This rule probably originated at a 
time when x11 was an on/off switch (and thus in line with our desire to turn as 
many useful features on as possible by default), rather than an either/or 
choice with quartz.


_______________________________________________
macports-dev mailing list
macports-dev@lists.macosforge.org
http://lists.macosforge.org/mailman/listinfo/macports-dev

Reply via email to