Re: xvfb not built on vax and x68k?

Antti Kantee wrote:
> On Tue, Feb 03, 2009 at 11:21:28PM +0000, David Brownlee wrote:
>> On Tue, 3 Feb 2009, Patrick Finnegan wrote:
>>
>>> On Tuesday 03 February 2009, Brad Parker wrote:
>>>> David Brownlee wrote:
>>>>> 	any
>>>>> 	objections to making those two ports consistent with all
>>>>> 	the others so all NetBSD ports build the Xvfb virtual
>>>>> 	framebuffer X server?
>>>> If it balloons out the build times I'd prefer not, at least on the
>>>> vax port.  None of the vaxen I use have frame buffers.
>>> Well, I think the point of xvfb is that it doesn't require or use a
>>> framebuffer in the machine.
>> 	Yup, exactly. The relative additional time for a build with
>> 	and without xvb is going to be marginal - if you want to
>> 	save build time you should skip building X11 altogether,
>> 	or crossbuild from a faster box. Given that x68k is going
>> 	to have xvfb added that would leave vax as the odd port
>> 	out.
>>
>> 	I'll time three vax builds - one without x11, one with x11
>> 	but no xvfb, and one with x11 and xvfb. These will all be
>> 	crossbuilds from a random x86 box, but as its relative
>> 	times we're looking at that should be fine.
> 
> Just do it.  No point in having kludges in the tree to make native
> build "fast" on the vax or wasting time even thinking about it.

Ooo. Thank you for that kind though. Yes. Doing native builds on a VAX 
is not something we should even consider.

The fact that native builds don't even work, and that this isn't even 
detectable by doing cross-builds is not something we care about either.

In fact, why even bother with the VAX?

I really like the way some people who work on NetBSD think... NOT!

	Johnny

-- 
Johnny Billquist                  || "I'm on a bus
                                   ||  on a psychedelic trip
email: [email protected]             ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol