On Jan 23, 5:39 am, "William Stein" <[EMAIL PROTECTED]> wrote:
> On Jan 22, 2008 7:17 PM, mabshoff
>
> <[EMAIL PROTECTED]> wrote:
>
> > But in the end it boils down to this: Do people want universal
> > binaries, considering the size tradoff? I do think so, but I hope this
> > isn't like the live-CD where everybody says that it is a good idea but
> > when push comes to shove few people step up and use it.
>
> Wow, that's a really good question.
:)
> Given that the entire point of universal
> binaries is to prevent people from accidentaly downloading the binary for
> the wrong architecture and getting frustrated when it doesn't work, maybe
> before (or in parallel with) actually working hard to do universal binaries we
> should think about other approaches to solving that problem. E.g.,
> is there a way to tell with high probability from javascript which
> architecture
> (ppc or intel) the visitor is using? (Probably yes, since many websites
> already seem to do this.)
Sure, that is possible.
> If so, we could just have the download page
> immediately give only a link to exactly the right binary. It seems like this
> would be well worth doing anyways and if it works right it will just solve
> the whole problem that we're trying to solve with universal binaries.
Well, I hate it when websites assume that I use the operating system I
am browsing with, so having something like:
You seem to be running X: [drop down box for binary X preselected]
[Download Button]
But still give the used an option to select something else.
Universal binaries is the Apple way to distribute things and there are
still plenty of PPC boxen around, so offering universal binaries if it
is little work seems to be worth it. The interesting question is what
we will do when we have x86-64 bit and maybe even later ppc64 support.
Do we offer a quad binary then? Other projects I follow do some times
offer arch specific binaries, but also duo or quad-universal binaries.
Does that seem like a workable compromise regarding the space/
usability effort?
> William
Cheers,
Michael
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
