config.guess weirdness

[Martin R]

I am forwarding an observation by Eric Bray, concerned with making FriCAS an optional (as opposed to experimental) package for sagemath.  It would be great if you could comment, ideally on https://trac.sagemath.org/ticket/23847?replyto=35#comment, otherwise here...

Weirdly, last time FriCAS [https://sourceforge.net/p/fricas/code/2251/ updated their `config.guess`], the version it was updated to does not correspond exactly with the [http://git.savannah.gnu.org/gitweb/?p=config.git;a=blob;f=config.guess;h=1f5c50c0d1529d50b94dc3533ca72a47f0fa5849;hb=47681e2a9b5e8236e38cffe03c85e571782fc080 upstream `config.guess` from the same date], almost as if it were hand-edited and mistakes were made.

[oldk1331]

Hi, you can see the change here:

https://groups.google.com/d/msg/fricas-devel/u7HT8s_lMF8/n6UGeTvyAgAJ

Quote:

They are slightly hacked to produce shorter names.

Yes, originals are from autotools. This version is hacked
to produce slightly shorter names, instead of say

x86_64-unknown-linux-gnu

it gives

x86_64-linux-gnu

END QUOTE


So it seems that we should stick to autotools default then,
since this hack causes problem in Cygwin, and also
update config.guess to latest version.

[oldk1331]

Hi Martin, I saw the "pr.children()" problem in your ticket,
a suggestion:

Why not use "fricas -nosman" instead of "fricas -nox -noclef"
to start FriCAS? Using "fricas -nosman", there's only 1
process, AXIOMsys.

[Martin R]

Thank you for the hint, I'll try this right now!

[Martin R]

For some reason, that doesn't work.  I can start fricas on the command line with fricas -nosman, and in sage manually, but for some reason it breaks the interface :-(

[oldk1331]

On Sat, Oct 7, 2017 at 5:49 PM, 'Martin R' via FriCAS - computer
algebra system <fricas...@googlegroups.com> wrote:
> For some reason, that doesn't work. I can start fricas on the command line
> with fricas -nosman, and in sage manually, but for some reason it breaks the
> interface :-(

From the ticket link, seems that it works now?

Well, I think there's something wrong with "sman", it also causes
a problem in TeXmacs interface.

[Martin R]

Yes, sorry not to notify.  The difference is in the line endings, sman sends \r\n, AXIOMsys only sends \n.  I made it insensitive now, because one may want to start sman to have graphics and hyperdoc.

Thank you!

Martin