***Server sets mode: +nt
***Server sets mode: +nt
***rekado_ is now known as rekado
<mwette>I'm working on a test script for a patch to guile-2.2.4. It works if I manyally start up meta/guile and load, but "make check fails". What gives? Here is evidence: http://paste.debian.net/1041331/ <janneke>mwette: i found a really small/silly bug with 0.86.0 while trying to create GuixSD packages <mwette> janneke: thanks -- I am not surprised. I made a number of changes. <janneke>mwette: i'm amazed how easy the upgrade was for me -- i tried quite some intermediate c99dev commits, tried to patch <mwette>janneke: c99dev is behind the times. I've working on ffidev and nxdev, but merging with master often. <mwette>janneke: Thanks so much. And happy to see your progress. Sounds like you have something fairly robust now? <mwette>janneke: sweet - adding it to the test-suite *janneke creates packages with patch for GuixSD and wip-bootstrap <janneke>mwette: robust...indeed -- my target for Mes is being compatible with Guile, while aiming for a small footprint and have enough to support Nyacc <janneke>as we experienced, there is a playing field between Mes supporting Nyacc, and Nyacc staying minimal (not using regexp in lex.scm) to make it easier for Mes <janneke>and indeed, i'm so happy that Nyacc 0.86.0 is a drop-in replacement, when using Guile and Mes alike! <janneke>...packages not built yet; let me be just a little bit careful while cheering :-) <mwette>wingo: Q on foldts*-values: If, in the DOWN procedure, I replace a tree '(ident "x") with '(lexical x x-123) do you think I should return (values '() '(lexical x x-123) ...) and handle '() in the UP procedure or return something like (values '(REPLACED) '(lexical x x-123) ...) and handle REPLACED in the UP procedure? The former seems more natural but it's then not SXML. <mwette>janneke: The fact that the (nyacc parse) code is running for you is good news to me. I had to rework that module a lot to get it work in interactive mode.