IRC channel logs


back to list of logs

<daviid>RhodiumToad: fwiw, i'd keep those as close as the original as possible, so i'd rather say: COMMENT= G-Golf is a Guile Object Library for GNOME and devel/g-golf/pkg-descr, somethig like: "G-Golf is a tool to develop fast and feature-rich graphical applications, with a clean and recognizable look and feel. Here is an overview of the GNOME platform libraries -
<daviid> - all accessible using G-Golf."
<daviid>RhodiumToad: maybe add that, i the descr, G-Golf comes with some examples, screenshots available at [ot something the like ...]
<daviid>RhodiumToad: but no big deal, the big deal is to get it in freebsd :)
<RhodiumToad>convention for COMMENT is not to include things like "$FOO is a ", but "Guile Object Library for GNOME" works
<RhodiumToad>hm, I should probably arrange to install the examples
<old>no way for pipeline to output to a desired port?
<old>instead of returning a new one
<daviid> RhodiumToad: imo "G-Golf - a Guile Object Library for GNOME" in the comment is better - i'd change the description as i suggested, can't you point to url in the description?
<daviid>RhodiumToad: wrt examples, not sure, as things are, they are not 'installed' upstream either, they are just part of the tarball - as a distro, i'd build anotrher package, g-golf-examples, or maybe g-golf-gtk-4-examples, g-golf-adaita-1-examples .. i don't know ...
<RhodiumToad>the comment isn't supposed to include the name of the package
<daviid>ok, then 'a Guile Object Library for GNOME.'
<RhodiumToad>it's always displayed along with the package name, e.g. "g-golf-guile30-0.8.0.a.4 Guile Object Library for GNOME"
<RhodiumToad>or possibly with the port name, as "devel/g-golf Guile Object Library for GNOME"
<daviid>guile30 is strange, i wish i could run that version as well :):) - freebsd, get yourself a trip in the future ... just joking ofc but more seriously, i'd invert the order, like g-golf-0.8.0-a.4-guile-2.2, g-golf-0.8.0-a.4-guile-3.0
<daviid>RhodiumToad: and fwiw, i'd keep the 'a' letter in the commet "a Guile Object Library for GNOME"
<RhodiumToad>I don't have a choice about how the package name goes
<RhodiumToad>starting with "a" is also not allowed for COMMENT
<daviid>ah ok - no problem
<RhodiumToad>(I didn't make the rules)
<RhodiumToad>the "guile30" is because I'm not sure that . is allowed in flavor names and in any event it is not conventional to use it
<RhodiumToad>I could have made it guile30-g-golf-0.8.etc
<daviid>oh no, g-golf must come first ...
<RhodiumToad>right, that was my initial thought :-)
<RhodiumToad>though other language flavors tend to use a prefix, like py39-foo or lua54-foo
<daviid>i am agsinst this 'flavor' - i think it should be prohibited by law to let distro change the original 'prefix name' :) and in this case, it is also ridiculous, since g-golf actually is an acronym that has guile 'in it' ...
<daviid>thanks god, debian doesn't do this ...
<RhodiumToad>the "flavor" in this case represents the guile version it was built for
<RhodiumToad>ghrm. and I can't use shebangfix to fix up the examples
<daviid>ok, you got it already though, the point is ... g-golf-0.8.0-a.4-guile-2.2, g-golf-0.8.0-a.4-guile-3.0 is much better imo ...
<daviid>what do you mean 'fix up examples' ?
<RhodiumToad>they all have exec guile -e ...
<RhodiumToad>"guile" might not exist, only "guile3" or "guile-3.0" etc.
<daviid>they were not supposed, as things are at least, installed, but you can change that to exec $GUILE no?
<daviid>*to be installed
<RhodiumToad>right, I just have to inplace edit them all
<RhodiumToad>normally that's done via an existing port tool called "shebangfix", but that only looks at the first line
<daviid>oh i see - can't you patch it?
<daviid>i mean patch "shebangfix"
<daviid>or write a mini guile script that does the change to what you need ... easy enough i guess
<RhodiumToad>a quick find ... -exec sed ... should do
<RhodiumToad>ok, examples work
<count3rmeasure>hey all, random lisp question that may have a guile implementation! Does anyone know of lisp bindings to libraries that would facilitate automated web accessibility testing? Think like selenium