On Tue, Jun 11, 2013 at 1:32 PM, Zach Beane xach@xach.com wrote:
Faré fahree@gmail.com writes:
Inferior-shell *should* be able to compile with ASDF 2. If not, it's a bug. Meh: not as easy for me to test as I'd like.
It uses the slash/package name thing. I'm not sure when it was introduced, but my build system doesn't understand it.
Autodetection of foo/bar as meaning "read it from foo.asd" is an asdf3-ism, but systems named like that work on asdf2 (and asdf1) just like any other system defined in a .asd — of which there used to be a lot even before this convention was invented. i.e. once defined asdf can use them, but before the .asd file is loaded asdf1&2 have no idea where to look for them.
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org Only a fool tests the depth of the water with both feet. — African proverb
Faré fahree@gmail.com writes:
On Tue, Jun 11, 2013 at 1:32 PM, Zach Beane xach@xach.com wrote:
Faré fahree@gmail.com writes:
Inferior-shell *should* be able to compile with ASDF 2. If not, it's a bug. Meh: not as easy for me to test as I'd like.
It uses the slash/package name thing. I'm not sure when it was introduced, but my build system doesn't understand it.
Autodetection of foo/bar as meaning "read it from foo.asd" is an asdf3-ism, but systems named like that work on asdf2 (and asdf1) just like any other system defined in a .asd — of which there used to be a lot even before this convention was invented. i.e. once defined asdf can use them, but before the .asd file is loaded asdf1&2 have no idea where to look for them.
I think that's the problem. inferior-shell-test.asd has :depends-on inferior-shell/test, which is defined in inferior-shell.asd.
Zach
Autodetection of foo/bar as meaning "read it from foo.asd" is an asdf3-ism, but systems named like that work on asdf2 (and asdf1) just like any other system defined in a .asd — of which there used to be a lot even before this convention was invented. i.e. once defined asdf can use them, but before the .asd file is loaded asdf1&2 have no idea where to look for them.
I think that's the problem. inferior-shell-test.asd has :depends-on inferior-shell/test, which is defined in inferior-shell.asd.
Interesting bug: because of how union-of-dependencies and maybe-add-tree work (ASDF1-isms), you need to specify inferior-shell *after* inferior-shell/test so it will be loaded *before* as a dependency and will find the other system.
I'm glad this stuff is gone from ASDF3.
In any case, I pushed to inferior-shell changes that will make the tests work with ASDF2 (the loading should already have worked before). I tested with clisp and asdf 2.26 while temporarily renaming away my asdf.asd.
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org To avoid criticism, do nothing, say nothing, and be nothing. – Elbert Hubbard
Faré fahree@gmail.com writes:
Autodetection of foo/bar as meaning "read it from foo.asd" is an asdf3-ism, but systems named like that work on asdf2 (and asdf1) just like any other system defined in a .asd — of which there used to be a lot even before this convention was invented. i.e. once defined asdf can use them, but before the .asd file is loaded asdf1&2 have no idea where to look for them.
I think that's the problem. inferior-shell-test.asd has :depends-on inferior-shell/test, which is defined in inferior-shell.asd.
Interesting bug: because of how union-of-dependencies and maybe-add-tree work (ASDF1-isms), you need to specify inferior-shell *after* inferior-shell/test so it will be loaded *before* as a dependency and will find the other system.
I'm glad this stuff is gone from ASDF3.
In any case, I pushed to inferior-shell changes that will make the tests work with ASDF2 (the loading should already have worked before). I tested with clisp and asdf 2.26 while temporarily renaming away my asdf.asd.
Thanks, build has improved now.
Zach
-
Faré -
Zach Beane