#general
- e
enough-analyst-54434
09/04/2015, 8:22 PMand modern pex has --help-variables - e
enough-analyst-54434
09/04/2015, 8:23 PMwhich would show something like you said - e
enough-analyst-54434
09/04/2015, 8:23 PMPEX_PYTHON i think is right for pexes - b
bored-art-40741
09/04/2015, 8:23 PMI'm wondering though if the chroots are a no-man's-land here - b
bored-art-40741
09/04/2015, 8:23 PMWhere pex env variables aren't respected for interpreter selection and our virtualenv isn't being used either - e
enough-analyst-54434
09/04/2015, 8:24 PMIf you want path control, just plumb it? - e
enough-analyst-54434
09/04/2015, 8:24 PMWe have done so for java distribution - e
enough-analyst-54434
09/04/2015, 8:24 PMits easy to do the same for our chroots - b
bored-art-40741
09/04/2015, 8:25 PMYeah, like I said this is strictly for the purpose of making pdb usable when debugging tests, I didn't want to go all feature-hammer on it - e
enough-analyst-54434
09/04/2015, 8:25 PMright now you can only use python-setup options to contrain version range of CPython - b
bored-art-40741
09/04/2015, 8:25 PMJust wanted a local hack to make debugging less painful - e
enough-analyst-54434
09/04/2015, 8:25 PMBut also its a feature you went off and wrote a whole new pex partially for! - e
enough-analyst-54434
09/04/2015, 8:25 PMpath control - e
enough-analyst-54434
09/04/2015, 8:25 PMwhich pex has - e
enough-analyst-54434
09/04/2015, 8:25 PMand pants just doesn;t plumb - b
bored-art-40741
09/04/2015, 8:27 PMSure... but just to be clear we wrote vex/vend for production path control. It's all about super carefully locking down an environment. I see that we can plumb this for tests... and then proceed to not actually use it in checked in code, which begs for it to be deleted. - b
bored-art-40741
09/04/2015, 8:27 PMWhich is why I was asking for a quick hack and not a feature - b
bored-art-40741
09/04/2015, 8:28 PMIn this circumstance it's strictly useful for locally selecting an interpreter that was compiled with readline - e
enough-analyst-54434
09/04/2015, 8:28 PMYeah - ok, right now you need --python-setup-- ? and an exact version - b
bored-art-40741
09/04/2015, 8:28 PMYeah I think i could get this to work - b
bored-art-40741
09/04/2015, 8:28 PMSince the patch versions are different - e
enough-analyst-54434
09/04/2015, 8:28 PMand that version at the head of PATH before others of the same version but not what you eant - e
enough-analyst-54434
09/04/2015, 8:29 PMI did this just a few weeks ago to bisect a bug from 2.7.4 to 2.7.10 - e
enough-analyst-54434
09/04/2015, 8:29 PMwhich turned out not to be a python patch level version issue at all of course - e
enough-analyst-54434
09/04/2015, 8:30 PMThis one: --python-setup-interpreter-requirement - w
witty-crayon-22786
09/04/2015, 9:49 PM@jolly-chef-92794: i'm still seeing the double-slashes in published jars - w
witty-crayon-22786
09/04/2015, 9:50 PMso it seems like the source of the issue might be elsewhere - w
witty-crayon-22786
09/04/2015, 9:52 PMoooh. - w
witty-crayon-22786
09/04/2015, 9:52 PMyour patch is in the java code, but it's not included yet. - w
witty-crayon-22786
09/04/2015, 9:53 PM@jolly-chef-92794: guessing you're already working on that release?