➜
This issue tracker has been migrated to GitHub,
and is currently read-only.
For more information,
see the GitHub FAQs in the Python's Developer Guide.
Created on 2017-02-25 05:59 by serhiy.storchaka, last changed 2022-04-11 14:58 by admin. This issue is now closed.
| Pull Requests | |||
|---|---|---|---|
| URL | Status | Linked | Edit |
| PR 289 | merged | ncoghlan, 2017-02-25 07:49 | |
| Messages (8) | |||
|---|---|---|---|
| msg288542 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) *  |
Date: 2017-02-25 05:59 | |
Importing webbrowser has a side effect. It outputs a message on stderr. >>> import webbrowser /usr/bin/xdg-settings: 1: /usr/bin/xdg-settings: kreadconfig5: not found This is a regression introduced in issue24241. |
|||
| msg288543 - (view) | Author: Nick Coghlan (ncoghlan) * |
Date: 2017-02-25 06:56 | |
I think an appropriate way of handling this would be to route both stdout and stderr in the subprocess to subprocess.DEVNULL. |
|||
| msg288544 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * |
Date: 2017-02-25 07:32 | |
Just stderr. stdout contains the requested value. And I think that an external program shouldn't be ran at import time. It would be better to defer its run until its output is required. |
|||
| msg288547 - (view) | Author: Nick Coghlan (ncoghlan) * |
Date: 2017-02-25 07:41 | |
Ah, and I think I'm not seeing this locally because I have both Gtk and KDE components installed (my main desktop is KDE, but Firefox and various other components are still Gtk apps). |
|||
| msg288548 - (view) | Author: Serhiy Storchaka (serhiy.storchaka) * |
Date: 2017-02-25 07:51 | |
I see this because my main desktop is KDE, but optional libkf5config-bin package is not installed. |
|||
| msg288549 - (view) | Author: Nick Coghlan (ncoghlan) * |
Date: 2017-02-25 07:56 | |
I considered that problem of "subprocess invocation as a side-effect of import", but in this case the output is needed by the immediately following "register_X_browsers()" call, as it affects how those browsers get inserted into webbrowser._tryorder. That means there isn't any obvious way to delay it without a fairly significant refactoring of how the module works :( |
|||
| msg288552 - (view) | Author: Nick Coghlan (ncoghlan) * |
Date: 2017-02-25 09:56 | |
The specific problem reported has been resolved, so closing this one. If anyone's able to figure out a way to defer this to post-import somehow, I think that would make a good enhancement proposal. |
|||
| msg290402 - (view) | Author: Nick Coghlan (ncoghlan) * |
Date: 2017-03-24 23:46 | |
New changeset 140792bd514ee4ba739fda899785bea3ce746f05 by Nick Coghlan in branch 'master': bpo-29644: suppress subprocess output from webbrowser (#289) https://github.com/python/cpython/commit/140792bd514ee4ba739fda899785bea3ce746f05 |
|||
| History | |||
|---|---|---|---|
| Date | User | Action | Args |
| 2022-04-11 14:58:43 | admin | set | github: 73830 |
| 2017-03-24 23:46:20 | ncoghlan | set | messages: + msg290402 |
| 2017-02-25 09:56:30 | ncoghlan | set | status: open -> closed resolution: fixed messages: + msg288552 stage: resolved |
| 2017-02-25 07:56:53 | ncoghlan | set | messages: + msg288549 |
| 2017-02-25 07:51:04 | serhiy.storchaka | set | messages: + msg288548 |
| 2017-02-25 07:49:45 | ncoghlan | set | pull_requests: + pull_request259 |
| 2017-02-25 07:41:13 | ncoghlan | set | messages: + msg288547 |
| 2017-02-25 07:32:07 | serhiy.storchaka | set | messages: + msg288544 |
| 2017-02-25 06:56:21 | ncoghlan | set | messages: + msg288543 |
| 2017-02-25 05:59:09 | serhiy.storchaka | create | |