Dealing with old PRs and Issues
Hi all, I was in a newcomers meeting with a few scipy maintainers the other day, and they mentioned going through old PRs or Issues as a way to help maintainers and contribute. Does anyone have any suggestions about the best way to do this or how to start handling those? For example, this issue (https://github.com/scipy/scipy/issues/1339) from 9 years ago doesn't have much activity, and I'm curious if I can do anything to assist with it since I don't have commit access and am not a maintainer. If anyone has advice or examples that'd be great :) Best, Will Tirone Duke University - Stat Sci
Dear Will, SciPy welcomes all people who want to contribute to the project with open arms. From my perspective there are several ways that one can contribute, depending on your background. 1) We're often in need of people with good domain specific knowledge. E.g. https://github.com/scipy/scipy/pull/15641 is idling because it needs someone who knows a lot about KKT multipliers for constrained optimization, so we can write some test cases. A high level of programming experience is not necessary. 2) There are lots of ways to contribute to documentation 3) If you have a good Python/programming foundation contributing to most issues/PRs is possible, even if you're not an expert on the maths in a given area. To get started I suggest reading: https://scipy.github.io/devdocs/dev/index.html https://scipy.github.io/devdocs/dev/hacking.html https://scipy.github.io/devdocs/dev/dev_quickstart.html Then roll your sleeves up and get going. It's definitely not necessary to have commit rights or be a maintainer to contribute. Are there particular areas/aspects of scipy that you have a passion for seeing improved? With 1339 there were two suggested solutions, one of them may involve input from the numpy team. Getting started on that could involve an email to the numpy dev list asking them what to do with that. Once you've had that discussion a solution may crystallise cheers, Andrew
Hi Andrew, Thanks so much for the reply! I've actually contributed a few times to bugs or documentation changes that are *current *but I'm more curious about what contributors can do for very stale or outdated issues that are still just sitting in the github issue tracker (or a very old PR that for some reason hasn't been closed). Apologies if that wasn't clear the first time. Thanks! Will On Thu, Jun 2, 2022 at 7:30 PM Andrew Nelson <andyfaff@gmail.com> wrote:
Dear Will, SciPy welcomes all people who want to contribute to the project with open arms. From my perspective there are several ways that one can contribute, depending on your background.
1) We're often in need of people with good domain specific knowledge. E.g. https://github.com/scipy/scipy/pull/15641 is idling because it needs someone who knows a lot about KKT multipliers for constrained optimization, so we can write some test cases. A high level of programming experience is not necessary. 2) There are lots of ways to contribute to documentation 3) If you have a good Python/programming foundation contributing to most issues/PRs is possible, even if you're not an expert on the maths in a given area.
To get started I suggest reading:
https://scipy.github.io/devdocs/dev/index.html https://scipy.github.io/devdocs/dev/hacking.html https://scipy.github.io/devdocs/dev/dev_quickstart.html
Then roll your sleeves up and get going. It's definitely not necessary to have commit rights or be a maintainer to contribute. Are there particular areas/aspects of scipy that you have a passion for seeing improved?
With 1339 there were two suggested solutions, one of them may involve input from the numpy team. Getting started on that could involve an email to the numpy dev list asking them what to do with that. Once you've had that discussion a solution may crystallise
cheers, Andrew
_______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org To unsubscribe send an email to scipy-dev-leave@python.org https://mail.python.org/mailman3/lists/scipy-dev.python.org/ Member address: willtironedev@gmail.com
Hi Will, I am happy to see that you are exploring ways to contribute :) We don’t have a “stale” PR/issue policy to automatically close things. As a contributor, what you can do is to: 1. Verify that the issue or change proposal is still relevant. After a few years, sometimes the function was removed or the issue was fixed elsewhere. 2. Read the comments to understand what are the blockers to move forward. As Andrew pointed out, most of the time we just lack reviewers with the expertise to do that critical job. 3. Do something. If you have the expertise, you can review, comment, ask for clarification on some points or even propose to take over a PR. For instance, if you identified that something should be closed, you can ping a maintainer on the issue explaining why it should be closed. Please keep in mind that we tend to stay on the conservation side when it comes to resolving issues/PR. We have a lot of half finished PRs and we don’t necessarily want to close them even if the original author disappeared. If someone, later, is interested in the issue at hand, it can help to not start from scratch and have a reference with a start of review/conversation with maintainers. Thanks again Will. Cheers, Pamphile
On 03.06.2022, at 04:32, Will Tirone <willtironedev@gmail.com> wrote:
Hi Andrew,
Thanks so much for the reply! I've actually contributed a few times to bugs or documentation changes that are current but I'm more curious about what contributors can do for very stale or outdated issues that are still just sitting in the github issue tracker (or a very old PR that for some reason hasn't been closed). Apologies if that wasn't clear the first time.
Thanks! Will
On Thu, Jun 2, 2022 at 7:30 PM Andrew Nelson <andyfaff@gmail.com <mailto:andyfaff@gmail.com>> wrote: Dear Will, SciPy welcomes all people who want to contribute to the project with open arms. From my perspective there are several ways that one can contribute, depending on your background.
1) We're often in need of people with good domain specific knowledge. E.g. https://github.com/scipy/scipy/pull/15641 <https://github.com/scipy/scipy/pull/15641> is idling because it needs someone who knows a lot about KKT multipliers for constrained optimization, so we can write some test cases. A high level of programming experience is not necessary. 2) There are lots of ways to contribute to documentation 3) If you have a good Python/programming foundation contributing to most issues/PRs is possible, even if you're not an expert on the maths in a given area.
To get started I suggest reading:
https://scipy.github.io/devdocs/dev/index.html <https://scipy.github.io/devdocs/dev/index.html> https://scipy.github.io/devdocs/dev/hacking.html <https://scipy.github.io/devdocs/dev/hacking.html> https://scipy.github.io/devdocs/dev/dev_quickstart.html <https://scipy.github.io/devdocs/dev/dev_quickstart.html>
Then roll your sleeves up and get going. It's definitely not necessary to have commit rights or be a maintainer to contribute. Are there particular areas/aspects of scipy that you have a passion for seeing improved?
With 1339 there were two suggested solutions, one of them may involve input from the numpy team. Getting started on that could involve an email to the numpy dev list asking them what to do with that. Once you've had that discussion a solution may crystallise
cheers, Andrew
_______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org <mailto:scipy-dev@python.org> To unsubscribe send an email to scipy-dev-leave@python.org <mailto:scipy-dev-leave@python.org> https://mail.python.org/mailman3/lists/scipy-dev.python.org/ <https://mail.python.org/mailman3/lists/scipy-dev.python.org/> Member address: willtironedev@gmail.com <mailto:willtironedev@gmail.com> _______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org <mailto:scipy-dev@python.org> To unsubscribe send an email to scipy-dev-leave@python.org <mailto:scipy-dev-leave@python.org> https://mail.python.org/mailman3/lists/scipy-dev.python.org/ <https://mail.python.org/mailman3/lists/scipy-dev.python.org/> Member address: roy.pamphile@gmail.com <mailto:roy.pamphile@gmail.com>
Hi Matt and Pamphile, Thank you SO much for outlining the above, that's extremely helpful. It seems like an interesting way to contribute, so I will start hacking away and see what happens. Thank you again for providing examples and advice! Best, Will On Fri, Jun 3, 2022 at 6:04 AM Pamphile Roy <roy.pamphile@gmail.com> wrote:
Hi Will,
I am happy to see that you are exploring ways to contribute :)
We don’t have a “stale” PR/issue policy to automatically close things. As a contributor, what you can do is to:
1. Verify that the issue or change proposal is still relevant. After a few years, sometimes the function was removed or the issue was fixed elsewhere. 2. Read the comments to understand what are the blockers to move forward. As Andrew pointed out, most of the time we just lack reviewers with the expertise to do that critical job. 3. Do something. If you have the expertise, you can review, comment, ask for clarification on some points or even propose to take over a PR. For instance, if you identified that something should be closed, you can ping a maintainer on the issue explaining why it should be closed.
Please keep in mind that we tend to stay on the conservation side when it comes to resolving issues/PR. We have a lot of half finished PRs and we don’t necessarily want to close them even if the original author disappeared. If someone, later, is interested in the issue at hand, it can help to not start from scratch and have a reference with a start of review/conversation with maintainers.
Thanks again Will.
Cheers, Pamphile
On 03.06.2022, at 04:32, Will Tirone <willtironedev@gmail.com> wrote:
Hi Andrew,
Thanks so much for the reply! I've actually contributed a few times to bugs or documentation changes that are *current *but I'm more curious about what contributors can do for very stale or outdated issues that are still just sitting in the github issue tracker (or a very old PR that for some reason hasn't been closed). Apologies if that wasn't clear the first time.
Thanks! Will
On Thu, Jun 2, 2022 at 7:30 PM Andrew Nelson <andyfaff@gmail.com> wrote:
Dear Will, SciPy welcomes all people who want to contribute to the project with open arms. From my perspective there are several ways that one can contribute, depending on your background.
1) We're often in need of people with good domain specific knowledge. E.g. https://github.com/scipy/scipy/pull/15641 is idling because it needs someone who knows a lot about KKT multipliers for constrained optimization, so we can write some test cases. A high level of programming experience is not necessary. 2) There are lots of ways to contribute to documentation 3) If you have a good Python/programming foundation contributing to most issues/PRs is possible, even if you're not an expert on the maths in a given area.
To get started I suggest reading:
https://scipy.github.io/devdocs/dev/index.html https://scipy.github.io/devdocs/dev/hacking.html https://scipy.github.io/devdocs/dev/dev_quickstart.html
Then roll your sleeves up and get going. It's definitely not necessary to have commit rights or be a maintainer to contribute. Are there particular areas/aspects of scipy that you have a passion for seeing improved?
With 1339 there were two suggested solutions, one of them may involve input from the numpy team. Getting started on that could involve an email to the numpy dev list asking them what to do with that. Once you've had that discussion a solution may crystallise
cheers, Andrew
_______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org To unsubscribe send an email to scipy-dev-leave@python.org https://mail.python.org/mailman3/lists/scipy-dev.python.org/ Member address: willtironedev@gmail.com
_______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org To unsubscribe send an email to scipy-dev-leave@python.org https://mail.python.org/mailman3/lists/scipy-dev.python.org/ Member address: roy.pamphile@gmail.com
_______________________________________________ SciPy-Dev mailing list -- scipy-dev@python.org To unsubscribe send an email to scipy-dev-leave@python.org https://mail.python.org/mailman3/lists/scipy-dev.python.org/ Member address: willtironedev@gmail.com
participants (3)
-
Andrew Nelson -
Pamphile Roy -
Will Tirone