Github templates #1954

soupault · 2016-02-20T19:57:29Z

Addresses #1949 .
Added templates, moved contributing guidelines from the root.

stefanv · 2016-02-20T21:36:13Z

Just double check, I think Josh might have already filed something similar
(sorry, not at a PC right now).
On Feb 20, 2016 11:57 AM, "Egor Panfilov" notifications@github.com wrote:

Addresses #1949 #1949
.

Added templates, moved contributing guidelines from the root.

You can view, comment on, or merge this pull request online at:

#1954
Commit Summary

Prepared templates

Moved templates

Merge branch 'master' of https://github.com/soupault/scikit-image
into gh_templates

File Changes

R .github/CONTRIBUTING.txt
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-0
(0)

A .github/ISSUE_TEMPLATE.md
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-1
(19)

A .github/PULL_REQUEST_TEMPLATE.md
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-2
(20)

M doc/source/contribute.txt
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-3
(2)

Patch Links:

https://github.com/scikit-image/scikit-image/pull/1954.patch

https://github.com/scikit-image/scikit-image/pull/1954.diff

—
Reply to this email directly or view it on GitHub
#1954.

JDWarner · 2016-02-20T21:41:38Z

Nope! I just raised the issue - it's been very busy for me of late, and I
figured I'd leave an easy PR for someone to tackle.
On Feb 20, 2016 2:36 PM, "Stefan van der Walt" notifications@github.com
wrote:

Just double check, I think Josh might have already filed something similar
(sorry, not at a PC right now).
On Feb 20, 2016 11:57 AM, "Egor Panfilov" notifications@github.com
wrote:

Addresses #1949 <
https://github.com/scikit-image/scikit-image/issues/1949>
.

Added templates, moved contributing guidelines from the root.

You can view, comment on, or merge this pull request online at:

#1954
Commit Summary

Prepared templates

Moved templates

Merge branch 'master' of https://github.com/soupault/scikit-image
into gh_templates

File Changes

R .github/CONTRIBUTING.txt
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-0
(0)

A .github/ISSUE_TEMPLATE.md
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-1
(19)

A .github/PULL_REQUEST_TEMPLATE.md
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-2
(20)

M doc/source/contribute.txt
https://github.com/scikit-image/scikit-image/pull/1954/files#diff-3
(2)

Patch Links:

https://github.com/scikit-image/scikit-image/pull/1954.patch

https://github.com/scikit-image/scikit-image/pull/1954.diff

—
Reply to this email directly or view it on GitHub
#1954.

—
Reply to this email directly or view it on GitHub
#1954 (comment)
.

JDWarner · 2016-02-21T02:12:12Z

Philosophically, I think we need to take a minimal approach to this. These templates are effectively barriers to contribution or even getting bugs reported, which are both behaviors we want to facilitate rather than have another big checklist.

Since this is an important public-facing aspect of our project, I'll be pretty fine grained in my review.

JDWarner · 2016-02-21T02:13:17Z

.github/ISSUE_TEMPLATE.md

@@ -0,0 +1,19 @@
+## Description
+[What is wrong? What the expected and current behaviours are?]


[What is wrong? How would you expect this to work?]

vighneshbirodkar · 2016-02-21T02:27:25Z

How are we supporting new feature requests through this ?

JDWarner · 2016-02-21T02:32:20Z

@vighneshbirodkar Great point. Basically, we have to make it clear nearly all fields are optional using language like "If possible..."

soupault · 2016-02-21T06:14:37Z

@vighneshbirodkar See my message in #1949 . I'm in favor of the approach where many of the blocks have double meaning/purpose.

soupault · 2016-02-21T06:46:43Z

@JDWarner thank you very much for reflecting on this! I've made all proposed changes, except one.
I'd insist that the Reference field should be really presented in ISSUE_TEMPLATE. It'll give a good chunk of related information and will be much appreciated for long-lived issues.
Withal, adding this shouldn't be a heavy effort for reporting person, because he or she is expected to work and know something on the topic. Perhaps, this will also perform as a soft filter to very immature and impulsive proposals/requests, which is (in my opinion) not a bad thing to have.

vighneshbirodkar · 2016-02-22T15:25:06Z

Should we also include a section for attaching images ? Comparison of new and old output makes a lot of things clear.

soupault · 2016-02-22T15:34:24Z

@vighneshbirodkar
For issues, there is already a mention: ...include example code and image(s)....
For PRs, I'd say that the vast majority of them doesn't really change the old behavour, but rather introduce something new. Those, which usually change it, are mostly related to the backend do not require images at all.

vighneshbirodkar · 2016-02-22T15:42:15Z

@soupault Then can we somehow request for images when new features are submitted ? I always try to attach the output from the example script

soupault · 2016-02-22T16:43:55Z

@vighneshbirodkar This should do.

UPD: I take your point, this should help a lot when people are submitting new features. Usually, nobody cares to pin an example of the result to the PR.

JDWarner · 2016-02-22T18:47:37Z

This LGTM. 👍

blink1073 · 2016-02-22T18:50:15Z

.github/PULL_REQUEST_TEMPLATE.md

+## Checklist
+[It's fine to submit PRs which are a work in progress! But before they are merged, all PRs should provide:]
+- [ ] [PEP8 compliance](https://www.python.org/dev/peps/pep-0008/)
+- [ ] Docstrings for all functions


Maybe add this link? https://github.com/numpy/numpy/blob/master/doc/HOWTO_DOCUMENT.rst.txt#docstring-standard

I wish Github had internal URL-shortener...

You can link to it within the document and put the full url at the bottom

We could, but yikes - perusing the actual standard is equivalent to reading a small book. Intimidating! Perhaps directing them to imitate existing docstrings would be more friendly? "Docstrings for functions, with style consistent across the package" or something similar?

I view this more as a getting-started pointer, rather than sending people out with homework ;)

@JDWarner I think the style should be actually aligned with numpydoc, not with random skimage example (which can become out-of-date one day).
I agree that the standard is TLDR 😃 , but (at the same time) is a fairly good document.

All well and true, but it's our responsibility as maintainers to ensure docstring style is up to date and consistent.

It's a good document, but it's just far and away more than any new contributor needs. If a docstring is subtly incorrect or needs a new feature, again, it's our responsibility to point them in the right direction in the course of the PR.

I do strongly feel we don't want to front load this. I was on the fence about including any links at all.

I understand.
Probably the best solution will be to keep the template as is for a while so to see which fields are actually used and which do cause obstacles.
Regarding the docsting guide: it should be better to address the detailed guide in CONTRIBUTING.txt.

blink1073 · 2016-02-22T19:05:04Z

@JDWarner, how do you feel about us adding something like python -m skimage.util.sysinfo that reports something similar to what IPython.sysinfo() provides: https://github.com/ipython/ipython/blob/master/IPython/utils/sysinfo.py

JDWarner · 2016-02-22T19:07:20Z

blink1073 👍, that would be worth investigating. Probably in a separate PR, though.

blink1073 · 2016-02-22T19:36:11Z

What if we linked to this, which demonstrates by example and is the canonical implementation: https://github.com/numpy/numpy/blob/master/doc/example.py

JDWarner · 2016-02-22T19:38:30Z

.github/PULL_REQUEST_TEMPLATE.md

+
+## Checklist
+[It's fine to submit PRs which are a work in progress! But before they are merged, all PRs should provide:]
+- [ ] Style in [compliance with PEP8](https://www.python.org/dev/peps/pep-0008/)


Changing my own mind here, but "Clean style in the spirit of PEP8" might be even better. Mainly considering PEP8 tells us to break it when we feel justified, and it's friendlier than 'compliance'.

ahojnnes · 2016-02-23T08:43:18Z

.github/ISSUE_TEMPLATE.md

@@ -0,0 +1,22 @@
+## Description
+[What is wrong? How would you expect this to work?]


I would reformulate, maybe "Describe your problem, feature proposal, etc."

soupault · 2016-02-23T21:30:27Z

@ahojnnes How do you feel about the updated version?

ahojnnes · 2016-02-24T11:21:12Z

.github/ISSUE_TEMPLATE.md

+[Please provide general introduction to the issue/proposal.]
+
+
+## Reproduction


"Reproduction" sounds slightly misplaced here, but I'll let native speakers comment :-)

It's not much off, but maybe 'Reproducing the issue' would be better.

or "Attachements"

"Steps to reproduce"?

Fine with @stefanv 's version.

ahojnnes · 2016-02-26T09:21:06Z

In general, I think, we should keep this as short as possible. Until now, I am not sure if we really had problems with too many incomplete issue reports?

soupault · 2016-02-27T11:43:53Z

@ahojnnes I suppose the goal is not just to integrate the new technique, but to exploit it to achieve some particular improvements.

soupault · 2016-02-29T20:56:03Z

All things are done. 👻

stefanv · 2016-02-29T20:59:16Z

.github/ISSUE_TEMPLATE.md

@@ -0,0 +1,16 @@
+## Description
+[Please provide general introduction to the issue/proposal.]


provide a general

stefanv · 2016-02-29T21:16:28Z

I think we're converging. Thanks for your patience with our bike shedding!

soupault · 2016-02-29T21:21:49Z

@stefanv I hope that the equilibrium has at least couple of words in the templates 😉 .

stefanv · 2016-02-29T21:25:03Z

@soupault If I had my way, very few ;)

soupault · 2016-03-15T21:10:45Z

One more iteration here 🔨 .

stefanv · 2016-03-15T23:06:05Z

That looks fine to me. Thanks for your patience @soupault !

soupault · 2016-03-19T11:03:03Z

@JDWarner @blink1073 @ahojnnes Any opinions, gentlemen?

blink1073 · 2016-03-19T11:17:10Z

Great, thanks @soupault!

Github templates

soupault added the 🔧 type: Maintenance Refactoring and maintenance of internals label Feb 20, 2016

soupault force-pushed the gh_templates branch from 8822646 to c568b01 Compare February 20, 2016 20:08

JDWarner reviewed Feb 21, 2016
View reviewed changes

blink1073 reviewed Feb 22, 2016
View reviewed changes

soupault force-pushed the gh_templates branch from f003661 to f6f5d5a Compare February 22, 2016 19:33

JDWarner reviewed Feb 22, 2016
View reviewed changes

ahojnnes reviewed Feb 23, 2016
View reviewed changes

ahojnnes reviewed Feb 24, 2016
View reviewed changes

stefanv reviewed Feb 29, 2016
View reviewed changes

Added GitHub templates, moved CONTRIBUTING.md to .github

81970ba

soupault force-pushed the gh_templates branch from c7cef7b to 81970ba Compare March 16, 2016 22:55

blink1073 added a commit that referenced this pull request Mar 19, 2016

Merge pull request #1954 from soupault/gh_templates

34ae680

Github templates

blink1073 merged commit 34ae680 into scikit-image:master Mar 19, 2016

soupault deleted the gh_templates branch March 23, 2016 18:26

soupault mentioned this pull request Apr 2, 2016

Add Issue templates and declutter root directory #1949

Closed

soupault mentioned this pull request Oct 8, 2022

Replace issue templates with issue forms #6554

Merged

		@@ -0,0 +1,19 @@
		## Description
		[What is wrong? What the expected and current behaviours are?]

		@@ -0,0 +1,22 @@
		## Description
		[What is wrong? How would you expect this to work?]

		[Please provide general introduction to the issue/proposal.]


		## Reproduction

		@@ -0,0 +1,16 @@
		## Description
		[Please provide general introduction to the issue/proposal.]

Github templates #1954

Github templates #1954

Conversation

soupault commented Feb 20, 2016

stefanv commented Feb 20, 2016

Added templates, moved contributing guidelines from the root.

JDWarner commented Feb 20, 2016

Added templates, moved contributing guidelines from the root.

JDWarner commented Feb 21, 2016

Choose a reason for hiding this comment

vighneshbirodkar commented Feb 21, 2016

JDWarner commented Feb 21, 2016

soupault commented Feb 21, 2016

soupault commented Feb 21, 2016

vighneshbirodkar commented Feb 22, 2016

soupault commented Feb 22, 2016

vighneshbirodkar commented Feb 22, 2016

soupault commented Feb 22, 2016

JDWarner commented Feb 22, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

blink1073 commented Feb 22, 2016

JDWarner commented Feb 22, 2016

blink1073 commented Feb 22, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

soupault commented Feb 23, 2016

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ahojnnes commented Feb 26, 2016

soupault commented Feb 27, 2016

soupault commented Feb 29, 2016

Choose a reason for hiding this comment

stefanv commented Feb 29, 2016

soupault commented Feb 29, 2016

stefanv commented Feb 29, 2016

soupault commented Mar 15, 2016

stefanv commented Mar 15, 2016

soupault commented Mar 19, 2016

blink1073 commented Mar 19, 2016