2017Q1 release #237

salkinium · 2017-04-02T23:33:54Z

Updates the Readme, Author files and adds a Changelog.

We're doing quartely releases now. Cos we're awesome.
This closes #160.

Please review changelog format and comment on release cycle.
I'm kind of hoping that GitHub renderes the PR numbers and commit SHA1 as a link as it does in the comments, but maybe that doesn't work.

cc @strongly-typed @Sh4rK @dergraaf @cajt @ekiwi

ekiwi · 2017-04-03T00:03:56Z

love the change log! cool!

dergraaf · 2017-04-03T08:23:37Z

👍
I really like the changelog! Writing those is always a lot of work, but worth it for everybody else. I guess it will only be possible to keep it to that level in the long run if we can extract them directly from the PR. Is there any support from GitHub for that?

Quarterly releases are fine, although releases every 6 month would also be sufficient.

What I don't like is the new authors file. With the number of commits it feels like a competition which it shouldn't be. Why not just list all authors independent of their number of commits? If you really need that information you can extract it quite easily as shown by your script.

salkinium · 2017-04-03T12:31:39Z

Is there any support from GitHub for that?

Not that I'm aware of. Maybe using the Milestones can be useful, don't know.

We are getting better with the PRs though, way more consistent, and usually have a summary in the merge commit anyways. So it's not too much work.

Quarterly releases are fine, although releases every 6 month would also be sufficient.

I'm gonna try quarterly releases for now and and see how that works. I think it fits nicely with the arm-none-eabi-gcc release schedule.

What I don't like is the Authors file.

Ok, I'll just make it an alphabetical list. I didn't think much about it, I basically just used what came out of shortlog. But I'm clearly the king and other things egomaniacs say about themselves. ;-P

chris-durand · 2017-04-03T15:11:16Z

Is there any support from GitHub for that?

There is a GitHub API. It's a bit slow but easy to use (with pygithub).

This prints a list of merged pull requests:

from github import Github
g = Github()

xpcc = g.get_repo('roboterclubaachen/xpcc')
pull_requests = xpcc.get_pulls(state='closed', sort='created')

for p in pull_requests:
    if p.merged:
        date_str = p.merged_at.strftime('%Y-%m-%d')
        print('### %s: %s' % (date_str, p.title))

Sh4rK

It's nice to see a release, there are a few interesting changes. I might actually update xpcc in the project I worked on a while ago :)

I added a few comments, sorry for being a bit too perfectionist, feel free to ignore them :D

Sh4rK · 2017-04-03T14:49:46Z

CHANGELOG.md

+chosen for their importance to xpcc users as judged by the maintainer.
+This means that not every small commit makes it in here.
+
+There are quarterly release summaries formattes as:


Typo: formattes -> formatted

Sh4rK · 2017-04-03T14:50:35Z

CHANGELOG.md

+
+## 2017-04-04: 2017q1 release
+
+As this is our first offical release it covers the last 12 months of development.


Typo: offical -> official

Sh4rK · 2017-04-03T14:52:07Z

CHANGELOG.md

+The `xpcc/architecture/util.hpp` macros have been renamed and
+properly documented.
+A new header `xpcc/architecture/legacy_macros.hpp` contains backwards
+compatible mappings for application code. All occurances of the non-


Typo: occurances -> occurrences

Sh4rK · 2017-04-03T14:53:22Z

CHANGELOG.md

+
+Three heap algorithms can be chosen from using the xpcc parameters:
+
+- newlib's dlmalloc (default): Chooses largest continous heap from page table.


Typo: continous -> continuous

Sh4rK · 2017-04-03T15:10:11Z

tools/authors.py

+# get the shortlog summary
+output = subprocess.Popen("git shortlog -sn", shell=True, stdout=subprocess.PIPE).stdout.read()
+# parse the shortlog
+for line in output.split("\n")[:-1]:


Maybe output.splitlines() instead of output.split("\n")[:-1]?

Sh4rK · 2017-04-03T15:10:41Z

tools/authors.py

+}
+
+# cleaned up dictionary
+commit_count = {}


You could use commit_count = defaultdict(int) here, remove the add_commits function below and just use commit_count[author] += shortlog[slauthor] to update the counts.

Sh4rK · 2017-04-03T15:21:46Z

tools/authors.py

+# merge authors
+for slauthor in shortlog.keys():
+    for author, aliases in author_alias.items():
+        if any(name in slauthor for name in (aliases + [author])):


What's the reason for checking for substrings instead of equality here? If equality checking works, this can be simplified to if slauthor in aliases:. No need to add author to the list, that will be handled by the else clause.

Substring allows me to be lazy and only specify part of the alias name. Adding [author] here so I don't need to specify it again as an alias. Coz I'm a lazy 🐒.

That doesn't seem to be too robust. What if someone else named Kevin (or any of the aliases) contributes? That would match the alias incorrectly.
As for [author]: you wouldn't need to add the "canonical" name to the aliases, since if it is not found in the alias list, the for loop would fall over to the else branch, and add the author there (as for every non-aliased author).

Good point about [author], I overread your first comment too :-(
Regarding aliases, I hope to control the aliases better in the future, a lot of them come from before xpcc had me as maintainer. There are still author emails if all else fails.

Okay, in that case go ahead with the release, I don't intend to stall this further.

Thanks for the review. I don't perceive this as stalling at all, I enjoy everyone reviewing my code ;-) I still have to update the changelog anyways…

salkinium · 2017-04-04T23:22:16Z

Thanks to all contributors for this release, especially the 7 first-timers!

Antal Szabó (@Sh4rK)
Christopher Durand (@chris-durand) 🎉🎊
Daniel Krebs (@daniel-k)
Fabian Greif (@dergraaf)
Georgi Grinshpun (@georgi-g)
Julia Gutheil 🎉🎊
Kevin Läufer (@ekiwi)
Michael Thies (@mhthies) 🎉🎊
Nick Sarten (@genbattle) 🎉🎊
Niklas Hauser (@salkinium)
Raphael Lehmann (@rleh) 🎉🎊
Sascha Schade (@strongly-typed)
Tarik TIRE (@7Kronos) 🎉🎊
Tomasz Chyrowicz (@tomchy) 🎉🎊

salkinium self-assigned this Apr 2, 2017

Sh4rK reviewed Apr 3, 2017

View reviewed changes

Update for 2017q1 release.

b505cbb

salkinium merged commit b505cbb into roboterclubaachen:develop Apr 4, 2017

salkinium deleted the feature/changelog branch April 4, 2017 23:50

strongly-typed mentioned this pull request May 10, 2017

Breathe (more) life into our release cycle #250

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

2017Q1 release #237

2017Q1 release #237

salkinium commented Apr 2, 2017 •

edited

Loading

ekiwi commented Apr 3, 2017

dergraaf commented Apr 3, 2017 •

edited

Loading

salkinium commented Apr 3, 2017

chris-durand commented Apr 3, 2017

Sh4rK left a comment

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

Sh4rK Apr 3, 2017

salkinium Apr 3, 2017

Sh4rK Apr 3, 2017

salkinium Apr 3, 2017

Sh4rK Apr 4, 2017

salkinium Apr 4, 2017

salkinium commented Apr 4, 2017


		## 2017-04-04: 2017q1 release

		As this is our first offical release it covers the last 12 months of development.


		Three heap algorithms can be chosen from using the xpcc parameters:

		- newlib's dlmalloc (default): Chooses largest continous heap from page table.

2017Q1 release #237

2017Q1 release #237

Conversation

salkinium commented Apr 2, 2017 • edited Loading

ekiwi commented Apr 3, 2017

dergraaf commented Apr 3, 2017 • edited Loading

salkinium commented Apr 3, 2017

chris-durand commented Apr 3, 2017

Sh4rK left a 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

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

salkinium commented Apr 4, 2017

salkinium commented Apr 2, 2017 •

edited

Loading

dergraaf commented Apr 3, 2017 •

edited

Loading