[nanomsg] Re: mangos - code integration rules?

  • From: George Lambert <marchon@xxxxxxxxx>
  • To: nanomsg@xxxxxxxxxxxxx
  • Date: Sat, 7 Feb 2015 01:15:37 -0500

Garrett - I think that your question is one of the most important and
opinionated in software projects today - and while there are multiple
schools of thought, I will share with you my very opinionated answer,
which is not necessarily correct - except for the fact this it is for
me, and I hope it sheds some perspective on your question.

I have recently become convinced that if you are using great version
control systems and continuous integration - There is a solid case for

******* pushing releases out very quickly, when all of your test cases
pass *******

I read a line over at io.js the other day that stuck with me.

"Releasing more frequently leads to a more stable product, not a less
stable product."

I had to reflect on that idea for a little bit, but the point rings
true the more that I think about it.

I saw this article a while ago - and thought it was very insightful

http://berk.es/2012/08/03/git-deploy-or-how-i-learned-to-stop-worrying-and-love-deployment/

If you are exceptionally concerned about your user base - you could
use two or three major tags and let your users decide for themselves.

Suggest that they all use Testing, and submit both bugs and test cases
for those bugs.  It will make your system more stable very quickly.

Imagine - letting them decide between

Stable:
Tested:
Daily:

I just had an experience which is best visualized with this github
chain - in an exchange with another open source project that had a bug
creep up.

Their initial response was: we have never heard of that before, and
are unable to replicate it on our machines.  4 days later, test case
provided,
bug identified, test case refined, blame assigned, and the iteration
loop in 4 days is that a previously unnoticed problem has been closed.

https://github.com/rethinkdb/rethinkdb/issues/3712#issuecomment-73351517



G.

On Sat, Feb 7, 2015 at 12:40 AM, Garrett D'Amore <garrett@xxxxxxxxxx> wrote:
> (If you don’t use mangos, ignore this message, unless you are a Go 
> programmer, in which case why aren’t you using mangos?)
>
> Up until this point, I’ve been basically integrating changes myself, with no 
> attempt at cross review (at least not much), although I *am* trying to use 
> things like Travis CI to apply professional practices to minimize risk and 
> improve quality.  But to date I have been just merging changes using a policy 
> sort of like Linus and the Linux kernel — if I think its right, then I push.
>
> Without abdicating control over mangos, I’m interested to know if people are 
> desirous enough to request that I slow down my rate of change, in order to 
> give other people a chance to review and comment (and even perhaps object to) 
> changes I make.  Admittedly with github anyone can fork if they really don’t 
> like where I’m going.  The real goals here are quality improvement and risk 
> mitigation, and I *do* believe in constructive code review as part of quality 
> improvement.
>
> I’m not sure if mangos has a big enough user base to warrant these concerns 
> yet or support a code review process/step, but if the github stars are any 
> sign, it suggests there *might* be.
>
> If you care about this sort of thing, and especially if you’re willing to 
> help cross review merge requests as part of the process, please let me know. 
> If nobody responds (or there is insufficient response to support a more 
> formal code review effort), then I’ll just keep on the course I’ve been on 
> til now.
>
> Thanks.
>
>         - Garrett
>
>



-- 
P THINK BEFORE PRINTING: is it really necessary?

This e-mail and its attachments are confidential and solely for the
intended addressee(s). Do not share or use them without approval. If
received in error, contact the sender
and delete them.

Other related posts: