www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Dmd pull requests coming in to fast?

reply "Benjamin Thaut" <code benjamin-thaut.de> writes:
At the time of this writing there are 120 open pull requests for 
dmd. Many of which don't have a single response. Are the pull 
requests coming in to fast to be handeled? Are there any plans to 
improve this situation?

Kind Regards
Benjamin Thaut
Feb 16 2013
next sibling parent Andrej Mitrovic <andrej.mitrovich gmail.com> writes:
On 2/16/13, Benjamin Thaut <code benjamin-thaut.de> wrote:

 Are there any plans to improve this situation?


More people reviewing -> greater chance of pull being merged and bugs
being found so a regression isn't introduced.
Feb 18 2013
prev sibling next sibling parent reply "Maxim Fomin" <maxim maxim-fomin.ru> writes:
On Saturday, 16 February 2013 at 15:42:54 UTC, Benjamin Thaut 
wrote:

 At the time of this writing there are 120 open pull requests 
 for dmd. Many of which don't have a single response.

 Kind Regards
 Benjamin Thaut


I would say that in fall 2011 the number was around 100 (as far I 
remember) and a year later it was around 110. It seems to be 
increasing. I consider this a revelation of a "too many problems 
too few people" problem.
Feb 18 2013
parent reply "Jonathan M Davis" <jmdavisProg gmx.com> writes:
On Monday, February 18, 2013 22:07:56 Maxim Fomin wrote:

 I would say that in fall 2011 the number was around 100 (as far I
 remember) and a year later it was around 110. It seems to be
 increasing. I consider this a revelation of a "too many problems
 too few people" problem.


With a difference of only ten, it could simply be that there were more in that 
month rather than it's been increasing overall. You need to look at the trend. 
That may show that we're falling behind, or it may we're not. I don't know. 
But even if the trend is that there are more pull requests, it could be that 
they're actually being processed faster, but it's just that there are more in 
the queue at a time, making it so that a pull request takes about as long to 
get processed now as it did a year ago. It very well be that there needs to be 
more people involved, but the fact that there are 10 more requests in the 
queue doesn't mean much.

- Jonathan M Davis
Feb 18 2013
parent "Maxim Fomin" <maxim maxim-fomin.ru> writes:
On Tuesday, 19 February 2013 at 00:29:46 UTC, Jonathan M Davis 
wrote:

 On Monday, February 18, 2013 22:07:56 Maxim Fomin wrote:

 I would say that in fall 2011 the number was around 100 (as 
 far I
 remember) and a year later it was around 110. It seems to be
 increasing. I consider this a revelation of a "too many 
 problems
 too few people" problem.


 With a difference of only ten, it could simply be that there 
 were more in that
 month rather than it's been increasing overall. You need to 
 look at the trend.
 That may show that we're falling behind, or it may we're not. I 
 don't know.
 But even if the trend is that there are more pull requests, it 
 could be that
 they're actually being processed faster, but it's just that 
 there are more in
 the queue at a time, making it so that a pull request takes 
 about as long to
 get processed now as it did a year ago. It very well be that 
 there needs to be
 more people involved, but the fact that there are 10 more 
 requests in the
 queue doesn't mean much.

 - Jonathan M Davis


Perhaps you are right. Auto-tester actually tracks some 
statistics 
http://d.puremagic.com/test-results/chart.ghtml?projectid=1. 
Judging by weekly pull creators - pull closures the difference is 
often positive, but small.
Feb 18 2013
prev sibling parent kenji hara <k.hara.pg gmail.com> writes:
2013/2/19 Andrej Mitrovic <andrej.mitrovich gmail.com>


 On 2/16/13, Benjamin Thaut <code benjamin-thaut.de> wrote:

 Are there any plans to improve this situation?


 More people reviewing -> greater chance of pull being merged and bugs
 being found so a regression isn't introduced.


I opened dmd/pull/773 at 2012/02/28
It is in page13 of my closed pull request lists.


https://github.com/D-Programming-Language/dmd/pulls/9rnsr?direction=desc&page=13&sort=created&state=closed
12(in p13) + (p1 to p12) * 25(per 1page) == 312 pulls closed in one year

Almost of my pull requests are for bug fix, so 'fixing bug speed' is
definitely not slow.

The reasons which those being kept opened are roughly divided into three.
1. Review is not still caught up.
    Most of my pulls, and some backend fix are belongs to here.
2. Already outdated.
    Need rebasing...
3. Because it is an enhancement.
    This is not less than you think.

Kenji Hara



 Feb 18 2013