BugzID references

[Alexander Shorin <kx...@gmail.com>]

Hi devs,

Question about importing Cloudant patches. Some of them has BugzID
references and short commit message which explains the change, but
doesn't explain what was wrong there. Is there public access to the
bug tracker to read about? Or if it's not, is it possible to export
the "contributed BugzID fixes" to somewhere for examination?

For instance, I wonder the reasons about 28257. And few others I
already lost in history.

--
,,,^..^,,,

Re: BugzID references

[Alexander Shorin <kx...@gmail.com>]

On Fri, Aug 1, 2014 at 4:44 PM, Robert Samuel Newson <rn...@apache.org> wrote:
> the change to the default values for max_write_delay in ticket 28257 was because of observations in production that the burden of updating the backing store that powers the global changes feed began to dominate. We modified that at runtime to a larger interval and observed a significant improvement, and thus we changed the default. This information should have been included in the commit message.

Thanks for clarification, this makes a sense.

> We won’t be opening our FogBugz tickets to the general public. I do want to address the issue you raise, however. Hopefully there are not too many commit messages that violate Cloudant’s internal standards (which I have pasted to the dev@ list in the recent past). Can we take this on a case by case basis? While we won’t provide open access to our bug tracking, we’ve no problem at all with extracting descriptions of any changes we’ve contributed to the couchdb project.

Sure, that's fine for me. Thanks!

--
,,,^..^,,,

Re: BugzID references

[Robert Samuel Newson <rn...@apache.org>]

Hi,

In general the commit messages should be descriptive enough, the bug number is for cross-referencing purposes and helps us coordinate release management and release notes.

the change to the default values for max_write_delay in ticket 28257 was because of observations in production that the burden of updating the backing store that powers the global changes feed began to dominate. We modified that at runtime to a larger interval and observed a significant improvement, and thus we changed the default. This information should have been included in the commit message.

We won’t be opening our FogBugz tickets to the general public. I do want to address the issue you raise, however. Hopefully there are not too many commit messages that violate Cloudant’s internal standards (which I have pasted to the dev@ list in the recent past). Can we take this on a case by case basis? While we won’t provide open access to our bug tracking, we’ve no problem at all with extracting descriptions of any changes we’ve contributed to the couchdb project.

B.

On 1 Aug 2014, at 13:18, Alexander Shorin <kx...@gmail.com> wrote:

> Hi devs,
> 
> Question about importing Cloudant patches. Some of them has BugzID
> references and short commit message which explains the change, but
> doesn't explain what was wrong there. Is there public access to the
> bug tracker to read about? Or if it's not, is it possible to export
> the "contributed BugzID fixes" to somewhere for examination?
> 
> For instance, I wonder the reasons about 28257. And few others I
> already lost in history.
> 
> --
> ,,,^..^,,,