Fixing bugs

[george_bbch]

Hi all,

I think that one of the top priorities at the moment is fixing the bugs in the 4.5 release. This is not an easy job, as there are a plenty of them.
I have some thoughts / suggestions for Sugar team:
- improve the bugs site: the access speed, search capabilities, an export feature
- clean the bug list of duplicates or similar bugs, update the status, resolution and fixed priority information (e.g. there are bugs marked as fixed in 4.2.1c, but they're actually still open, etc.)
- ask for community's feedback about the importance of the bugs in a direct, quick and easy manageable way and use this to internally adjust priorities for fixing them
- when a savvy programmer from the community suggests a workaround, don't hesitate to test it internally and add it to the upcoming patch; think about a way of encouraging and rewarding such efforts
- plan and release a patch every week until all the major bugs have been fixed
- think about a way to detect and preserve at patch installation third party code customizations, whenever it's possible
- make / keep patches cumulative, so that patch C includes all bug fixes in patch B, which includes all fixes in patch A, etc.: users who can afford to wait will have less work and problems for patching their systems
- keep us informed about future patches, their expected release dates and any changes related to them

Thanks,
George

[WayneSugar]

- when a savvy programmer from the community suggests a workaround, don't hesitate to test it internally and add it to the upcoming patch; think about a way of encouraging and rewarding such efforts

As a developer we do. The best way to get 'code' to us is to attach it to a bug

- plan and release a patch every week until all the major bugs have been fixed

Every week is a torrid schedule. We aim for quality over quantity of patches. I'm sure most sysadmins appreciate this.

- make / keep patches cumulative, so that patch C includes all bug fixes in patch B, which includes all fixes in patch A, etc.: users who can afford to wait will have less work and problems for patching their systems

We do.

[alsutton]

Every week is a torrid schedule. We aim for quality over quantity of patches. I'm sure most sysadmins appreciate this.

Wayne,

From where i'm sitting your failing on aiming for quality over quantity. Major releases are made that still have major bugs in from previous releases, so you can't expect us to take the quality argument seriously.

A weekly schedule isn't unrealistic if you reduce the amount of work you take on. You could do the following which would give you small, bite sized chunks to fix and provide your customers with the fixes they want;

1) Add voting to the bug tracker, that way you can see which bugs are important by the number of votes it has. At the moment the only way of showing the importance of a bug is to register duplicates of it which wastes everyones time.

2) Commit to fixing the top bug (and possibly more of the top 10) in each patch. If the top bug takes more than a week to work on then people will wait because what we want and need is being worked on (unlike at the moment where the biggest thread on the forum covers email bugs which have been sitting around for months).

I work in software development and have for over a decade, so please don't try to do a "Ohhh that won't work" reply, I've been there, seen it, done it, and I know it does.

Al.