Re: Thunderbird’s future: a modern addressbook
eyalroz at technion.ac.il
Sat Mar 17 18:30:35 UTC 2018
On 3/17/18 6:44 PM, ace wrote:
>> Final note - can you elaborate a bit (either on- or off-list) about the
>> "processes" you mentioned an extension author gets into if his/her
>> extension gets integrated into the core code?
> I meant concepts that may pose changes to his current workflow like these:
> 1. there will be reviews of the code he intends to check in.
Code reviews are a _gift_. Any developer worth his salt should love
having experienced people looking over the code you write, helping you
find bugs or potential for improvement.
> 2. there are some coding style rules in the code that need to be
> followed, so that others can also understand the code easily.
I guess having to give up your favorite brace style could be a bit
annoying; but it's not much trouble otherwise.
> 3. if the code is integrated, the nightly users always get the newest
> code. If there is some breakage/API change from m-c, he may be asked to
> fix it ASAP instead of working on some new feature.
Yeah, so this is where the pain is at. As an extension author you only
get this with new releases; having to attend to nightlies would be more
effort. On the other hand, _other_ people would have to try and make
sure they don't break the new address book.
Also, this will becomes harder as the core code changes/disintegrates
due to what's happening with FF.
> 4. similarly there are the release cycles and there will be periods
> where it may be preferred to work on polishing and stabilisation bugs,
> not new features. And again on fixed deadlines.
Couldn't one just work on new features in a separate branch?
More information about the tb-planning