Integrating Kallithea with TortoiseHg (plus some other comments/suggestions from my team)

[Thomas De Schampheleire]

On Tue, Feb 16, 2016 at 8:36 AM, Angel Ezquerra
<angel.ezquerra at gmail.com> wrote:
> On Mon, Feb 15, 2016 at 1:42 PM, Thomas De Schampheleire
> <patrickdepinguin at gmail.com> wrote:
>> On Mon, Feb 15, 2016 at 12:04 AM, Mads Kiilerich <mads at kiilerich.com> wrote:
[..]
>>
>> I find the request to skip directly to the approve button very
>> strange, to be honest. Code review is not just about
>> approving/rejecting, but also on giving code comments (which are most
>> useful when placed inline in the code, really).
>>
>> Why would you want to skip to that button and give the impression that
>> you cannot/should not comment on the code?
>
>
> Both you and Mads make similar comments, so I´ll answer here to both of you :-)
>
> It does not make sense, as you say, to want to directly approve or
> reject a revision without reviewing it first.
> However it does make sense to review the revision _outside_ Kallithea
> (e.g. using TortoiseHg and your favorite editor) and then want to just
> mark the revision as Approved or Rejected in Kallithea.
> That is basically the reason for the request. People are used to their
> tools, and they'd want to use them to review the code (in addition to
> being able to do it offline). Ideally they'd like to be able to right
> click on the revision on TortoiseHg and select "Approved" on some sort
> of Kallithea or Code Review menu. Since that is not possible the
> second best solution is to just open the page and jump straight to the
> main form.
>

But 'doing the review in TortoiseHg' means that you cannot make
comments there, correct? (you'd have to write a TortoiseHg plugin that
allows commenting from the UI and send the requests to Kallithea).

So this means that either users will not write any comments and just
respond with approved / rejected; or they are 'forced' to accumulate
all comments in the main comment form without automatic line
references.
I think it would be better to try and convince your users about the
benefits of real inline code review, than to 'help' them stick too
non-ideal processes.

Anyway, I am not here to force an approach to others, this is just my
suggestion. It's probably possible to resolve your request with some
minimal HTML addition, so I suggest you propose a patch and we can
then discuss on that technically.

/Thomas