[PATCH 2/2] dns: Support larger queries when asking forwarder

[Michael Adam]

Hi Matthieu,

On 2013-05-29 at 00:47 -0700, Matthieu Patou wrote:
> On 05/29/2013 12:30 AM, Michael Adam wrote:
> >
> >you may have missed that these patches have already gone to
> >master. (See the discussion on the following set of 2 patches
> >to add RP and HINFO records.
> 
> Well I hate the way we do review because there is no way to see if
> someone has already done the review or a way for the one waiting for
> the review to see that someone dare to pay attention to his patch.

Well, since doing proper review usally involves taking the proposed 
patch and applying it to current branch (master), you would have
noticed in this case since the patch was already in...


You are of course right in that the system is not perfect,
but how can we improve it? This is an important topic.
Here are some thoughts of mine about the reviewing workflow:


Of course there is a way to notify others of review in action,
review done, etc: By sending email. But as with any other
technical means, this depends on the will of the developers to
participate. (And how much mail is reasonable?)

As an example of a different mechanism, in the beginning, Simo
has set up this patchwork framework. I also used this initially,
but since most of the people did not participate, I quit doing it.

So even with a better means than mail, you will not get 100%
coverage. Isn't your complaint rather about the way people are
acting with reviews than about our technical enforcement, right?

Here is what we can/could do:
- we could enforce review by blocking patches from landing in
  master without sufficiently many team signoffs or reviews.
- we can ask people to communicate/use tools better

Here is what we can not do (at least I don't know how):
- We can not force people to use indicators that they are doing
  review.
- We can not prevent patches from landing in master without the
  pusher having previously indicated that she is doing review.
- We can not prevent anyone from doing review while someone else
  is already reviewing.

So you can propose a better reviewing process.  I would be really
interested, but I doubt you will come up with something waterproof.

Cheers - Michael