[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [xen 4.6 retrospective] Possible solution together with the comments will be helpful

> On 31 Aug 2015, at 09:33, Jan Beulich <jbeulich@xxxxxxxx> wrote:
>>>> On 31.08.15 at 10:24, <feng.wu@xxxxxxxxx> wrote:
>> = Issue / Observation =
>> Sometimes the review comments are quite open, it doesn't contain a possible 
>> solution or a clear direction,
>> so it is not clear for the contributor on how to effectively address them. 
>> At least, in Linux kernel and KVM side, if the maintainers have
>> objection to the implementation of the patches, they will give a possible 
>> solution or a direction which is very
>> helpful for the contributor to address the comments. Hence this will make 
>> the review discussion more effective and productive and save both reviewer 
>> and developer's time.

I am wondering how big the difference between linux/kvm and xen is. And if 
there is a big difference (taking Jan's comment below into account), what the 
root cause is.

>> = Possible Solution / Improvement =
>> Try to give some possible solutions with the comments, especially for some 
>> big changes which affect a lot
>> to the whole patch-set.
> I think when a solution can be thought of in the context of reviewing,
> it is being given. I believe I know which case you allude to here, and
> I'm afraid it's not always reasonable for the reviewer(s) to do the
> contributor's work of finding a solution when none is obvious.

Again, maybe in this case, it would be helpful for the reviewer to call out the 
fact that there is no obvious solution for a problem he/she has raised. In that 
case, it gives the co-maintener (of which we have more now at least in core 
areas) an opportunity to step up. 

There is of course also a learning angle: to some degree, educating and helping 
new contributors, should eventually make them better, which in the long term 
leads to more efficient reviews. So to some degree, the upfront investment by a 
reviewer in educating a new contributor, will lead to less work for the 
reviewer in the long run. 

Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.