From: Phillip Wood <phillip.wood123@gmail.com>
Date: Mon, 10 Jun 2019 10:40:30 GMT
Subject: Re: [GSoC][PATCH 3/3] cherry-pick/revert: update hints
Message-ID: <72b44513-1182-8172-94aa-ca69730d9a22@gmail.com>
In-Reply-To: <20190610052823.17013-1-rohit.ashiwal265@gmail.com>

Hi Rohit

On 10/06/2019 06:28, Rohit Ashiwal wrote:
> Hey Phillip
> 
> On Sun, 9 Jun 2019 19:03:02 +0100 Phillip Wood <phillip.wood123@gmail.com> wrote:
>>
>> Hi Rohit
>>
>> On 08/06/2019 20:19, Rohit Ashiwal wrote:
>>> [...]
>>> @@ -2654,8 +2654,8 @@ static int create_seq_dir(void)
>>>  {
>>>  	if (file_exists(git_path_seq_dir())) {
>>>  		error(_("a cherry-pick or revert is already in progress"));
>>> -		advise(_("try \"git cherry-pick (--continue | --quit | --abort)\""));
>>> -		advise(_("or  \"git revert (--continue | --quit | --abort)\""));
>>> +		advise(_("try \"git cherry-pick (--continue | --skip | --quit | --abort)\""));
>>> +		advise(_("or  \"git revert (--continue | --skip | --quit | --abort)\""));
>>
>> If the user has already committed the conflict resolution then we don't
>> want to recommend --skip as there is nothing to skip.
> I think it is more about suggesting what are all the possibilities
> you can try and not about intelligently suggesting what you should
> do.

Previously all the suggested options were viable, --skip is not
applicable if the user has committed a conflict resolution. The idea of
the advice is to help the user, suggesting options that wont work is not
going to help them.

> ofc, we can not use `revert --<option>` while cherry-picking.(

As I suggested in patch 1 we should tailor the error message to the command.

> we should not be able to do so in ideal conditions, but the world
> does not work as we think it should). Still we are suggesting so
> here.

Yes because you have the power to easily make that change. It is normal
to try and improve the code base when we make related changes.

> Also, I think it is more reasonable to make "this" a part of patch
> which will cover "tailored" advice messages which is also a topic
> of discussion as I described here[1].

That might make sense, but it is a pretty self contained change as part
of this patch.

Best Wishes

Phillip

> 
>> Best Wishes
>>
>> Phillip
> 
> Thanks
> Rohit
> 
> [1]: https://public-inbox.org/git/20190609200038.GD28007@hank.intra.tgummerer.com/T/#mbb071f6e29c69f291ecd9c61c71b889774ff33b2
>