[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [ecf-dev] JavaDoc (or lack thereof)

I think I've made it incrementally better.

I like your suggestion about the request, but am always hesitant to ask
committers to provide anything more than the bare minimum. I'll have to
stew over that for a while.

I think we should move this off the ECF mailing list. Perhaps I'll open
a bug.

Wayne

On 05/27/2011 01:48 PM, Alex Blewitt wrote:
> On 27 May 2011, at 18:35, Wayne Beaton wrote:
>
>> Actually, the way that you're describing it is how I understand it. 
>>
>> I guess that the wording on the page may be confusing? I think that "Also set the author name and email to the contributor's information." is pretty clear. Am I wrong?
> It's not tremendously clear. An example would help. You bounce back and forth in that statement between the committer and contributor, and it's not always clear.
>
> For example, when you pull a change from someone else, the committer information is already set. You don't need to change anything. 
>
>> My concern about signed-off-by is that it sounds like it may require some extra work from committers (until we get Gerrit running, I guess). Traditionally, getting committers to do extra work tends to have inconsistent results.
> I think you can use EGit to add the signed-off-by for the commit amend without problems.
>
> It's also  unclear what this means:
>
> "Project committer adds a comment that includes URL pointers to the refs in the eclipse.org repository
> 	• e.g., http://git.eclipse.org/c/ecf/org.eclipse.ecf.git/commit/?id=0e184beca9dacdc26ec009dbc54be108a5c5723e";
>
> You can't have a commit message pointing to itself, because the commit ID is both a hash of the message and not known until after commit. Or do you mean a comment in bugzilla? That would be clearer.
>
> So, to summarise - I think it would be better to represent it as a set of commit messages. For example:
>
> Stephan wants to commit the following change to EGit:
>
> commit d2fea522c1cf2e5f09542b3d2c3624c56aeffcdc
> Author: Stefan Lay <stefan.lay@xxxxxxx>
> Date:   Tue Mar 8 14:31:02 2011 +0100
>
>     Fix Build after JGit change 9013e9e99
>     
>     Change-Id: I48b61df5575a9351c415bc6f9860d9e12eae233d
>
> You create a bug, pointing to http://.../d2fea52c...cdc
>
> Shawn wants to pull this change into EGit. Email address 'stefan.lay@xxxxxxx' is not a committer on the EGit project. So, he adds a signed-off-by with his e-mail address to acknowledge that the change is OK:
>
> commit fda2c6f3916822f8476a2a4bc0a26d512cf0a1b1
> Author: Stefan Lay <stefan.lay@xxxxxxx>
> Date:   Tue Mar 8 14:31:02 2011 +0100
>
>     Fix Build after JGit change 9013e9e99
>     
>     Change-Id: I48b61df5575a9351c415bc6f9860d9e12eae233d
>     Signed-off-by: Shawn Pearce <spearce@xxxxxxxxxxx>
>
> Although stefan.lay@xxxxxxx isn't a committer on EGit project, spearce@xxxxxxxxxxx is, so this change can be pushed as-is to the EGit repository.
>
> That's the way I think it should work.
>
> Alex
> _______________________________________________
> ecf-dev mailing list
> ecf-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/ecf-dev