Forums : Suggestions for Ohloh 2.0
Dear Open Hub Users,
We’re excited to announce that we will be moving the Open Hub Forum to
https://community.blackduck.com/s/black-duck-open-hub.
Beginning immediately, users can head over,
register,
get technical help and discuss issue pertinent to the Open Hub. Registered users can also subscribe to Open Hub announcements here.
On May 1, 2020, we will be freezing https://www.openhub.net/forums and users will not be able to create new discussions. If you have any questions and concerns, please email us at
[email protected]
Contributors in SVN that were
Contributors in SVN that were attributed in the comment due to lack of commit-permission.
Due to SVN's model, the 'committer' is the contributor, which may (and in some cases usually) is not the case, since they 'merely' incorporated the changes/patch that someone else did.
Hi harningt,
I understand this limitation - its unfortunate. We could either:
- try to automagically understand the commit comment and assign the commit to a different person
- allow people to claim individual commits
- any other suggestion?
I seem to recall there was a suggestion somewhere allowing the commit 'owner' on ohloh to assign individual commits to other users, and that sounds ideal for the situation of the projects I've been involved in, and perhaps less open to problems/abuse than allowing people to claim individual commits themselves.
hi Alyssa,
Now that you mention it, I also recall that suggestion. It does seem like the most equitable solution. Ok - we'll aim for that.
It should perhaps be possible to request attribution for a commit from a committer, to make it easy for the committer to find the right account if nothing else.
A repository property, for those SCM systems which support it, would also be a candidate. While there would be an uphill march to establish use of it, it is the same with any metadata, and would be offset by the committer having to come to Ohloh to assign the commit anyway. If Ohloh becomes trusted enough (or a good, multi-provider credential system like CardSpace takes off), it might even be able to write metadata using the committer's delegated credentials.
Hmm. What would be used as the value the property? I mean, if the contributer had an SVN account, wouldn't they have committed the code? Anyway, couldn't committers let the requests pile up for a while (a week, a month) before handling them? Wouldn't be such a hassle then...
It could be either way - do it now or later. I think doing it later is a hassle. It's probably because when I'm committing a contrib, I do it all at once so I don't have to remember anything later. A property wouldn't necessarily be the truth of the metadata, just a source. Since it is custom metadata, for use by Ohloh, it could be whatever. Ohloh could create an ontology of metadata sources so that tracking data could be enriched, and because they make the ontology, they can pick what it is. If people want to participate, just use properties (or even commit or source comments for that matter) which are formatted in a particular way. TortoiseSVN does this with the bugtraq:
property metadata convention. It works. From here, it's easy to see that a good mapping would improve the processing of code semantics a lot - people could tag code in comments, or use attributes or repository properties or even annotate build scripts, and all of that would allow a mapping of code meaning to emerge.
Hmm, looks like git could use this just as much as SVN... it's hard to figure out how to use git correctly :-(. For instance, see the GIT commitdiff linked below (though I did mess up sending the patch, and it was my first...):
Re the GIT comment.... git does have an 'author' of a commit built in. So in GIT there's the 'committer' and (optionally different) 'author'.
For a solution... offering a method of claiming ownership of a commit would be nice... if the comitter was a member of Ohloh, then a notification of sorts would be in order.
Also... if a project enabled some option for the commit, a regex could be used on commit messages, perhaps w/ a list of 'common' regexes and a field for custom...
Most gnome project and definitely GStreamer uses one ChangeLog entry per commit and in that ChangeLog one can find Patch by:
to show that this has been contributed by someone else that the committer.
I think having to manually attribute commits is a bit painful. Who is going to bother? Then again, I understand the difficulty of trying to do it automatically. I'm not sure about this one. Perhaps ohloh could define a standard format that people can use to do the attribution in the commit message?
Wells,
It seems to me the best approach would be to let users enter it manually first (and through our API). Then, after, we could write some automation tools to do this automatically whilst parsing repository logs.
If the project containts GNU-formatted ChangeLog files, those are the place to look for who contributed which change...
Excellent point: changelogs, if present, generally list the actual author rather than the committer.
Have anything changed since this discussion? I want to claim a patch, but it seems there is no UI for that.
Hi Andrey,
Sorry, no, this feature has been perpetually pushed down in the stack by other demands. To be honest I don't think this will happen anytime soon.
Hmm, just found this thread, after posting on another thread about the same thing. +1 vote for the ability to manually reassign/claim ownership of individual commits... manual work is never pretty but at least it'll help folks some method for setting the record straight if they care enough to make the effort. Plus the manual reassignment functionality must be much easier to implement than automagical log-parsing reassignment.
