Forums : Feedback Forum
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]
Developers commiting to repository for others.
My project Dawn of Light have very few people with SVN commit access, code is suggested or submitted by anyone willing to contribute to these people directly, this is needed for quality control, yet under the current anlysis code I presume they will be recognised as coming from the same developer. Let's look for example at a log file
We use the structure *) Code from xxx
then the list of changes. This I imagine is very hard to handle individually, but could be handled with another parsing layer through each of the log descriptions instead of just the author, with predefined project structure
definitions comparing the log description with the predefined structure to count the developers. Unfortunately this as with everything would be very open to abuse.
Other issues I have encountered is that when trying to edit the tags for the project, I want to add C#, I type C, it gives me a recommendation for C#, whether I press it, or just type it myself and press add, it comes up with the message: Name may include only letters, numbers, and the special characters + - _ .
Hi Etaew,
Your problem is more common than you might think. I have seen several other projects which attribute code to other authors in their comments.
We are currently trying to solve the general problem of a person who used two or more different names within a single project, or a person who worked on several separate projects. We'd like to be able to attribute all of this work to a single real person. By putting a mapping layer between commit logs and actual people, we open up the possibility do what you are proposing, which is to extract an author name from the commit comments based on some kind of pattern recognition. It's relatively simple in theory but the trick is achieving this at massive scales. I must admit this is probably going to take a while to achieve.
I reviewed the Subversion log for Dawn of Light, and I wanted to point out a complicating factor: It seems that a single commit often includes attributions to several different developers. This means that it will be impossible from the source control to figure out who wrote what code -- we can only pull out a single diff from Subversion, but with multiple names attached to that diff we don't know where to attribute the changes. To implement your feature request, at a minimum, you would need to be sure to attribute all code in a commit to a single person.
If you really like the model of random patches contributed to official reviewers who do the actual commit, I'd recommend you take a look at Git. Git is used by the Linux kernel team to support exactly this development process. When people around the world send you a patch for your project, as an administrator you can review or amend that patch, but Git will retain the original authorship information in the log.
Finally, regarding the validation problems you are having with tags, I'll take a look at this right now. It sounds like a very simple oversight in some recently added code.
I really appreciate the feedback. Thanks for coming to Ohloh!
Robin
