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.synopsys.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]
For programming languages, how
For programming languages, how about tracking the number of files as well?
I like the idea. It's a feature that would take a day or two. I'll add it to our internal feature list.
In a similar train of thought, what do you think to tracking the number of comments per programming language?
Could be interesting to know that in a project's line count, 70% of the CSS code is comments but only 10% of the PHP is. Since ohloh works on basis of per file, could add blank lines to that too.
It occurs to be that your analyser must have those numbers, it's just a matter of storing them heh.
Interesting, Daniel. On a related note, I think I already suggested somewhere that the Ohloh parser may try to distinguish between structured comments (machine-readable documentation à la phpdoc) and raw comments.
Structured comments allow detailed code documentation to be generated from the source and are understood by some editors for edit support and autocompletion (e.g. Eclipse plugins). The percentage of structure comments (if any) per project would be a very interesting metric IMO.
That's be quite good... would be another project thumbs up badge for if it's code contained a reasonable amount of doxygen/similar commenting, though that might take a little longer to implement scratch head
Oh, another one! While we're on the subject of per language... wouldn't the ohloh parser have a reasonably easy way to spit out this info (comments, blanks, language composition, etc) on individual commits? Could be useful to be able to look at a graph showing how much a paticular language has been changed in the last X time frame... for, say, the html documentation or what ever?
As an extension to that... it would be interesting to be able to tell at a glance that 90% of the lines in the commits over the last month have consisted of adding blank lines and comments.
these are all awesome ideas. Grrr - why must a day only contain 24 hours? ;-).
dartar: Our code parser is decent, but a little primitive. I would love to upgrade to something more modern - and definitely make it open source. Do you know anyone with reasonable parsing/compiling skills who would be interested in spearheading this effort?
daniel: we have this data -- almost. For every file in every checkin, we know how many lines were added and deleted,by languange and by code/comment/blanks. The part that we don't know about is whether the lines deleted were previously your lines or someone else's. This means that we aren't able to determine, in a project, what percentage of the code is due to which developer. We know how many lines they added and deleted, but not how many they're responsible for. We'd need to run something like a 'blame' command and keep track, on a line-by-line basis, who editing who's code. I'd like to get there - its just something that keeps slipping under higher-priority issues. If someone could help us with the code parser - we might just get there sooner tho!
Do you know anyone with reasonable parsing/compiling skills who would be interested in spearheading this effort?
I understand the basic theory of language parsing/compiling, but I've never actually tried writing one heh. Probably the closest I've gotten is bbcode parsing and simplistic script interpreters heh. Then again, time is something I'm not exactly overflowing with :P yes, curse those 24 hour limitations.
Having said that, I'll have to keep an eye on the idea ;)
The part that we don't know about is whether the lines deleted were previously your lines or someone else's. This means that we aren't able to determine, in a project, what percentage of the code is due to which developer. We know how many lines they added and deleted, but not how many they're responsible for.
hmhmm *thinks* So you could spit out those numbers for the project (graph/pie chart/wherever), which would be rather interesting.
I hadn't thought of doing them for the user commit graph, though I can see how what I said leads to that idea. It's a little late so my memory isn't working well, are there any metrics available for contributor's LOC changed? I seem to remember everything on users talking about commit count rather than lines.
Do you know anyone with reasonable parsing/compiling skills who would be interested in spearheading this effort?
I'll spread the news, what language is your parser written in?
FYI the specifications for PhpDoc/Doxygen docblocks can be found here:
I guess you probably wouldn't need to parse the whole content of docblocks (although this might be a very interesting extension), you would just need to detect the beginning and the end of a docblock and count the # of lines it contains.
dartar: ruby and C.
Hmmm, think it'd be possible to spit out some sort of graphing based on the Language Summary bit of the commit page?
Or even expose it in the api so I can have a play with it? :P