Privacy and KudoRank

Is a benchmark good, if it only includes programs that explicitly applied to be included?

When you talk about comparisons, please fix the activity measurings first. It hurts projects, and therefore potentially people's businesses, if complete nonsense like "decreasing development activity" etc is in the summary, just because a project develops primarily in branches or whatever. That's far more important.

Tobias: A really strange point of view, to compare humans with programs.

From a marketing point of view assigning Kudos to non-registered developers makes sense: Seeing the ranks you assigned to my individual contributions (without asking me!), made me curious what's my real Kudo rank.

Still it's my opinion, that Ohloh would be a much better service, if stats (KudoRank, number of commits, ...) would be an Opt-In service: You want it? You get it. You don't want it? Also ok, no reason to argue. I absolutly do not see, what's the benefit of publishing contribution stats for people, who don't want to be that transparent.

Hi Mathias!

Well, I see your point. But then you need to argue the same way with Googles page-ranke, for example. Most people don't opt-in for being ranked there. Nevertheless Google indexes their page and publishes the page rank.

I have to admit that it does not really matter to me, if Ohloh keeps up kudo-ranks for non-registered users. I just posted my first thought on this issue. :)

From the marketing point of view, Ohloh can also state "Wanna know your kudo-rank? Just register and see it!", if they stop publishing it for non-registereds.

Regards, Toby

Tobias: I don't think you can compare Google's page-rank and Ohloh's kudo-rank: When creating web-sites, indexable by Google you do this for publicity. You want your page to be found, otherwise you would not link it, or you'd put it behind some wall. So page-rank, as an essential feature of Google's search algorithm - helps you.

For free software contributors situations is different. Publicity not necessarily is a motivation for contributing to free software. I contribute for example, because I enjoy the chance of improving my daily tools, because I enjoy working with people and code. Of course insist in giving and taking, that's why the free software idea, especially the GPL family of licenses is important to me.

Well, but publicity really just is a side effect of all this work. I'd use a pseudonym, if I'd start from scratch.

If you create or contribute to an open source project you also do this in and for the public and therefore it gains you publicity. Also if it is just a side-effect, you still provide publically available information. Google processes some of the information you put into public and Ohloh does so, too. And of course there are a lot more institutions out there which take information that is publically available on the web and process it in much worse ways than Google or even Ohloh does.

So, as said above, it does not really affect me, if Ohloh keeps it like it is or changes it. But I think it does not really affect anyone to have this information on the web. From a pure privacy point of view I agree with you, though.

Hello Mathias,

as soon as you contribute to an open-source project, you do sth. public: Your codes are published along with your nick (identity in versioning like SVN, CVS, etc.). Additionally, most projects require you to add yourself to the list of authors (usually at the top of the modified source file).

Ohloh just indexes public information which can be obtained by everyone without Ohloh as well - the sources with their versioning infos simply are public in all open-source projects.

Btw. most developers use pseudonyms anyway. I don't know one single developer who uses his full name as nick (are you the first?) - if no pseudonym then at least an abbreviation or simply the first name is used.

If Ohloh would give you the possibility to hide your data, it would still be publicly available as long as the source repositories of the projects are online (or mirrored).

Best regards, Marco :-)

Mathias, you're already 'transparent'! As GNOME developer your name is reported somewhere for sure, just think of public repositories.

The only thing Ohloh does in addition is grouping your contributions and assign a number (kudo rank) that everybody with a little portion of brain working immediately identifies as a sham stuff to concretise what cannot be evaluated in a different way.

Also, it doesn't make any sense saying you've not being asked for being ranked:

1. you decided yourself to register at Ohloh and claim your work (you can undo it, if you wish),
2. your name is not linked in any way to your real life, EVERYBODY is allowed to use the alias 'mathiashasselmann' and, unless you EXPLICITLY claim your commits, Ohloh does not assign you ('you' real person) any judgement.

Giulio: I complained first, but after discussing the issue and reading comments I decided it's better for me to to register at Ohloh.

Registering at services like Ohloh absolutly doesn't mean, that you agree with the existance of this services, or that you like them. Lame comparison: Guess you pay taxes - don't you? But does that mean you pay them voluntary?

Regarding the comparison between Ohloh's KudoRank and Google's PageRank: I think that comparison works quite well here, actually. Some people publish web pages at least in part because they want visibility, popularity, or to provide their content to the world (or simply appreciate these when they occur); Google's PageRank may help those people. Some people throw content up on a web server because it let them or others conveniently get at it, and some of those people find it surprising that Google indexed their page. (If that sounds surprising, consider how the pre-search-engine web might react to a search engine.) Similarly, some people publish their FOSS contributions at least in part because they want visibility, popularity, or to provide their contributions to the world (or simply appreciate these when they occur), and Ohloh's ranking may help those people; others publish their FOSS contributions exclusively because they wanted to solve a problem or provide their software to a particular community, and some of those people find it surprising that Ohloh indexed their software. (With tongue firmly in cheek: "If that sounds surprising, consider how the pre-Ohloh FOSS community might react to an Ohloh." :) )

Also notable: Google provides several opt-out mechanisms (both by directly telling Google via a form, and by adding appropriate flags to pages), but does not provide an opt-in mechanism. Similarly, I personally think Ohloh would do well to provide an opt out mechanism to satisfy people uncomfortable with Ohloh's indexing of their contributions or projects, but I don't think Ohloh needs to go as far as requiring opt-in.

At the same time, if Ohloh's KudoRank went opt-in, I don't think that would damage it greatly now. Originally, Ohloh's ranking of unregistered contributors provided something useful to rank against; it makes little sense to provide a KudoRank of a small handful of people against only each other. However, with the number of people registered on Ohloh at this point, a KudoRank containing only registered users would probably not prove that much less interesting than one including unregistered users.

Meanwhile, I'm very motivated to change/alter our language or add reasonable disclaimers. Any suggestions?

I'd suggest a disclaimer near any information about unregistered contributors, along these lines: "This information reflects contributions automatically discovered by Ohloh when indexing a registered repository. However, this contributor may have made other contributions which this information does not reflect, either to other projects (including those not indexed on Ohloh) or to this project under different aliases, . Ohloh does not automatically tie together contributions from contributors; contributors can do that by registering and claiming their contributions."

I dont feel that its an invasion of my privacy; commits arent secret. Sometimes they are embarrassing -I was in the top 10 of VB coders one month, after committing 20 lines of open office script to our project. More embarrassing would be measuring the percentage of times my commits broke the build -which is something Apache Gump can do, if it chose. They choose not to, as it runs the risk of penalising risk takers.

What I worry about is more house security. Its easy to tell when I am travelling, as my commit rate drops to 0. I should automate something to make me look busy during months I am away.

I'm on the "public information is public information" camp, so I don't believe any opt-in or opt-out is required. Providing an (easy) opt-out is, however, polite. And you want to be polite, as you depend on goodwill for the site to work.