Are rockstars good for your project?

[jono]

This is where you can discuss the segment in Season 5 Episode 5 in which we discuss whether rockstars can be good or bad for an Open Source project.

[Allix]

If 2008 is going be the year of the linux desktop, then we are going need more rockstars.
On a serious note as Mr Bacon said we need a good mix of both, you can have the best hackers in the world, but if they cannot get the word out what they are creating its not going go anyway and the same with fantastic publicity to a vapourware has a chance a good chance of killing a product before its even out and then those people might be seen as unable to ever create a project.

I think in the proprietary world they use both, many can argue that they spend more on advertising however the software must to some extent be of a standard otherwise it will fail. As much as people complain about Microsoft, people do get they jobs done using it.

Its proberly not a good idea comparing the open source community to Microsoft , it was just a thought on the software perception from outside.

Maybe i ve taken the segment the wrong way and its just looking at how it is concieved from within the community , which in that case, my post is invalid .

[ozgeek]

I think this is one of those questions that is highly dependant on your definition of the relevant terms.

So for the prupose of my spiel, a rockstar is a programmer:
a) who is well known/famous (as opposed to infamous) for his project
b) credible/widely accepted in the community
c) has kicka ss kacking skills,
d) has good vision
e) probably has good organisational managment skills as a result of managing the project that mafde him famous
f) probably has developed (or maybe even started with) good people skills again, possibly as a result of his epxeriences manging the project that made him famous

As opposed to a programmer who has all of the above, except maybe a and b.

I don't think you need rockstars for a good project.
I do however think that good projects may produce rockstars.

I mean most of the well known/popular projects did not have rockstars to start with.

Typcially/traditionally a project's founders were relative zeros (fame wise) but became rockstars when their projects took off and got well known and accepted Or at least this was true until a couple of years ago. Now we are startig to see rockstars leave or branch out from their firstborns and starting/joining new projects.

Who knew Linus in '92 or even in '96?
Did that stop linux? Nope. He successfully started what is without doubt in my miond, the most popular, well known free software project around, but he was a realtive zero back when he started it.
He was not as far as I can tell even well known in academic circles.

Once the project is up and running, i.e. once the founder (s) (who may now be a rockstar(s)), has got a decent implementation out and in wide-ish spread use, and the project is accepted and has attracted a team to hack on it, then you also probably don't really need a rockstar.

At that point having a famous lead might be useful to sell new ideas/radical changes (and paradigm shifts!! ) or new projects, but is probably not anymore useful than just decent programmers/architects/team members who put in their time grinding out the boring stuff: bug fixes; docco; incremental new feature introduction etc.

As alluded to above, I think the value of a rockstar is probably not so much for the project that originally made them famous, their value is probably in subsequent projects, because:
a) people will listen to them and be more rceptive than they might from some bloke who is as yet unkown
b) they now have one project under their belt and the experience and knowledge that goes with that and
c) they still have the same vision and coding ability that they had before.

E.g. Linux could probably live without Linus now from all perspectives other than there needs to be a crown prince to replace him. But if Linus went off and started abnnother project (e.g. git) it will probably take off (or take off faster than if a relative newcomer/noone started it) and benefit from his fame and his now well known management skills.

Another example is emacs.
RMS does not really need to have anything to do with emacs (and I think he does nto have much to do with emacs dev now), but he has managed to leverage his rockstar status from emacs on subsequent projects, where his fame/credibility and experience hae proven useful. E.g. FSF.

I'm not saying that having a rockstar is not possibly useful for a project, but I am saying that I don't think it is needed for a good project.
I guess I think a rockstar probably speeds up acceptance for projects features, but probably does not make them any better from a technical point of view or a group dynamic point of view.

[mrben]

c) has kickass kacking skills,

The brown note rears it's ugly head again....

[Foaf]

I took Thomas's post to not to be about specific people or projects but about an attitude. The rockstar is the person who always wants to take their art to the next level often to the point discarding what came before and ignoring the wishes of the existing fanbase.

That sortof happened with Gnome. V1 was a promising debut. V2 was (almost?) a complete rewrite and a bona fide hit. Thomas's "rockstar" would now discard the V2 codebase in an effort to create the next big thing.

To stretch the musical analogy, it seems to me like Thomas would rather work with a tight set of session/covers musicians who know the songs, know what the fans want and can put on a good show. They may not be making drastic changes or game-changing innovations, but they are making it better and making people happier, one bug, one feature at a time.

[haakin]

I think that in the segment somebody said that a rockstar programmer is the one who is too big to correct bugs. I know several examples of these situation in research. A good researcher, after several years working hard, becomes leader of a project. In this precise moment it's almost impossible for that person to continue doing research. Instead of that he or she must spend the time in budget meetings, filling applications, giving lectures, participating in conferences, etc. This person cannot work in the lab anymore, but not because he or she doesn't want, it's because the rest of the tasks fills all the available time.

Probably it's the same for rockstars programmers. They are good IT engineers that must start working as people engineers because their status.

[QuantumG]

Rockstar programmers are "rockstars" for the same reason that "rockstar singers" are "rockstars".

They're big ass self promoters. They toot their own horn.

It doesn't matter what they do. It doesn't matter if they do nothing but talk.

[mrben]

Rockstar programmers are "rockstars" for the same reason that "rockstar singers" are "rockstars".
They're big ass self promoters. They toot their own horn.

Hmmm - I'm not sure that's true (in this context). I would have said they have become "rockstars" through actual ability and contribution to a project, and when the project has received press they have received much of the "reflected glory" without necessarily being self-promoters.

[Aq]

Rockstar programmers are "rockstars" for the same reason that "rockstar singers" are "rockstars".
They're big ass self promoters. They toot their own horn.

Hmmm - I'm not sure that's true (in this context). I would have said they have become "rockstars" through actual ability and contribution to a project, and when the project has received press they have received much of the "reflected glory" without necessarily being self-promoters.

Absolutely. Our question wasn't really meant to be "does your project need a self-aggrandizing arsehole who steals all the credit and never does any work?" We'd have all agreed instantly that that was a bad idea, and moreover I personally don't think that there are many (if any) people like that in the free software community. Instead, it was "does your project need someone who has high visibility and fame (even given that that high visibility and fame is justified by their contributions)?"

[QuantumG]

No offense to all the rockstar programmers out there, but if their response to media attention was "that's nice, but I have work to do" or simply "no comment" then they wouldn't be rockstars. And yeah, it's not really about taking all the credit. It's about having a personality and associating that personality with the project.

So I guess your question really is, what do you gain from a "cult of personality" and do you need it?

I think the success of Apache and other rockstar free projects shows that you don't *need* it but it certainly helps to get new projects off the ground - "have you heard what Theo is working on now?!"

[Foaf]

Brian Behlendorf's work Apache httpd certainly game him "rockstar" status, which no doubt helped when he started collab.net

[Nightwish]

I think that from that point of view you can say it takes the spotlight off others, for good and for bad.