cvs commit: www/en/projects/ideas ideas.xml

Robert Watson rwatson at
Thu Mar 12 05:24:01 PDT 2009

On Thu, 12 Mar 2009, Alexander Leidinger wrote:

>> I think there could be a useful project to be done on fleshing out kernel 
>> SDT providers, but that we need to come up with a more specific description 
>> if we want it to be useful for GSoC.  We also need to identify a potential 
>> mentor for the project who can field questions during the proposal process, 
>> potentially mentor the project, and so on.
> To have it on the ideas list we don't need a GSoC mentor... basically I say 
> that I think that the DTraceToolkit stuff should stay on the ideas list in 
> some form (doesn't matter to me if it is marked up fro the GSoC or not).

On the ideas list, perhaps, but I feel increasingly strongly that no idea 
without a technical contact should be tagged as GSoC-appropriate, and that we 
should also have mentors in mind for every project there.  Pre-submission 
interaction improves not just the quality of the proposal, but also gives us 
an early sense as to whether they have the right skills so that they can tune 
the proposal to what it turns out the students are able to do. I also feel we 
shouldn't invite people to submit proposals on an idea if we don't have 
someone willing to mentor them, because we've had that happen and it's 
unproductive :-).

Robert N M Watson
Computer Laboratory
University of Cambridge

> That aside, as I already added a lot of dtrace probes in my linuxulator user 
> branch (slowed down a little bit ATM because of other high priority tasks), I 
> feel confident to mentor some DTrace probes related GSoC project (if there's 
> some interest out there that this should be done), so the mentor part should 
> not be a showstopper.
> Bye,
> Alexander.
> -- 
> If it sits on your desk for 15 minutes, you've just
> become the expert.
>    Alexander @ PGP ID = B0063FE7
>       netchild @  : PGP ID = 72077137

More information about the cvs-all mailing list