[Clam-devel] Re: GSoc project idea.

[dgarcia at iua.upf.edu]

Some tips for your application (and whoever read this in your same
situation).

First of all, you can apply for several proposals but for each one be
clear and concrete on the scope. Main questions we are interested about
the proposal is: you know which is the current CLAM state related to your
proposal, which CLAM state will be after your proposal (3 months), why
this step is interesting to CLAM and why you are the one to fill the gap.

You have your own project. This has come to be a frequent situation for
this year applicants. Having a project of your own is good for the last of
the questions above as it demonstrates experience and it might be a source
of code. But applicants in such situation like you tend to mix up their
proposal with their own project. Remember that you are applying to CLAM.
After reading your mail i am not sure if you are talking about what you
have coded for your own project or if you are talking about what you are
proposing to do for CLAM. Stressing too much a different project in the
goals of your proposal is not a good sign for us. We have given some
resources from Google to fund CLAM development and we want to spend them
in CLAM and we want involved students to stay after GSoC. Summarizing the
tip: target CLAM on your proposal and use your previous experience and
available code base as such.

I also might missread your mail and in fact you are not doing any proposal
at all just presenting your credentials, which is ok, but now let's do a
proposal :-)

A second advice in the same line: what kind of project?

Choosing ideas from the GSoC_Ideas wiki page is good because those are the
ideas CLAM team is interested in. Original ideas are good too but have
less chance to get accepted. Spend your creativenes in dressing those
existing ideas.

A rule of thumb with some milleage variation: as project our goals are in
this order:
* Polishing and enhancing the CLAM core (subnetworks, plugins, backends,
widgets, networkeditor facilities...)
* Make existing applications more usable (normally in the Ideas List)
* Fixing/enhancing existing processing modules
* Having new end user applications based on CLAM
* Adding fancy (not in the list) features on existing applications
* Adding new processing modules and widgets

If you are to work in an application, a fancy proposal must be a good (but
still feasible) one to scale positions on our priorities. There exists an
scale (NetworkEditor>Annotator>SMSTools) on how finnished they are. The
less finnished the greater gap between fancy and basic proposals. Adding
new processing modules and widgets not a need by itself but something that
other goals would require.

Of course, that's a rule of thumb, an IRC talk is useful to explore our
interest in a completely original idea.

David.


> I forgot to add this to my last mail.
> The suggested project of chord extraction, can also be made
> use of. It can even be used to change the scale of the sound, along with
> the other effects.
>
> On 29/03/2008, Abhishek Batra <abhshk.btra at gmail.com> wrote:
>>
>>
>> I have been working on a project independently. It is a sound processor
>> ( coded in python ). A version of it has already been demonstrated in
>> Birla Institute of Technology and Sciences' tech fest in 2007. Though it
>> is currently just for win32, I have been working on development for
>> other
>> platforms.
>>
>> The most important feature of my project is that it has the capability
>> of  processing sound on the fly. Hence. mixing of various media is
>> possible. Other features/ effects that it supports are: Distortion,
>> feedback effects, echo, reverb and equalization.
>>
>> I am fluent with c++ and python, and in general with algorithms.
>>
>> I'll be grateful if someone could help me out with the final proposal if
>> this
>> sounds good.
>> --
>> Abhishek Batra
>>
>
>
>
> --
> Abhishek
> _______________________________________________
> Clam-devel mailing list
> Clam-devel at llistes.projectes.lafarga.org
> https://llistes.projectes.lafarga.org/cgi-bin/mailman/listinfo/clam-devel
>