[Owasp-board] OWASP Summer Code Sprint Proposal

Matt Konda matt.konda at owasp.org
Wed Mar 4 12:31:18 UTC 2015


Hi all,

This is an interesting discussion.

I would vote yes to do 30K for GSOC like project sponsorships, HOWEVER I
actually think that OWASP should not be in the business of building
software.  That may seem contradictory but let me explain.

I would vote yes for near term funding project sponsorships because:

   - Certain innovative projects push the industry forward
   - Having a bit of additional work on a given project can help mature them
   - It is a great way to get people involved
   - It is great to be able to support and lend community to ambitious and
   motivated leaders

Collectively, I think these aspects make it like a little appsec petri dish
where people can learn and good things can happen.

The reason I don't think OWASP should aim to be "building software" is that
it is really hard to do well even with money, 30 person dev teams, QA,
marketing teams and sales teams - PER PRODUCT.  There are also a lot of
really hard choices that go into that.  I think if we underestimate how
hard it is to build really good software, we do a disservice to ourselves
in the long run.

We could also end up undermining our credibility on the information front.

I don't care much or at all about disrupting existing vendors that are
sponsors.  If OWASP tools can disrupt them with almost no money and some
volunteers, how great can their product be?

That being said, I've used dependency-check and ZAP and they are some of
the better tools from OWASP, but in my experience, they are not at the same
level as commercial tools.  For those two there are lots you can't even
start to use.  Would we become totally focused on ZAP?  I honestly believe
its actually better to have more immature projects that highlight appsec
areas, get people involved and interested and could potentially be
references for commercial ventures later than to focus on one tool and try
to build an entity around that.

All this being said, I am open to *considering* an apache like model where,
as I understand it, most of the development is volunteer driven and the
projects are not funded - but they are presented in an organized way.  I
think this is a lot of work too - but I'd be open to it if we could define
it and demonstrate how the value works.  Take tools like brakeman or
repsheet or modsecurity - why shouldn't they be things we reference and
support - but what can we give them that is of value?  Why would they want
to be associated with OWASP?  Unless we can give them a great answer, we're
probably wasting our time.

I would also say that since the apache foundation started, there is a
totally different landscape for collaborating on open source software.
Back then, someone had to run cvs/subversion, support distributing
binaries, provide licensing options and provide a web page.  Now, we get
that for free with a click on github.  What is the material support we can
provide?

Regards,
Matt


On Tue, Mar 3, 2015 at 2:30 PM, Jim Manico <jim.manico at owasp.org> wrote:

> +1 I agree with Joshs perspective on this. I'd personally vote no if I had
> to make a decision on these funds today.
>
> --
> Jim Manico
> @Manicode
> (808) 652-3805
>
> On Mar 3, 2015, at 2:25 PM, Josh Sokol <josh.sokol at owasp.org> wrote:
>
> To some extent, I think this gets back to the "should OWASP pay people to
> work on it's tools" debate.  In my personal opinion, I think that the
> answer is "yes", provided that we:
>
>    - Have a pre-defined scope for the opportunity with specific
>    milestones required
>    - Have a pre-defined award for completing the opportunity
>    - Publicly publish any and all opportunities so that anyone can
>    express an interest in them
>    - Have a formal selection process with ideally a committee of leaders
>    making the selections
>    - Those involved in the selection process cannot also submit
>    - Those involved in the selection process are also responsible for
>    assessing completion
>    - All work produced is provided under the same open source license as
>    the project
>
> If we have agreement on these points, then I would suggest extending
> Fabio's proposal to be a much broader OWASP call for ideas (not just GSoC
> submissions).  Put a two week limit on submissions and, once expired, put
> all reasonable ideas someplace public.  Submit a press release stating that
> we are looking for students interested in tackling these challenges and
> providing the details.  As long as this is no longer GSoC, then we get to
> make up our own rules, and I think that we should take a step back to
> evaluate how WE would want this to work.  What goal do WE want to
> accomplish with this initiative.  I'm all for allocating $30k here, but
> don't just want it to be OWASP's rejected rehashing of GSoC.
>
> ~josh
>
> On Tue, Mar 3, 2015 at 1:49 PM, Fabio Cerullo <fcerullo at owasp.org> wrote:
>
>> Dear all,
>>
>> As you probably know by now, we have not been accepted to Google Summer
>> of Code this year.
>>
>> Usually, this is a major push for projects during the year as experienced
>> by ZAP, OWTF, Appsensor, Hackademics, Seraphimdroid, etc. For a full list
>> of ideas in 2015 please check the following URL:
>>
>> https://www.owasp.org/index.php/GSoC2015_Ideas
>>
>> In order to keep the momentum going and progress those projects, I would
>> like to request an extraordinary budget allocation of 30K USD to cover up
>> to 10 student slots at 3K each. Usually Google pays 5500 USD per student
>> during GSOC. We will use the same structure as previous years with
>> Kostas/me as org admins, the project leaders who usually participate in
>> GSOC (Core team) will pick the best student submissions and then a group of
>> dedicated OWASP volunteers who every year act as mentors for the students.
>> We could establish a mid-term and full term evaluation where if a student
>> is failed mid-term he/she will only receive half the funds (1500 USD). If
>> the student is approved full term, he/she receives the full amount (3000
>> USD).
>>
>> I understand this is a non-planned expenditure, but considering the
>> importance of GSOC in the last couple of years to progress OWASP coding
>> projects, I think is imperative to take some action considering the current
>> scenario.
>>
>> If you have any questions, please let us know.
>>
>> Thanks
>> Fabio
>>
>> _______________________________________________
>> Owasp-board mailing list
>> Owasp-board at lists.owasp.org
>> https://lists.owasp.org/mailman/listinfo/owasp-board
>>
>>
> _______________________________________________
> Owasp-board mailing list
> Owasp-board at lists.owasp.org
> https://lists.owasp.org/mailman/listinfo/owasp-board
>
>
> _______________________________________________
> Owasp-board mailing list
> Owasp-board at lists.owasp.org
> https://lists.owasp.org/mailman/listinfo/owasp-board
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.owasp.org/pipermail/owasp-board/attachments/20150304/d2823d87/attachment.html>


More information about the Owasp-board mailing list