[Openchemistry-developers] MolView for GSoC 2017

Sat Apr 1 11:47:31 EDT 2017

Dear Geoff,

Dart is a language developed by Google and mainly used within Google (the
outside community is fairly small despite some efforts to make it more
popular). I couldn't find any Dart related stuff on GSoC. I know 3Dmol.js
and related projects (and have worked with WebGL and GLSL extensively in
the past), I have been wanting to integrate it into MolView as it is more
active (currently relies on a modified version of GLmol, 3Dmol was forked
from GLmol). There is actually one project in there I have been wanting to
do for years (fragment based compound resolving), in fact I communicated
about this with David Koes last year, but at least then my 'advanced
mathematics' skills were not sufficient to get started. The reason I was
(and still am) interested in this is because it would allow client-side
computing of 3D structures (thus MolView can be used offline, and resolving
is probably quicker).

Anyway, I understand your position. I'll see what I can do (I'm very late
with this, but I didn't think of it earlier, I know some people at Google
because of working with Dart).

Regards,
Herman

Op za 1 apr. 2017 om 17:08 schreef Geoffrey Hutchison <
geoff.hutchison at gmail.com>:

> Dear Herman,
>
> We certainly understand your interest in using GSoC to re-start MolView -
> that's one of the goals of Google in running GSoC is to foster open source
> development. And we have an interest in the broader scope of open source
> chemistry. (A good browser-based editor would be a very nice tool.)
>
> The problem is that GSoC is a very specific program from Google. To be
> honest, we do not set up the rules or even pick student projects. We
> *recommend* student proposals to Google, who ultimately makes the final
> decision. Beyond questions on the license, there *must* be a specific
> mentor for a project. While we encourage students to propose new projects,
> we still have to find a good fit with a mentor.
>
> We have no mentors familiar with your code, nor with Dart.
>
> If you wanted to propose a project for browser-based 3D editing using
> 3Dmol.js or Avogadro, that's a completely different thing - we have mentors
> that can serve.
>
> But even a truly outstanding proposal without a mentor will not be
> selected by Google. And it's not just one mentor - Google wants to see an
> organization like OpenChemistry have a pool of ~2x mentors as students
> (e.g., mentors travel and take vacations too).
>
> I haven't looked - does Dart have a summer program? I think you can
> propose "unattached" projects through a catch-all Google organization too.
>
> Best regards,
> -Geoff
>
>
> On Apr 1, 2017, at 10:15 AM, Herman Bergwerf <hermanbergwerf at gmail.com>
> wrote:
>
> Dear Marcus,
>
> Although changing the license is not the primary problem for me, there are
> indeed no other people involved in this project for a role as mentor. There
> is another project that is currently consuming my time and when a friend
> gave me a heads up about GSoC I thought it might be a way to start MolView
> development again, not so much as it turns out.
>
> Thanks for your response anyway, and good luck with GSoC.
>
> Regards,
> Herman Bergwerf
>
> Op za 1 apr. 2017 om 15:48 schreef Marcus D. Hanwell <
> marcus.hanwell at kitware.com>:
>
> Hi Herman,
>
> I don't see a good fit here for a few reasons, first and foremost being
> that a student cannot mentor themselves. The Open Chemistry organization is
> formed around a number of projects where we know and have worked with the
> projects/mentors for quite a few years (in loose online forms in some
> cases, but still). This enables us to build up a level of trust, and enough
> overlap such that we have good confidence that we can offer the good
> experience for the students and mentors.
>
> Finally, we favor permissive, OSI-approved licenses in general that
> encourage reuse in the wider context. While there are some exceptions for
> historic reasons I think as a group that is the direction we (and certainly
> I) would like to follow. The AGPL is very much in the opposite direction,
> which is your project's choice but I don't feel it is well aligned with our
> organization.
>
> I wish you well in your endeavors to develop MolView, but as stated I
> don't see a good fit. Even if there were we decided on included projects in
> the organization proposal period, were selected on our proposal, and intend
> to offer the projects with the mentor group originally proposed.
>
> Yours sincerely,
>
> Marcus
> --
> Marcus D. Hanwell, Ph.D.
> Technical Leader, Kitware Inc.
> (518) 881-4937
>
> On Sat, Apr 1, 2017 at 9:06 AM, Herman Bergwerf <hermanbergwerf at gmail.com>
> wrote:
>
> Hi,
>
> I am Herman Bergwerf, developer of http://molview.org. I have a fairly
> simple question. Would it be possible to sign up my development on
> molview.org for GSoC under the Open Chemistry organization?
>
> I developed MolView in highschool and have been looking into improving it.
> Last year I did a failed attempt at building an architecture from scratch.
> Basically I started of a little too enthusiastic and wrote a completely
> modular, message based frontend architecture that was intended to be super
> modular. But after almost a year of working on this I decided I was going
> nuts and discarded it. However a lot is to be done. I almost get emails
> weekly about people enthusiastic about MolView or requesting new features.
> An old roadmap of ideas is here:
> https://github.com/molview/molview-1st-gen/wiki/Roadmap. I created
> mockups for a new, even more immersive user interface more than a year ago (
> http://blog.molview.org/posts/2015/07/23/material-design/). I abandoned
> development on MolView almost a year ago in favor of working on some other
> projects (also OSS). However I still retain interest in MolView. It is open
> source (AGPL) and so far developed on 0$ funding (except for my own funding
> to pay for hosting, and I hate ads so never added them). In the past 3/4
> years MolView has been visited over 400.000 times by more than 230.000
> unique users. In the past years I developed an interest in the Dart
> language. I would love to rebuild MolView from scratch in Dart,
> implementing the new user interface (as shown in the mockups), along with a
> set of new features (more embedding options, OBJ export for 3D printing,
> DNA sequence viewer, etc.), and drastically improving the code (compared to
> now I was a very bad coder back in highschool ;-D).
>
> I know MolView is not officially a open chemistry project (I'm not sure
> how the organization works), but it would be really awesome if I can work
> on MolView this summer while also making some money. If this is possible I
> can write a proposal with a list of technologies, the basic architecture,
> and an initial set of features.
>
> Thanks!
> Herman Bergwerf
>
> _______________________________________________
> Openchemistry-developers mailing list
> Openchemistry-developers at public.kitware.com
> http://public.kitware.com/mailman/listinfo/openchemistry-developers
>
>
> _______________________________________________
> Openchemistry-developers mailing list
> Openchemistry-developers at public.kitware.com
> http://public.kitware.com/mailman/listinfo/openchemistry-developers
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/openchemistry-developers/attachments/20170401/fc335a2c/attachment-0001.html>