Sean (Jan 18 2024 at 20:22): Sean (Jan 18 2024 at 20:22):@all ! There was a great Google kick-off meeting last week for admins and mentors that I know at least @Daniel Rossberg and I were at, maybe others as well. This is going to be the 20th year of the program! I'm going to send the usual call-for-mentors e-mail out here in a bit, but there's at least one topic that might be good to discuss more fluidly, namely how our project ideas are presented.
This was one of the topics discussed at the kick-off meeting that had me reflecting on the challenges we've had the past couple years attracting student proposals across all partner orgs. Google specifically called out not just pointing candidates at an issue tracker -- unless we really tailor it to the program, which we do, but the underlying point remains about making the list more reader-friendly and scannable. To address that, I'm wondering how much work it would be to "publish" the combined list to a single page?
We also need to balance the number of topic ideas so that there's 1-3 ideas per mentor, namely that there's a specific mentor behind the proposed idea and that mentor is specifically (and contemporaneously) advocating for that project idea, not just a big list of things needed in the past that we'll find a mentor for if we get a good proposal.
The end goal for our ideas list, assuming a goal of 6-8 slots, would be an ideas list with about 20-25 ideas total, ideally randomly presented on each view, with titles and summary descriptions curated to be easily understood (and then perhaps linking to the issue page with more details).
Sean (Jan 18 2024 at 20:24):If you are interested in mentoring this year, please do indicate interest here (include name and which partner-org) or on the mailing list so we can get a rough head-count and list of partner-orgs.
Divyanshu Garg (Jan 19 2024 at 07:20):Hey @Sean
I am interested in mentoring this year
Torsten Paul (Jan 19 2024 at 09:07):I think the dedicated issue list with tag filtering is actually quite good and would take a big amount of work trying to host that as separate application. What we might be able to do without too big of additional effort is extracting that issue data into a file and using that for backing a single web page.
That would let us keep all the maintenance UI via github issues and also the link to those filters, while giving a better entry page than the currently very static one while keeping that part simple as it's view-only.
Erik (Jan 19 2024 at 11:35):I'm interested in helping with BRL-CAD. I found a little node utility that pulls the issues down as CSV at https://github.com/gavinr/github-csv-tools , the only complication I see with that side of it is that the body is in markdown. I think it'd be pretty trivial to read the CSV and print a static webpage (then cron it or whatever), mostly just deciding on presentation. Randomizing per view would be a bit more involved, needing some kinda code somehow to generate a dynamic view.
Torsten Paul (Jan 19 2024 at 23:57):Fetching the github issue list as JSON and presenting that (ignoring the markdown for now, but that should be possible too via the highlight library I believe) is not that complicated.
For testing I've created a repository for the site generator at https://github.com/opencax/project-proposals which deploys to https://opencax.github.io/project-proposals/ - Question is now what to present and how.
Torsten Paul (Jan 20 2024 at 00:27):So the markdown is pretty much working too, it may just need a couple of tweaks to some of the issues.
Torsten Paul (Jan 20 2024 at 21:24):Ok, I'm done for now - updated https://opencax.github.io/project-proposals/ - waiting for feedback.
Sadeep Darshana (Jan 21 2024 at 13:07):hey @Sean @Daniel Rossberg
I'm interested in mentoring for BRL-CAD.
I prefer to mentor someone working on Arbalest (BRL-CAD GUI) since I initiated that project as part of GSoC 2020.
I'm ok with mentoring for a different project if Daniel or someone is already mentoring for that.
Himanshu (Jan 22 2024 at 13:30):Sadeep Darshana said:
hey Sean Daniel Rossberg
I'm interested in mentoring for BRL-CAD.
I prefer to mentor someone working on Arbalest (BRL-CAD GUI) since I initiated that project as part of GSoC 2020.
I'm ok with mentoring for a different project if Daniel or someone is already mentoring for that.
I can also join for this one. Love to see the progress of what I left.
Sean (Jan 22 2024 at 17:37):Torsten Paul said:
Ok, I'm done for now - updated https://opencax.github.io/project-proposals/ - waiting for feedback.
Oh my gosh that looks absolutely fantastic @Torsten Paul ! I love the detailed drop down, sorting, and link through to the issues.. That really is exactly what I had in mind (or better even). Thank you so much! That should really help.
Sean (Jan 22 2024 at 17:37):@Himanshu Sekhar Nayak and @Sadeep Darshana that's great to hear. Are you both already set up on the mailing list?
Himanshu (Jan 22 2024 at 18:27):Sean said:
Himanshu Sekhar Nayak and Sadeep Darshana that's great to hear. Are you both already set up on the mailing list?
Nope, I am not in the mailing list. Where I can opt in?
Emmett Lalish (Jan 23 2024 at 02:45):Joining in from the manifold thread, I believe @pca006132 @Johnathon (zalo) Selstad and I are interested in mentoring for the Manifold project. We've already added some project issues to the opencax github.
Sean (Jan 23 2024 at 15:47):Himanshu Sekhar Nayak said:
Sean said:
Himanshu Sekhar Nayak and Sadeep Darshana that's great to hear. Are you both already set up on the mailing list?
Nope, I am not in the mailing list. Where I can opt in?
Once I've gathered names for this season, I send an invite which you should get via e-mail
Sean (Jan 23 2024 at 15:48):Sadeep Darshana said:
hey Sean Daniel Rossberg
I'm interested in mentoring for BRL-CAD.
I prefer to mentor someone working on Arbalest (BRL-CAD GUI) since I initiated that project as part of GSoC 2020.
I'm ok with mentoring for a different project if Daniel or someone is already mentoring for that.
I'm perfectly fine with mentors attaching themselves to the projects they wish to mentor, and it's okay if there is overlap. We typically strive for a 2:1 mentor to contributor ratio if possible anyways, but 1:1 is also okay.
Sadeep Darshana (Jan 23 2024 at 18:15):Please add me too @Sean when you send the invite.
darshanasadeep@gmail.com
Torsten Paul (Jan 30 2024 at 19:14):A note on the project list webpage, the github action is scheduled once a day https://github.com/opencax/project-proposals/commit/5b700bdda42c285a35ffe7cf90b4669a3cd0120b - so changes in the project issue list will be picked up 9:17 UTC.
Sean (Feb 08 2024 at 16:34):I just wanted to share that the 2024 application was submitted and the new project-proposals page is what was used for our project ideas page. It looks really great! That also means it's time to confirm mentoring and finalize/update project descriptions.
For all mentoring, please check over the list of projects and make sure your name is on at least one of the topics. Make sure that topic has a very clear title, and any extra details can go into the description.
Himanshu (Feb 09 2024 at 04:02):Sean said:
I just wanted to share that the 2024 application was submitted and the new project-proposals page is what was used for our project ideas page. It looks really great! That also means it's time to confirm mentoring and finalize/update project descriptions.
For all mentoring, please check over the list of projects and make sure your name is on at least one of the topics. Make sure that topic has a very clear title, and any extra details can go into the description.
I am interested for 2 projects - New BRLCAD GUI and Physically-Based Rendering (PBR) advanced shaders. For the 2nd project, I want to learn side by side with you and whoever chooses that project. It will be helpful for me.
Himanshu (Feb 09 2024 at 04:05):I have little experience while I was following Raytracing series of Peter Shirley. So may be a good start for me here as well in this project.
Sean (Apr 03 2024 at 12:32):@Matteo Balice for submitting a code change that goes with your proposal, generally speaking one is required. Any code change is better than no code change, however small. Code changes that actually fix a bug or are a feature improvement are best. Code changes that are a fix or an improvement and also relates to your proposal are a bonus.
There are literally hundreds of potential ideas in the top-level source repository's BUGS and TODO files of vaying complexity. There are also various FIXME notes throughout the codebase that you'll come across reading code. There are also issues on GitHub. You're also welcome to simply run one of the tools, discover something that could be fixed/improved, and work on that.
It's a very large system, so the possibilities are practically endless. Part of the task is assessing your ability to navigate code and talk with other developers.
Sean (May 01 2024 at 21:35):Thank you to all our applicants and congratulations to our four selected BRL-CAD contributors, and five other umbrella participants. You all made the selection process difficult with such variety and quality of proposals. It was definitely a challenging process, but kudos to those selected for your efforts communicating with mentors, working on code contribution(s), and of course for your proposal writeups. Looking forward to working with you.
Matteo Balice (May 01 2024 at 21:56):@Sean Thanks for the opportunity. I noticed that @fall Rainy was also accepted for the same research project of neural rendering as mine. Was the idea for us to collaborate? If so, I think it would be even better, and it would probably lead to better results. I look forward to your response.
Sean (May 02 2024 at 02:36):Hi @Matteo Balice and @fall Rainy , the idea is not to collaborate as GSoC projects must be independent but I did want to talk to you both about your options. I was actually just starting to articulate my thoughts in an e-mail which I hope to send here in a bit -- keep an eye out for it.
Sean (May 16 2024 at 17:44):Sean said:
Hi Matteo Balice and fall Rainy , the idea is not to collaborate as GSoC projects must be independent but I did want to talk to you both about your options. I was actually just starting to articulate my thoughts in an e-mail which I hope to send here in a bit -- keep an eye out for it.
On that note, please feel free to check out the latest papers from Siggraph 2024 on nerfs and radiance fields! https://radiancefields.com/siggraph-2024-program-announced
fall Rainy (May 17 2024 at 05:18):Thanks! I will continue to follow this conference, especially the "NeRFs and Lighting" sections. Regarding collaboration, I have an idea:
@Matteo Balice and I can approach the project in different ways. For the sampling method, he will use the "bounding sphere method," while I will use the "boundary-based active learning approach." In terms of network design, he will employ various methods to improve the network, whereas I will design an entirely new network, with one possible option being the KAN network. We can maintain consistency across various interfaces, which will facilitate communication and help us find the best solutions.
Matteo Balice (May 17 2024 at 09:39):fall Rainy ha scritto:
Thanks! I will continue to follow this conference, especially the "NeRFs and Lighting" sections. Regarding collaboration, I have an idea:
Matteo Balice and I can approach the project in different ways. For the sampling method, he will use the "bounding sphere method," while I will use the "boundary-based active learning approach." In terms of network design, he will employ various methods to improve the network, whereas I will design an entirely new network, with one possible option being the KAN network. We can maintain consistency across various interfaces, which will facilitate communication and help us find the best solutions.
Yes, I agree with you.
Himanshu (May 21 2024 at 14:42):I have updated GSoC 2024 wiki page where you can add your dev log, project plan and abstract. Here are some samples you can see from past years. 2023, 2022.
Vidit Jain (May 23 2024 at 15:19):Himanshu said:
I have updated GSoC 2024 wiki page where you can add your dev log, project plan and abstract. Here are some samples you can see from past years. 2023, 2022.
Ig the hyperlinks are wrong in this.. its pointing to https://brlcad.zulipchat.com/
Sean (May 23 2024 at 16:58):Vidit Jain said:
Himanshu said:
I have updated GSoC 2024 wiki page where you can add your dev log, project plan and abstract. Here are some samples you can see from past years. 2023, 2022.
Ig the hyperlinks are wrong in this.. its pointing to https://brlcad.zulipchat.com/
Link is https://brlcad.org/wiki/Google_Summer_of_Code/2024 and change that to 2022 or 2023 for previous years.
Himanshu (May 24 2024 at 03:20):Vidit Jain said:
Himanshu said:
I have updated GSoC 2024 wiki page where you can add your dev log, project plan and abstract. Here are some samples you can see from past years. 2023, 2022.
Ig the hyperlinks are wrong in this.. its pointing to https://brlcad.zulipchat.com/
Oops didn't see that I had pasted brlcad zulip chat. My bad.
Last updated: Oct 09 2024 at 00:44 UTC