Feedback on PurlDB Popularity Scoring Proposal #299
Closed
hiyasanghvi
started this conversation in
GSoC
Replies: 2 comments
-
|
@hiyasanghvi you don't have contributions which have some amount of working, tested code, and this is key for us to be able to even consider a proposal for GSoC as this is the strongest way of you showing us you can finish the proposed project. You have a single contribution which is on docs and scripts, and this is not enough, without this we are not able to provide any feedback on your proposal. https://github.com/aboutcode-org/aboutcode/wiki/GSOC-2026#your-contributions |
Beta Was this translation helpful? Give feedback.
0 replies
-
|
Hi, thank you for your feedback. I understand that my contributions are
currently not sufficient. I had my final exams recently, which limited my
ability to contribute earlier. However, I have now started working on code
contributions and will be submitting PRs shortly. Please let me know if
there are any specific issues I should focus on. I would really appreciate
your guidance.
…On Mon, 6 Apr 2026 at 5:51 PM, Ayan Sinha Mahapatra < ***@***.***> wrote:
Closed #299 <#299> as
resolved.
—
Reply to this email directly, view it on GitHub
<#299>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/BIO2CUBFFB5FH6RQ2AC5V434UOOMBAVCNFSM6AAAAACXDEWHW6VHI2DSMVQWIX3LMV45UABFIRUXGY3VONZWS33OIV3GK3TUHI5E433UNFTGSY3BORUW63R3GI2TCMJUGI2Q>
.
You are receiving this because you were mentioned.Message ID:
<aboutcode-org/aboutcode/repo-discussions/299/discussion_event/2511425@
github.com>
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hi @JonoYang @pombredanne @AyanSinhaMahapatra @tdruez @TG1999 & ,
I'm Hiya, a GSoC 2026 applicant working on a proposal to improve package popularity scoring in PurlDB.
Before drafting the proposal, I explored the codebase and implemented a working prototype:
• Built and ran PurlDB locally using Docker
• Implemented a ranking pipeline that computes popularity scores and stores them for real packages
• Identified the version-fragmentation issue (e.g., requests==2.28.0 vs 2.31.0 being treated separately) and handled normalization at the package level
• Explored integration points for exposing scores via API and scheduler jobs
Repository: https://github.com/hiyasanghvi/purldb
I’ve shared my draft proposal here:https://github.com/hiyasanghvi/purldb/blob/main/PURLdb_GSOC_proposal.pdf
I’d really appreciate feedback on one key point:
→ Does the current approach (version-collapsed scoring + heuristic ranking) align with PurlDB’s long-term direction, or should I design this with ML-based ranking in scope from the beginning?
Thanks for your time!
Beta Was this translation helpful? Give feedback.
All reactions