-
Notifications
You must be signed in to change notification settings - Fork 726
Static site for rustls.dev, including home for performance reports #2201
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Benchmark resultsInstruction countsSignificant differencesThere are no significant instruction count differences Other differencesClick to expand
Wall-timeSignificant differencesThere are no significant wall-time differences Other differencesClick to expand
Additional informationCheckout details:
|
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #2201 +/- ##
=======================================
Coverage 94.63% 94.63%
=======================================
Files 102 102
Lines 23928 23928
=======================================
Hits 22645 22645
Misses 1283 1283 ☔ View full report in Codecov by Sentry. |
Did you consider picking up something like oranda? Seems like it might be a good fit. |
I haven't heard of that, will look into it. |
OK, I spent an hour or so with oranda. I like the way it can present github release notes, but unfortunately it only really supports a handful of built-in themes (and limited CSS customisation on top of that). I filed an issue about that, but in the meantime I think that possibility need not block this? |
The |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Zola seems pretty heavy weight for this case and I think I'd prefer the Makefile personally. A Rust script seems OK.
e632875
to
1e5e4aa
Compare
intel.com specifically returns 403 from GHA runners.
1e5e4aa
to
3eaa70b
Compare
Zola seems to work pretty well, stays out the way, and has good docs. I'm minded to write the absolute minimum of code for this. |
My enthusiasm for static site generators was reduced by the experience of Hugo and Jekyll. Maybe Zola will be better, but if it's like those two then there's a non-zero amount of headache related to keeping up to date with the tooling and associated breaking changes. This branch is also hardcoding a Zola version in CI in a way that I think excludes Rennovate from helping us keep it in-sync with upstream releases. Is there a way to avoid that? I don't feel strongly enough to block this getting merged but given there's only one page with no significant site structure to contend with it feels like dragging in a whole site generation framework is overkill. |
I noticed Curl has some special markup in some of its CI YAML that looks like it configures Renovate to update versions, e.g.: https://github.com/curl/curl/blob/774844ab5db84e2368389e6cd9f0b8f81852b738/.github/workflows/linux.yml#L38-L57 That might be an option if we stick with Zola. |
I think these two points kinda cancel each other out a bit? For the purposes of static site generation, I'm happy to fix on a working version until we have a reason to change (for my own blog, I basically did that with jekyll, only upgrading it 3 times in 11 years). |
I feel like that just shifts the pain, collecting it all up at the point where you have to jump many versions at once. Put another way: if we adopt Zora we're on the hook for future update pain. We can either pay it in small increments to keep up with the tooling as it's released (generally easier IME but still annoying), or we can pay it in large lump sums when we're forced to for feature-needs or because the tooling stopped working on our systems. We're using so few of the features of Zora that it seems excessive weighed against a small makefile directly invoking the one thing we do want, markdown -> HTML for ~a single file but if I'm in the minority here I'm happy to go along with the crowd. |
I'm sympathetic and don't mind either route. That said, I'm optimistic about the Zola route just based on the notion that the Rust culture around backwards compatibility is pretty good. I think Joe will most likely end up being on the hook for maintenance so ultimately his call? |
I think one of two things will happen:
|
This is a more permanent and visible home for https://github.com/rustls/rustls/wiki/Performance-measurements-(2024%E2%80%9010%E2%80%9018) -- with more to come (for arm64, and multithreaded cases).
At the same time, I started a landing site for rustls.dev. Feedback on what we say there gratefully accepted!
To see what this looks like, schedule https://github.com/rustls/rustls/actions/workflows/docs.yml for the branch
jbp-landing-page
and then look at https://rustls.dev/ . Each timemain
changes (and nightly) this will be lost and you'll see a github 404.