Hacker News Reader

Comment by rwmj | Hacker News Reader

It does seem very strange. If I'm looking at a PR, especially a historical one which has been merged already, I want to see all the comments which were made about the commit in one place.

If it was confusing to have these comments turn up "out of context" then maybe that's something that github could have improved instead, by visualising the context better.

View on HN

They seem to be removing heavy features from often-used pages. I would imagine this change creates quadratic savings in data fetching for the timeline.

This could be so, but wouldn't a better solution be to prerender popular pages or cache generated HTML (a la memcached etc)? (Having said that I have no idea how modern web apps work, nor what technology stack github is using.)

Depends how long their tail is, I guess.

Honestly, if that was the reason ("we need to make our mysql faster"), I would strongly prefer if they'd just say that. Certainly more understandable (if not relatable) than this sort of communication.

Maybe they'll bring them back in a little while as a premium feature

I'm still not understanding why this is a justifiable reason. It sounds like something like this is a prime candidate for caching.

Caching what? All possible comments everywhere? That’s just another db, not a cache. And having infinite caches is not feasible, the point is to store widely and frequently accessed data, not anyone’s comments on any commit in the repo.

They've been in a roll nerfing stuff like this ever since MS bought them. See "load more comments" ajax links when there are more than 10 comments in a PR review, or "load diff" when a file has changed more than a few lines. All of these degrade the ability to actually look at changes and review them, which is kind of the whole point of pull requests. But alas, the enterprise must enterprise.

Reply on news.ycombinator.com