Giters

meain / gh-issues-to-rss

Convert github issues and prs into rss feed

Home Page:https://blog.meain.io/2021/releasing-gh-issues-to-rss/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Not understanding `We invalidate internal cache only every 12 hours`

TehPeGaSuS opened this issue · comments

commented

Does We invalidate internal cache only every 12 hours means that the RSS feeds will only be updated every 12h?

If so, can we invalidate the cache manually more often?

Cheers

commented

Does We invalidate internal cache only every 12 hours means that the RSS feeds will only be updated every 12h?

Yup, you are right. We save the GH response for 12 hours and only fetch again after 12 hours. This was a quick hack to prevent GitHub from rate limiting. It does not have to be this aggressive, but it was mostly for my personal use and this was working pretty well.

This would be another thing which could benefit from a flag to configure, something like --cache-timeout <interval>. It is unfortunately fixed as of now

If so, can we invalidate the cache manually more often?

Let me see if I can get around to adding a flag over the weekend, but one quick hack would be to delete the cache files located at /tmp/gh-issues-to-rss-cache with a cron job or something. Super hacky, but it would work.

commented

I know that GitHub has rate limits, but I'd use this to follow several GitHub repos, so having less cache time would be greatly appreciated.

Accordig to this docs, when using personal access token, we have way higher rate limits.

Here's some ideas:

  • It would it be worthy to have a way to add/specify a personal access token, via some flag like --access-token <TOKEN>.
  • 1h caching would, probably, be enough for most of the users. If we specify a token, then the standard rate limit would be ignored.

I know that this would mean re-writting stuff, but would be a great improvement.

Cheers!

commented

I know that GitHub has rate limits, but I'd use this to follow several GitHub repos, so having less cache time would be greatly appreciated.

FYI, the cache is per repo.

Accordig to this docs, when using personal access token, we have way higher rate limits.

Yup, you are right. I just never got around to do it as this was mostly a personal project and I didn't really have use for it. Now sound a good time as any to implement it.

commented

@TehPeGaSuS this should now be fixed