Can't edit silver, but can edit passing & gold

[rvagg]

When logged in, visiting https://bestpractices.coreinfrastructure.org/en/projects/29?criteria_level=1 and clicking on "Edit", which should take me to https://bestpractices.coreinfrastructure.org/en/projects/29/edit?criteria_level=1, I get redirected to https://bestpractices.coreinfrastructure.org/en. But this doesn't happen for criteria_level=0 and criteria_level=2.

(Ref: https://github.com/nodejs/security-wg/issues/953)

[david-a-wheeler]

That is.... weird. I'm not sure why that's happening. Thank you for the report.

[david-a-wheeler]

@andrewfader - any ideas?

[david-a-wheeler]

Can you try again? I'm not having any issues. I hate problems that are intermittent :-(.

[rvagg]

yes, sorry, I thought you fixed it because it started working again!

[david-a-wheeler]

I guess we'll leave this closed because I can't figure out how to duplicate the problem to track it down. Thanks for the report, we will keep looking to see if we can keep it from happening again.

[rvagg]

This issue is back again and the team is keen to update their Silver status.

These work, give me 200's and let me edit the pages:

• https://www.bestpractices.dev/en/projects/29/edit?criteria_level=0
• https://www.bestpractices.dev/en/projects/29/edit?criteria_level=2

This doesn't, and gives me a 302:

• https://www.bestpractices.dev/en/projects/29/edit?criteria_level=1

I've cleared cache, tried a completely different browser, it's persistent.

[david-a-wheeler]

Well, ugh. I'm very sorry you're having this problem (again). I got a different problem report that I suspect is caused by the same problem. So let me fix that, and hopefully your problem is "fixed for free" :-). But if that doesn't work, I'll need to track this one down too.

[rvagg]

I had a bit of success today when I thought I'd try varying the request: https://www.bestpractices.dev/en/projects/29/edit?criteria_level=1&boop gets me in and seems to have worked in letting me edit the form; so I guess this might be a caching problem somewhere in the stack that's holding on to the redirect and just busting the cache gets past the problem?

[david-a-wheeler]

I still can't reproduce the problem, sigh, which makes this hard to fix.

I think you're probably right, there must be a subtle caching problem. We've made a few tweaks to the cache invalidation that hopefully resolves this.