Wladimir Palant

Not sure what’s inconsistent here. In my test database I can see only one ISO date: `_LAST_MODIFIED` key stored under ``. That’s a non-standard entry by KeePassXC. Everything else is...

Ah, I see. Yes, I was also surprised when I read about performance being the reason for this change. I cannot imagine there being enough dates in a database to...

Note that a weak Etag has performance advantages which is likely why nginx uses it (`{last_mod_time}-{file_size}` with both parameters in hex). In particular, it’s a prerequisite if one is to...

The way I read the code, Buttercup is currently using PBKDF2-HMAC-SHA1. So it would require an even more considerable increase. Using PBKDF2 has a severe downside: while PBKDF2 computation doesn’t...

Yes, there is no native support for anything other than PBKDF2 which is an issue. I’ve found https://github.com/antelle/argon2-browser to have okay’ish performance in the browser however. Don’t know whether better...

> Also, we're using SHA256 there - https://github.com/perry-mitchell/iocane/blob/master/source/symbols.ts - not SHA1. I see, I misunderstood the default algorithm here.

> salt length above standard recommendation Salt length is irrelevant, as are rainbow tables if a per-user salt is used. Strong master passwords would help however. Personally, I [arrived at...

> but I'm honestly not sure it's worth the effort right now unless someone would contribute it. I can certainly relate. Back [when I did this](https://github.com/palant/pfp/issues/58) it was quite some...

I think that’s a duplicate of #2384.

I see. That’s the change I meant here: > I considered changing Endpoint::tcp (and Endpoint::unix) method to automatically recurse for TLS endpoints instead but I doubt this is how the...