> Can I request all of this information be deleted?
Yes, the ones that are on HN. I'm not sure how it works for 3rd parties that obtain your data
> does this include other comments that referenced my username?
I don't think so, this is unlikely, especially as you didn't create it and HN doesn't link this (as opposed to reddit)
> What about comments that might have linked to my GitHub profile?
I'd say that being required is even less likely as HN has no way of knowing what's your GH profile
GDPR is what they know about you. If they're actively trying to link pasted GH profiles and usernames then this would apply, otherwise no.
> What if there are server logs that include my IP and a time which can correlate to when I posted a comment or something?
That thing with IPs being PII I'd say this would apply, but then again, this doesn't bring any new information.
So if they keep track of users access times then yes, but if this information is rotated, sent to /dev/null then no.
You're not obligated to connect all the dots, or track user login times. That being said, IP (especially + times) are PII so better anonymize it and discard once not needed.
Arguably much or all of the entries with FKs to those rows as well, transitively. Unless you don't believe in normalisation :)
(I'll leave the normalization debate for later :) )
If they ask for something specific in an informal way, that can be provided
But from the GDPRs data portability point of view, it's everything that's linked to the account. Export your Facebook data for a good example of this.
HN example: it would be the information in your profile, the links/text you submitted (but not the content of the link itself), the comments (the comment IDs) you upvoted, stories flagged/upvoted (for the lenght of time this is kept, for example, if after a while user ids that upvoted a comment are erased and only the score is kept, that's fine) and maybe some other background information (for example: password hashes/access logs/etc)
> what format that information needs to be in,
Machine readable format. HTML/XML/JSON is fine.
> how to locate it and package it, and whether new infrastructure needs to be created to manage this request pipeline.
Well that's not the problem of the law, is it?
You know your data scheme. You know whether you can run this in your existing infrastructure or not.