What were the problems with the previous site? I used it without major problems.
Wouldn't a browser plugin require you to actively be using HN, or at least have a tab open? Getting an email notification four hours later that someone replied to you on HN, so you may want to take a look is nice (if you don't want to be a slave to HN, don't check your email as often or don't deal immediately with this email).
Additionally, in a browser extension, there's then a lot of people grabbing the comment history rather than a single program, which necessarily puts a higher load on HN.
Major problem with it going offline, hence this new tool, which may or may not go offline at some point again. I am sure you could build an extension that polls the HN API every couple of seconds or so to check for replies and have it show a popup when you get a notification as well as send an email.
> Major problem with it going offline, hence this new tool, which may or may not go offline at some point again.
I'm not sure I understand this sentiment. It's a tool that provides a capability that wasn't there before, so event if it goes away later, you still get the utility it provided while it worked. Since the cost is just providing and email address, that seems a worthwhile trade to me. If it was a trade between this and some other implementation that worked, and it's possible closure at a later date would force you to set up with another service afterwards and you wouldn't if you chose the correct service in the first place, then that's a different story.
> I am sure you could build an extension that polls the HN API every couple of seconds or so to check for replies and have it show a popup when you get a notification as well as send an email.
Well, I'm not sure it would be nice to HN to query them an extra time every few seconds X the number or people that might install the extension. If it's a query every 10 seconds, and 1000 people run it, you've just increased their load by 100 queries a second.
Which is a distinction I don't think really matters. They have an agreement, and firebase provides the API, but that agreement could become strained or need to be renegotiated if there's a feeling that the HN API is taking more resources than it's worth, or if HN pays directly for usage, then we're back at it affecting HN again.
Ultimately, the bits we consume aren't free. Someone pays for them, it just commonly not the consumer in the current model. At the point where you make the person that is paying for them feel it, then the person who pays might end up shifting to you. Or the resource goes away, and everyone loses.
I'm not saying that's likely in this case, but even if it isn't likely, doesn't mean we shouldn't think about it.
Well in a browser extension, you wouldn't necessarily have a lot of people grabbing the comment history. The program could still run as it does, calling the HN API every ten minutes for new replies. The browser extension could just call the service to see if it has found any new replies for you particularly.
Wouldn't a browser plugin require you to actively be using HN, or at least have a tab open? Getting an email notification four hours later that someone replied to you on HN, so you may want to take a look is nice (if you don't want to be a slave to HN, don't check your email as often or don't deal immediately with this email).
Additionally, in a browser extension, there's then a lot of people grabbing the comment history rather than a single program, which necessarily puts a higher load on HN.