[citruspi]

> If it fails it raises an Exception... If your server is down, my application would crash too

So first off, let me just say that I completely agree. If this were the case, that'd be fucking atrocious and would definitely be a blocker for using it.

But I'm curious how you came to that conclusion.

It took me < 2 minutes of looking at the source code[0] to determine that your claim was incorrect. Not only does it appear to gracefully handle the server being unavailable, the developer literally commented that code explaining that they wanted to ensure users could continue uninterrupted if the errorship server is unavailable.

> We give people the benefit of doubt.

> We only consider people to be not authorized if the backend comes back with an authoritative answer to that effect.

> Else, any errors or any other outcome; we assume authorization is there and also assume they belong to the highest pricing plan: Enterprise

> # failure of errorship should not lead to people been unable to ship exceptions

And it took even less time than that to run a new Python Docker container, install the library, run the sample code, and validate my assumptions[1] (the first attempt fails because the key is invalid, I disabled Internet access for the second attempt and it succeeded).

So I'm legitimately curious - did I miss something? Is there another failure case I didn't catch or test for? Or did you just make an assumption and not bother to verify it? And if it's the latter, why? What was the point? Like, to be frank, if this was a news piece I could understand the (possibly inaccurate) commentary. But why take the time and energy to write your comment and tear down someone's personal project with seemingly inaccurate claims?

(To be clear, no affiliation with errorship, I'm not even a DataDog user. Just a random dev browsing HN).

[0] https://gitlab.com/errorship/errorship_python/-/blob/master/...

[1] https://gist.github.com/citruspi/16d359ac2dafef6fc876e2dd101...