Maybe it’s just me, but this feels like a lot more of a pain than just using an awk or sed one-liner (or even the chains of cuts and heads and greps the post author seems to think are the only existing option) to parse some arbitrary text. This is a solution that fully structures semi-structured output from certain commands into JSON. I’m not convinced that learning the exact hierarchies it parses any given command’s output into is worth it. You’re already looking at a nightmare if you put this into a script and release it – the thread for this post on HN has a bit where someone was unable to get it to work, and the author says it’s because their locale isn’t set to en_US.utf8, as the parser is expecting.
Semi-structured output is semi-structured, the only reason this program even works is because it’s trivial to parse the outputs of UNIX commands as text and get the desired information. So why not just do that, especially since you’ll still have to do that when interacting with anything other than the set of built-in commands understood by this program.
First, there’s no official client. The fact that it’s so easy to implement a client means there’s a Cambrian explosion going on, and the filtering die-back has not yet occurred. This might change in the medium future.
Maybe it’s just me, but this feels like a lot more of a pain than just using an awk or sed one-liner (or even the chains of cuts and heads and greps the post author seems to think are the only existing option) to parse some arbitrary text. This is a solution that fully structures semi-structured output from certain commands into JSON. I’m not convinced that learning the exact hierarchies it parses any given command’s output into is worth it. You’re already looking at a nightmare if you put this into a script and release it – the thread for this post on HN has a bit where someone was unable to get it to work, and the author says it’s because their locale isn’t set to en_US.utf8, as the parser is expecting.
Semi-structured output is semi-structured, the only reason this program even works is because it’s trivial to parse the outputs of UNIX commands as text and get the desired information. So why not just do that, especially since you’ll still have to do that when interacting with anything other than the set of built-in commands understood by this program.
as opposed to what official client for the web?
the web at least has standards…
like this?
no. I mean a real technical specification, not a first-draft quality half-baked pseudo-specification that calls itself feature complete.
👌
I guess you haven’t read where the document itself refers to itself as a pseudo-specification…
I guess you haven’t noticed it basically hasn’t changed in like a year plus because it’s stable at this point.
I have noticed it hasn’t changed, and I’ve noticed that it lacks things that could really make it a better protocol
It’s worth reading Tangled Web
Content is still valid as an ebook. :)
Nostarch ebooks are DRMfree when you buy from their store if that’s a concern