Recent searches
Search options
The fact that 99% of my RPC and event serialization has to be done in json is blowing my mind.
Where did we go wrong??
“It’s so simple” they say, wtf?
This is the direct consequence of “I program with vim and 80s unix tooling is the best we’ll ever need” thinking, imo
@mnl Would you prefer something that exists, or something better that doesn't?
I've used grpc extensively and do find JSON interfaces valuable for the simplicity and ease of debugging
@llimllib something existing. XML if you need the whole schema/validation/streaming/transformation thingamajig (but the tooling is rotting so it's ... pragmatic to maybe not)
- protobuf/flatbuf/anything decent binary for minimal parsing and memory overhead, streaming and just simplicity.
- yaml for human readability (i usually convert JSON to YAML for when I need to actually look at things.
@mnl We have exactly opposite tastes in data formats!
- I find YAML amazingly unreadable
- protobuf (etc) is imo only useful for high volume/performance sensitive applications. I had good luck using it in grpc, but it is vastly harder to debug than json and that caused real pain in my team
- XML is rightfully mostly-dead
(Also I do use vim and Unix tools and resent that crack a bit)
@llimllib funny! I find json really verbose when indented , no comments, no multi line strings, “” everywhere hard to read
I also think xml should very much be alive, with better tools. the libraries are quite alright from a use standpoint. It suffered (as did Java, design patterns and agile) from being conflated with the oo architecting and nascent “at web scale” software engineering fumbles of those days.
@mnl 
take: in a lot of ways, XML was and is better than JSON: it has far better tooling around it (schemas, XSLT, etc), and is still at least somewhat human readable… but everyone decided they hated all the angle brackets, so instead we got JSON.
@edmistond my argument. The tooling is rotting... :/
@edmistond @mnl Everybody decided they hated the parse time, if one is being charitable. Everybody decided they hated having to think in trees, if one is being practical. Everybody decided they hated having to follow complex rules when their intent was simple, if one is being unkind.
@graydon @edmistond which is why we have json and json schema and json ld and bjson and oO
@mnl @edmistond Rather inevitably, yeah.
They could have at least have made JSON isomorphic to the subset of XPath maps it can represent.
@graydon @edmistond @mnl "Instead of regarding the obligation to use formal symbols as a burden, we should regard the convenience of using them as a privilege: thanks to them, school children can learn to do what in earlier days only genius could achieve." - Dijkstra
He was criticizing the idea of programming in natural languages. Now that that has arrived, it is clear he was right.
@ravenonthill @graydon @edmistond how is it clear that he was right?
Even as programmers, we program in natural language all the time, in documentation, comments, variable names, rfcs, GitHub issues, planning meetings, etc…
Dijkstra himself was a great natural language communicator, if it weren’t for his cheap quips.
@mnl @graydon @edmistond look at machine learning systems, which the very rich, who resent talent they don't have and can't buy, hope to replace actually knowledgeable people with. They're fine, until they start hallucinating, and then they're not fine at all.
If we want correct software, we have to use precise formal systems.
@ravenonthill @graydon @edmistond I think we’re going to have to disagree on these points :)
