@benthor I mean I'd try to change your mind, but I agree with you (or rather the subset of what you mentioned that I understand well enough to have a proper opinion).

@benthor I like IPSec, though most implementation suck.. Decently designed protocol (with too many choices) #changemymind

@hannesm How can IPSec be decently designed if it effectively off-loads parts of the "design" aspect on the administrators?

@benthor no no, first there's the spec. Then there is the implementation where the dev decides what to implement and how to present the config to the administrator...

@hannesm Well, that isn't any better, is it? If the spec doesn't prescribe sensible minimal requirements, how are the implementers to blame?

@benthor: I 100% agree with "". The workflow for the 21st century? I'm esp. impressed & hopeful about 's few-years strategy: about.gitlab.com/strategy/#seq > "BHAG"

Disagree: MacBooks before the "butterfly" keyboards.

Missing IMHO:

@gittaca Wow, I hadn't known that GitLab is apparently one of the more thoughtful and progressive companies out there. I think I'm already a fan from what I've read so far.

@benthor @gittaca I really like gitlab as a product; but the team behind it (and by extension the company that provides the framework for its development) simply is awesome.

I read so much on their blog and genuinely enjoy their policy about keeping their business internals open and accessible. Just check out their public (internal) monitoring: monitor.gitlab.net/dashboard/d

@gittaca Wait, you _liked_ everything by Apple before the butterfly keyboards?

@benthor No, I just disagreed with "anything" & "later" ;-) Both their products I own were made between 2006 & 2015.

@benthor that HTML is based on XML can generally be regarded as a bad choice.

@marble What other good choice was there in 1989? ASN.1? *shudder*

Of course, I personally would have loved s-expressions... But it's unclear if those would have fit the purpose.

@marble json as a technology did not exist for the first ten years of the history of html. Hence it does not count.

@benthor yeah,but i also mean in the current context/time

@marble You wouldn't seriously propose using json as a drop in replacement for HTML though, would you?

@marble I'm sorry, that's not a permissible argument.

@benthor I generally think that most file formats could be changed to json.
I think so for e.g kicad and svg.

XML is just PITA

@marble That is a mere statement, not a proper argument.

(You'll need to show me concrete examples of how encoding markup in json instead of XML is any less painful.)

@benthor for one: HTML is not a regular language and hence cannot be parsed by regular expressions

HTML tags lea͠ki̧n͘g fr̶ǫm ̡yo​͟ur eye͢s̸ ̛l̕ik͏e liq​uid pain

@benthor json is simple like
You got lists and objects and that's about it.

@marble And how is the ability to parse a markup language by regular expressions relevant?

@marble I fail to see how this is in any way relevant to your initial assertion that HTML shouldn't have been built on top of XML.

You can write quick&dirty crawlers using just bash, curl and grep.

@benthor Nice idea, gave it a try.

Failed to come up with such a nice spilt. Too many criteria (privacy, centralization, cost, ownership, mono-/oligopoly risks, ease of use developer/customer end, military potential, and that‘s just for starters...).

If I managed to make up my mind how to weigh the items, a #feelgoodabout might score something like 57:43.

My Apple #feelbadabout is roughly 40:60. Having started in 1984 I claim senile obstinacy and keep buying anyway...

Sign in to participate in the conversation
chaos.social

chaos.social - because anarchy is much more fun with friends.
chaos.social is a small Mastodon instance for and by the Chaos community surrounding the Chaos Computer Club. We provide a small community space - Be excellent to each other, and have a look at what that means around here.
Follow @ordnung for low-traffic instance-related updates.
The primary instance languages are German and English.