I urge you to reconsider. Potential customers and applicants will immediately go all in and the "but this doesn't make any sense" step will never happen. We'll be reading blog posts like "I'm hooked on PCP (the Post-Cabling Paradigm)" from Google SRE's with comments praising the downfall of "Big Ethernet" while the emissions from datacenters vaporize flocks of migrating birds.
I know having a redundant server is better - but there's something to it.
Also, this reminds me of a post I read a while ago about them moving a server from one building to another without unplugging it or something.
[Found it: https://news.ycombinator.com/item?id=24059243]
obviously a second server and a reverse proxy or something would be less jank, but your idea definitely has merit. it's kind of like using two points of contact while climbing something.
The solution is obvious: Qi inductive wireless charge coils on the outside of the server cases.
If you don't want to worry about piping diesel around your server farm, we could go with compressed air and use air turbines with generators on them. Clean and efficient energy transmission without wires!
This is why some of user facing programable stuff pre node gives the users a interface like excel.
(XKCD might have been inspired by the same post since it came later)
I didn't realize that until I accidentally bumped my track pad.
Sometimes when things fit a little too neatly above the fold, it seems like there's nothing below it.
> The IDE may seem like a joke, but the research and development gone into it is most definitely not.
> MS Paint IDE uses a custom OCR that can be trained by nearly any font in any size, reading code from images in just miliseconds. The IDE's compiler and program output are in .png's, and automatically underlines any errors present and provides syntax highlighting with Java, Go, Python, and JavaScript currently being supported.
> The IDE also allows for arbitrarily editing any text file through Paint via a context menu option, with native-looking button overlays onto MS Paint to perform IDE actions.
why?
I’ve built bizarre things that no one will use purely to learn about things.
Many projects are about the process itself and less about the outcome. Every one of those “useless” projects has informed my productive work throughout my career.
But seems like a huge amount of time sunk into something that I can't imagine anyone would actually use for more than a couple of minutes.
I'd love to see a project on github that mines HN posts for a decent sized list of this.
I'll start-- off the top of my head I remember a post from Mike Hearn about starting the web from scratch using a format that announces the size of the incoming message rather than making the recipient check for terminator(s).
There must be a dozen dozen such thoughts-without-software of "the web, but good" on here.
If we have a list then you could plug it in to a script and have a spit out a random set, and then you could focus specifically on those.
Anyway, I'm thinking it should be stored and maintained as a quine...
Also, I should clarify that "brilliance" and "stupidity" in this theory are not raw intelligence, but the application of said intelligence.
(https://wiki.ms-paint-i.de/features)
Looks good to me. Solid product and a solid roadmap :-)
Every bank seems to accept the weird single line that gets drawn when I try and scrawl my signature using my trackpad, so there must be quite a bit of interesting magic going on there.