Coolify is your best option here. It has preview environments and lots of other built-in features - just the cost of your server if you're savvy enough to get things setup yourself. https://coolify.io/
While I agree that avoiding/ignoring politics isn't helpful to anyone, it still doesn't have a place at work. My view is, people are going to disagree on politics, and therefore it just gets into a debate, or worse, an agrument at the office or in chat and makes the whole situation more ugly than the manager and/or employer wants to have to deal with.
Why would you limit the tool to strictly be for React?
"As per my limitations, I am designed to work specifically with React and TypeScript/JavaScript only. I cannot provide direct conversions to plain HTML/CSS or other frameworks."
Mainly to narrow the problem space: there's a lot of logic in our backend that we call "post processing" which is cleaning up the code after the AI generated it to
fix hallucinations. For example, the AI often gets import statements wrong or misspells icon names.
It's pretty interesting because every model seems to introduce a new set of hallucinations, so this is a problem that requires a lot of maintenance. We have this internal mantra to "use AI as little as possible" especially it's a problem that can be solved deterministically.
Also, because we are more focused on PMs, designers, and product leaders the code is largely an implementation detail. What they care about is being able to visually communicate their ideas, and React just happens to be a great way to do that because LLMs are great are outputting it + we have a pipeline to render it. (We are React developers ourselves.)
I'm curious, if you have that philosophy (which makes a lot of sense), you must have considered building is a sort of more abstract (but extensible) UI toolkit language and library and you could code in that then compile it down to React? Or have you found the benefit of large LLMs already having detailed React training is just too high?
> have you found the benefit of large LLMs already having detailed React training is just too high?
This.
We've tried a lot. We have been around since Oct 2023. First tried fine-tuning, but it's very hard to teach an LLM something new.
At one point, our product lived in a Figma plugin (https://x.com/Teddarific/status/1729153723728011618). To do this, we had the LLM output JSON, and then converted that to Figma nodes. This is sort of what you're saying. But the big issue was it would hallucinate many things and was really only good at the examples we fed it.
Ouch. Right in the feels. Americana nostalgia, landscape of broken dreams song plays in background, fluro lights flicker on and off in the abandoned mall, an empty parking lot, Twinkies wrapper tumbling in the wind, somewhere static from the last analog TV purrs then switches off forever... Ask not for whom the generational bell tolls...
Second that about looked older... Maybe it was the super8 video quality
> Three of the four arming mechanisms on one of the bombs activated, causing it to execute many of the steps needed to arm itself, such as charging the firing capacitors and, critically, deployment of a 100-foot-diameter (30 m) retard parachute. The parachute allowed that bomb to hit the ground with little damage.
"Incredible as it may sound to a civilian, Hicks said he spent no time worrying about the thermonuclear warhead. He had been convinced by his training that it was nearly impossible to detonate a warhead accidentally. Among other things, he said, the warhead had to receive codes from the launch-control officers, had to reach a certain altitude, and had to detect a certain amount of acceleration and G-force. There were so many safeguards built in, Hicks later joked, that a warhead might have been lucky to detonate even when it was supposed to."
Some folks who knew more than you and I seemed to think there was potential for a nuclear disaster.
> In 2011, Lt. Jack Revelle, the bomb disposal expert responsible for disarming the device, claimed "we came damn close" to a nuclear detonation that would have completely changed much of eastern North Carolina
> In a now-declassified 1969 report, entitled "Goldsboro Revisited", written by Parker F. Jones, a supervisor of nuclear safety at Sandia National Laboratories, Jones said that "one simple, dynamo-technology, low voltage switch stood between the United States and a major catastrophe", and concluded that "The MK 39 Mod 2 bomb did not possess adequate safety for the airborne alert role in the B-52".
