4 ms·
I had a similar impression of it, the best I can come up with is I think it was a bet based on a prediction of the future that is kinda true, but just doesn't m
by gonehome 4y ago
I had a similar impression of it, the best I can come up with is I think it was a bet based on a prediction of the future that is kinda true, but just doesn't matter as much as they thought.
1. Personal computers are dead and just a means to access a browser via dumb terminals
2. Applications will be Figma style - web based entirely.
3. Intel/AMD chips are bad and not getting that much better (this one is true, but missed the mark entirely because of Apple silicon throwing a curve ball)
4. Given that your personal computer is basically a dumb terminal to a web browser you could imagine a world where it doesn't matter and it's worth paying $35/mo for a super computer to run your web apps streamed to whatever local hardware you're using, since your local machine is basically useless as a machine and only serves to load a web browser (which is where your actual applications are). This super computer will also remember state so any devices you access mighty on is exactly as you left it. Long term, this could end with a Chromebook like devices of some sort that just connects to mighty and are really nice low power hardware.
--
Personal bias aside (I hate this future and I'm actively working on software trying to bring personal computing back in a from first principles networked way) - it was a bet on trends and fits squarely into the PG concept of a good startup being a bad sounding idea that might have recently became good because of changes in the environment that have not yet been really recognized. I think Apple made it less relevant. I also agree there's a weird issue of who is the actual target market at that price point (the exact people that would maybe need it can also afford good hardware).
That and there is obviously fertile ground in AI right now and the founder sees a bigger opportunity there with a lot of low hanging fruit.
- hn_throwaway_99 4y ago> Intel/AMD chips are bad and not getting that much better (this one is true, but missed the mark entirely because of Apple silicon throwing a curve ball) Define "bad". I'm on a 2019 MB Pro with an i9 with 32G of ram. I pretty much never have browser performance issues that bother me. I have Chrome open with 2 windows and a total of 39 tabs, plus some other apps that people always like to bitch about resource-wise (Slack, VSCode). Granted I'm not a designer so I don't have what would be some of the most resource-hungry apps like Figma, but if there is one group of people who I've always seen spend a shitload on their desktops/laptops it's designers. I think Mighty was always a weird project that pretty much only solved a problem for people that would never be willing to pay for it.
- wmf 4y ago2019 MB Pro with an i9 with 32G of RAM You know that's a very expensive laptop, right? Like more expensive than 99.9% of all laptops. Of course you don't have a problem. I think Mighty was always a weird project that pretty much only solved a problem for people that would never be willing to pay for it. Amen. If it could have been, say, $10/month or less it would have been great for people who can't afford a fast computer.
- icedchai 4y agoExpensive, but basically obsolete now, thanks to the M1. I had a 2019 MB Pro (i9, 32 gigs RAM) as a dev machine. It was terrible: fans running constantly, throttling. My M1 Pro machine is soooo much better.
- hn_throwaway_99 4y agoYes, I know it was a high end laptop at the time, but you can get refurbished ones now for about 800-900 dollars, putting it squarely in a mid-range laptop market. Would be even cheaper with, say, 16G of RAM, and still be totally capable as long as I just closed some unused browser tabs a bit more frequently.
- Kwpolska 4y agoYou are overblowing the Intel/AMD/Apple Silicon thing. Intel/AMD CPUs that aren’t complete cheapo garbage have more than enough performance to run all those fancy apps.