One solution I'm considering is a heavy duty home server (since I want to get into serious astrophotography image processing and machine learning) paired with a really cheap chromebook
Conversation
The server will be a power-hungry-work machine and can be a minimalist linux environment otherwise. Chromebook for mobility, with all writing work done entirely online... basically weaning myself out of the thick desktop world altogether while waiting for someone to reinvent it
1
2
15
So basically:
✓ heavy-lift server
✓ consumption devices (phone/tablet)
✓ browser-typing heavy device (chromebook)
✓ experimental smart environment
✕ ditch thick desktop environment with "apps"
3
11
Maybe the thick desktop as an environment is basically uncanny valley and dead right now. Caught in the badlands between touch-consumption, servers, and browser-native typing.
4
1
11
Replying to
true... I don't. Closest I'll get is video and image processing, which are doable on a server type box.
1
2
eg i am pretty sure has a rig at home for his ml art stuff and then the heavy heavy lifts he buys ec2 time for
2
You’re unable to view this Tweet because this account owner limits who can view their Tweets. Learn more
i bet, like you can just send the media and the batch instructions up and get back finished video
2
In some cases you can, in other cases, the sensor and state-update loops are too tight, and the diffs not sufficiently formalizable to do updates. There are large batch regimes, yes, but also large interactive regimes. Image processing is surprisingly artisan...
i'm just thinking all the 3d animation and video editing i've ever done, rendering the final is a unitary task you could throw tons of cpu at and parallelizes well but the in-situ stuff can be handled fine on a desktop

