I don’t think OpenAI gets enough credit for exposing GPT via an API. If the tech remained only at Google, I’m sure we would see it embedded into many of their products, but wouldn’t have held my breath for a direct API.
Yeah, for all that people make fun of the "Open" in the name their API-first strategy really did make this stuff available to a ton of people. They were the first organization to allow almost anyone to start actively experimenting with what LLMs could do and it had a huge impact.
DeepMind wrote the paper, and while Google's API arrived later than OpenAI's it isn't as late as some people think. The PaLM API was released before the Gemini brand was launched.
Microsoft funded OpenAI and popularized early LLMs a lot with Copilot, which used OpenAI but now supports several backends, and they're working on their own frontier models now.
Yeah and it was Open AI that scaled it and initiated the current revolution and actually let people play with it.
> while Google's API arrived later than OpenAI's it isn't as late as some people think.
Google would not launch an API for Palm till 2023, nearly 3 years after Open AI's GPT-3 launch.
Yeah let's not pretend Open AI didn't spearhead the current transformer effort because they did. God knows how far behind we would be if we left things to Google.
I've never had any WiFi issues, and I think most people who started using Linux in the mid 2010s have the same experience.
Graphic though, yes (I had as much sleep issues on windows than on Linux). Especially the dual graphics intel/Nvidia. I still have to force environment variables to launch my games with the correct GC
Lol. Yes. I would like to say that's not true... and for quite a while my Dell XPS would do all three, but it wasn't a cheap device. I think their driver team isn't supporting my 8 year old XPS anymore as sleeping is... problematic. And power management on linux laptops has always been worse than windows.
But... I will gnaw my left arm off before I go back to Mach or WinNT. (Maybe I'll try using HaikuOS as my daily driver...)
Though... fwiw... I've been running a non-x version of leenucks and then booting into a X and experimental Wayland FreeBSD VM via KVM and it seems to work well. I can halt the machine and save state in about a minute and then turn off the hardware. I come back later and restore. It's not a seamless operation, but I'm happy to live with it. It's also pretty easy to checkpoint the virtual disk before installing the bazillion packages I sometimes have to install to test out various python extensions. So all I have to do is revert to a checkpoint and all that crap is gone. I don't have to worry about remembering which packages I have to manually uninstall.
AFAIK, Microsoft and/or Intel pushed to remove the usual sleep S3 state and use a less sleepy state to be able to access network and display notifications. As if it was a tablet or a Macbook.
This is (of course) badly done, and tested as well the rest of Windows, so it results in laptops waking up in bags, choking thermally, and not going back to sleep.
Just like the other replier, people who put words into others mouths are extremely annoying. And in both of your cases, come off as fanatical. I'd love to run Linux on a laptop (and have tried many times) but have actual work to get done.
Dell XPS? They were pretty good there for modern-ish devices. Not so much for random Inspirons. Lenovo had fairly decent support for their midrange on up models. HP makes crap, so it's unlikely I'll every touch another HP laptop in my lifetime.
But... I think the poster above should have said something like "pick any two (for a depressingly large number of laptop models.)" Also see my post above about what seems to be XPS models falling out of support after eight years or so.
They enshittified/Dropboxified their core Docker Desktop app so much that OrbStack — I believe a single person initially — managed to build a better product. I love this outcome.
Core is not Django-specific, but it has an optional integration. Sync and async, retries/cancellation/etc., very extensible, and IMO super clean architecture and well tested.
IIRC think the codebase is like one-tenth that of Celery.
If you like Procastinate, you might like my Chancy, which is also built on postgres but with a goal of the most common bells and whistles being included.
Rate limiting, global uniqueness, timeouts, memory limits, mix asyncio/processes/threads/sub-interpreters in the same worker, workflows, cron jobs, dashboard, metrics, django integrations, repriotization, triggers, pruning, Windows support, queue tagging (ex: run this queue on all machines running windows with a GPU, run this one on workers with py3.14 and this one on workers with py3.11) etc etc...
The pending v0.26 includes stabilizing of the HTTP API, dashboard improvements, workflow performance improvements for workflows with thousands of steps and django-tasks integration.
We moved all our celery tasks to procrastinate at work for all our django backends since almost two years now and it has been great.
Having tasks deferred in the same transaction as the business logic stuff is something that helped us a lot to improve consistency and debugability. Moreover, it's so nice to able to inspect what's going on by just querying our database or just looking at the django admin.
For those wondering, procrastinate has no built-in alternative to django-celery-beat, but you can easily build your own in a day: no need for an extra dependency for this :)
The "Unix Domain Sockets breakthrough" sounds kinda suspect to me. I'd imagine you should be able to push at the same rate to stdout with a smart MPSC design.
We’ve learned security through obscurity is bad. Not using these will be security through ignorance.
Hopefully it will push us to not only fix individual issues but close entire classes of possible gaps, once P(discovery) gets much higher.