tgtweak

18 ms·

241.
by tgtweak 2y ago
You mean aside from the fact it's ex-post ...
242.
by tgtweak 2y ago
It's a carry over from powerbi actually, separate function entirely.
243.
by tgtweak 2y ago
Csv suffers from similar encoding and platform specific gotchas surrounding newlines and delimiter escaping.
244.
by tgtweak 2y ago
I feel the same way about elastic. That being said I noticed .parquet as an export format option on Shopify recently and an hopeful more providers offer the choice.
245.
by tgtweak 2y ago
Remotefx is a really big feature addition. I still think modern alternatives like parsec are better in almost every regard but RDP is extremely mature.
246.
by tgtweak 2y ago
This is actually very helpful to me! Thanks for posting!
247.
by tgtweak 2y ago
Feels like a catch-up feature to chatgpt... honestly the biggest holdback for me on anthropic is the output token limit on sonnet... 8000 tokens max output is really limited (and 200k tokens in) compared to other offerings - especially cons
248.
by tgtweak 2y ago
This is pretty much the hallmark of antitrust/monopolistic behavior - apple watch (and airpods) are a significant contribution to Apple's bottom line - by restricting these functions under the guise of security or optimization, an
249.
by tgtweak 2y ago
The fact most WHOIS is private these days makes it more or less useless. I think rdap with a request/response authentication on the requestor but that the provider can't mask would be more practical. Also requiring that registrars
250.
by tgtweak 2y ago
There was a moniker for the few people with high speed back then - LPB - low ping bastards. All those fortunate enough to live in a city with adsl or cable high speed in the early days (or gaming at work or university on the T1)
251.
by tgtweak 2y ago
Playstation was in development and was released in japan almost a year before 3dfx released their first card. It's reasonable to draw that someone could have had experience working on the playstation graphics system prior to moving ov
252.
by tgtweak 2y ago
Tribes was the game I remember that showcased the card best - you had a legitimate competitive advantage if you were running this card. (also, shazbot)
253.
by tgtweak 2y ago
I remember upgrading to a creative labs 3dfx voodoo banshee and it was actually stunning - I don't think there has been a generational leap quite as apparent as seeing everything before and after that upgrade. I think I had a matrox c
254.
by tgtweak 2y ago
+fabric But the documentation and every other reference to it will retain the old name.
255.
by tgtweak 2y ago
Not really interested until this can run locally without api keys :\
256.
by tgtweak 2y ago
You can use claude computer functions to actually play it on an emulator with no programming at all - but that kind of feels like cheating :D
257.
by tgtweak 2y ago
The issue with "mining" the rom for shaders is they're not defined in a consistent way across the games. Games using the same base engine could have their shaders stored and referencable in a consistent way - but that would
258.
by tgtweak 2y ago
In those gl/dx games (built for non-specific hardware) all the textures and shaders are compiled either during the game's build OR before you get into the scene. Many console systems, particularly Nintendo, do that precompilation
259.
by tgtweak 2y ago
The Rust criticism is valid - the hidden "breaking changes" have been a hallmark of rust releases and anyone maintaining a serious codebase in rust needs to really be on top of testing and validating if they want to stay current -
260.
by tgtweak 2y ago
There is no official support for x86 in android any more - the Android-x86 project was the last I know that supported/maintained it. Last release was 2022. For apps that use Vulkan natively, it's easy - but many still use and rel
261.
by tgtweak 2y ago
A few things that seem like they're consistently missing from these projects: Hardware 3d acceleration from the host in a version of OpenGL ES + Vulkan that most phones have natively. Lastly, many apps have built-in ways of detecting
262.
by tgtweak 2y ago
https://xyzlabs.substack.com/p/berkeley-researchers-replicat... Given they've reproduced earlier model's and vetted it - I think it's probably safe to assume that these new models are not out of thin air
263.
by tgtweak 2y ago
I'd be really curious about the split in hardware for training vs inference - I got the read that it was a very high ratio to the point the training is not a significant portion of the requisite hardware but instead the inference at sc
264.
by tgtweak 2y ago
I'd torrent it :D
265.
by tgtweak 2y ago
I think they also published the training methodology as well - that others have reproduced, no? The only thing that I'm not sure is their low level nvidia CTX training code was released under the license - but in order for a third par
266.
by tgtweak 2y ago
Wasn't there an internal google email or memo that stated as much as well? That open source was moving faster and more efficiently than the best private teams and that it was accelerating - basically calling this out about 18 months ea
267.
by tgtweak 2y ago
Hype buyers are also Hype sellers - anything Nvidia was last week is exactly what it is this week - DeepSeek doesn't really have any impact on Nvidia sales - Some argument could be made that this can shift compute off of cloud and onto
268.
by tgtweak 2y ago
It's kind of like using an infinite money cheat in a game - it's a guaranteed way to get bored of it quickly.
269.
by tgtweak 2y ago
The sarcasm was abundantly clear in the first bullet point.
270.
by tgtweak 2y ago
I think he's mostly referring to inference and not training, which I entirely agree with - a 4x version of this card for workstations would do really well - even some basic interconnect between the cards a la nvlink would really drive

More ›