I can't be the only person annoyed by this.
Whenever I have a ton of mastodon tabs open in the background, they are *constantly* pinging the server to the point of being a DoS that makes the server stop loading or accepting posts from me.
Surely there is a better way to engineer the mastodon web client so this doesn't need to happen.
The Torvalds I anticipated wouldn't take "I'm not good at <language>" as a reason to let someone else program his hobby project for him, instead of rolling up his sleeves and jumping in.
Never thought I'd see the day when Torvalds himself would become a vibe coder. Is he so out of practice after transitioning to a management role for so many years that this was the easiest path forward? What of the joy of learning in what is supposed to be a hobby?
@gfxstrand maybe a dumb question, but would a virtualized GPU that uses NIR(?) as it's ISA and presents a virtIO device with operations that are typical of most uAPIs be viable?
Something like DRM native context (maybe even built on it), but
- agnostic to the host hardware (by leveraging an IR and thin layer of abstraction over job submission and memory management) and
- more flexible to the client/host driver versions.
@bcantrill @ahl Perhaps https://www.youtube.com/watch?v=hEIoCeqMbWU ought to be revisited some time. 😅
@bcantrill I discovered today I have one of these SIL3114 cards. (Apparently chinese manufacturers are still pumping them out.) Supposedly there are some kernel workarounds for them now, but I think it might be worth looking at other hardware. 😅
Perhaps the Promise 300 TX4? 😂
@bcantrill Is it possible the silicon image data corruption you mentioned in https://www.youtube.com/watch?v=srKYxF66A0c was https://web.archive.org/web/20070621094529/http://home-tj.org/wiki/index.php/Sil_m15w
There was an errata affecting the SIL3112, another (DMA issue!) affecting the SIL3512 and SIL3114, and then also whatever nvidia chipset thing was going on here: https://lists.debian.org/debian-user/2009/01/msg00323.html
@bcantrill "When hardware and software development is divided among different organizations, as it is in the Windows, VMware, and Linux worlds (or, alternatively, the Intel and AMD worlds), exploiting error injection technology for product improvement is much more difficult."
@bcantrill "Of course, companies such as IBM and Oracle that control both the hardware they sell and the software that supports it are best positioned to take advantage of error injection technology to improve the handling, diagnosis, and containment of errors by their respective systems, as having the hardware and software people all in a single organization allows the necessary continuous interaction between them as new hardware and software is developed."
@bcantrill Of particular note:
"Since the e-cache parity crisis, error injection has become a core competency of what is now Oracle's Systems organization. As new processors and their supporting ASICs are designed, error and diagnostic reviews make sure they have the appropriate ability to inject errors into their internal structures, and the error injector is enhanced to inject those errors so that we can test our error handling, diagnosis, and containment software in an end-to-end fashion."
@bcantrill I'm sure many at oxide would find this quite interesting (if one can overcome the ecache traumas :P)
"Injecting Errors for Fun and Profit"
"Error-detection and correction features are only as good as our ability to test them."
@brewsterkahle It's become very acutely apparent that the Internet Archive is an enormous single point of failure. Nearly none of the Archives collections are replicated externally by other orgnaizations. The eggs are all in one basket.
Should the Archive fail for any number of legal, policy, or financial reasons, most of the archived information ever collected will be erased.
@libreleah Do you know if it's possible to back up the vendor BIOS of the T1650 in service mode?
Failing that, do you know what sizes of chip-clip will work if using a pi pico or something?
@lofty eat banna? 🍌
@bcantrill @ahl Spurred by this funny little bit of historical revisionism I stumbled over today, and the recent news of 20% cuts to workforce:
Here's one for the Oxide and friends:
Is intel the new DEC?
The sprawling empire, the marriage to a legacy architecture, the failure to follow a market shifting to smaller computers, and now the canibalization to stem losses and possible acquisition target.
And, in true Oxide fashion: do they share any cultural problems?
@karolherbst I don't know if you're exactly the right person to ask, but do you think that chipstar might be able to pick up the older 32-bit cuda support that nvidia dropped with the 50XX series? Right now the GTX 580 is beating the crap out of the RTX 5080 in titles like Mirror's Edge which have absolutely horrible CPU support in physx. (see: https://www.youtube.com/watch?v=h4w_aObRzCc)
@foone Do you have a readable archive of your old tweet threads anywhere? The links on your wiki to the best ones go nowhere now and I thought they were pretty good. In particular I loved the one about backing up mouse drivers. (https://x.com/Foone/status/1146135405793669121)
I found gallery-dl is decent for getting tweets. The basic shape of the script I used to back up my own was https://pastebin.com/wwVgeFt4 which grabs cookies from firefox. It dumps out media and json files. Not sure how to limit to specific threads.
@connections A few years back the museum received a coin control frame. Have you had the opportunity to restore it to service? Even if not, is it possible you could do a video on it?
@libreleah There's coreboot support for the T1650 with ECC support, but would it be cost effective to extend this port to the T3610? It supports a similar ivybridge CPU and chipset, but a ton more ram. (128 GB or even 512 GB for the T7610)
Seems to me that it's a lot more capable, though maybe more rare/costly to obtain and expensive to ship.