5,392 karma · joined October 31, 2011
It's like saying "steal from artists, but only the good ones"
In the days of these massive AI harvesting projects publishing anything on the web feels like just feeding the machine that will take your livelyhood in a few years.
That may be true around the water cooler. Try publishing any result that goes against a peer-reviewed paper (and, as an author, not be an affilieated with a well-known US university) and you quickly realize that published results are pretty much settled in stone.
I'm not talking about ground-breaking stuff. Something on the level of "after lots of careful work in our group we were unable to replicate the results published in the relatively well-cited paper in a very niche topic" is largely unpublishable.
Sn/Cu/Ag solder has a narrow window where the solder is melted and wets pads but the FR4 substrate isn't yet degrading. If you have sensitive plastic parts like connectors, etc. the window is even narrower. This is doable in an industrial reflow owen. However when doing hand work or repair, it's impossible to control the temperature this well.
For example, after years of experience, I'm still unable to desolder large BGA or PGA parts without visibly degrading the board (i.e. there's usually a brown mark left on it - unfortunately the boards I commonly work with these days use white soldermask, which makes this even more obvious). That's with a temperature controlled pre-heater and a hot-air station. I won't even attempt to solder such parts manually with Sn/Cu/Ag.
A quick search shows it's about 3 times the price of common Sn60/Pb40 lead solder in small quantities: 0.3 EUR/g vs. 0.1 EUR/g.
Typical lead-free tin-copper-silver solder you find in factory mass-produced electronics is hard to work/rework manually because of its high melting point. That is the reason most hand-made and low-volume circuits still use lead solder.
If it touched at any time any kind of confidential data or algorithms it was probably shredded into tiny pieces.
I don't know about Croatia, but usually such a change in service allows you to break the contract on your end without any penalty (as in, end subscription right there and not pay the rest of the 2 years).
It might be hidden in the fine print though and they probably don't want you to know that option exists legally.
I'm often wondering if those are some clever trick in the map or if the map is running on some newer variant of the Doom engine that has features not available in the original.
Sometimes they do explain it in the video, like in that one map that had a randomly chosen math question the player had to answer by pushing a button. I wish they would do that more often though.
A few years ago when I was working on a project, the factor was in low two digits. It was a no-brainer to choose micro B over C.
I wonder how this works. How do you safely check for an instruction that "locks ups unrecoverably" on some hardware? Will Linux hang on boot on such systems? That is still better than hanging randomly later, but I can see why Microsoft didn't want to go for that approach.
EDIT:
The check in Linux literally just called hlt a bunch of times and hoped it didn't crash. There was no recovery in case it failed (note how there's no code path to print anything else than "ok"). Searching for "checking hlt instruction" you can actually find people asking on forums why their boot stoped at that point. The check has been removed in recent kernels, so I had to go back in history a bit to find it.
https://github.com/torvalds/linux/blob/521cb40b0c44418a4fd36...
As soon as it's worthwhile, there will be RSS spam targeting search engines.
Many companies already have near zero-useful-content blogs because having blobs of text that search engines consider fresh increases their ranking. If RSS-based search becomes a thing, you'll get that kind of content pushed to it.
Or, if that doesn't work because of feed moderation or something, expect getting sponsored entries in your favorite indie blog's feed. Sponsored content is already a thing, but mostly to get "yes-follow" links. Now it will be to get content pushed to RSS search engines.
To me, the peer reviews look like something you might get from a reviewer today, but I'm certain the Einstein's performance review linked on the very top is fiction.
It seems the designers were aware of this problem, because it has a key lock feature. Unfortunately it's hillariously ineffective. To activate/deactivate key lock you need to touch two buttons simulateneously. They forgot however that in practice it's almost impossible to activate two buttons at the exactly the same time. So the machine in most cases interprets this as pushing two buttons separately, usually further messing up some setting you wanted to protect by keylock.
Mozilla the non-profit owns the for-profit corp who owns Thunderbird.
In their own words from the footer of https://www.thunderbird.net:
> Thunderbird is now part of MZLA Technologies Corporation, a wholly owned subsidiary of Mozilla Foundation.
Interesting idea to have separate parametric and functional test indicators. I've made a few custom production test setups in the past years (nothing as nearly as complicated as the one described here) and always went with the greatest simplicity possible - start button and a single red/green go/no go, exactly to avoid the kind of problem described later in the article.
I can see why they would do that though. When reviewing test runs, it's always the parametric tests that give the most false-positives. If a parametric test fails, the part functions as intended, but some measured parameters are outside of the designed range. Parametric failures are often not failures in the parts themselves, but failures in the test procedure (e.g. bad connections to test points due to dirt or worn out contacts in the fixture, etc.).
A separate indicator for parametric failure would give the operator an early indicator that excessive failures might be false positives and re-test the parts, or clean or exchange the test fixture.
I fail to see how the fact that the Latin alphabet has upper and lower case is to blame for this. This is more about the coding standards the author has to abide to than anything else.
If your coding standard would require you to name the property MMAXTEXTURESIZE instead of mMaxTextureSize, would there be any less busy work?
I understand that not having models in a centralized repository brings problems with discoverability. So maybe a better project would be a search engine for 3D objects hosted on the long-tail web?
The recommendation against using multiple connectors on a single board seems to be largely ignored in the wild though. I've seen similar products that use two or even three DF40-series connectors on the same PCB. Here's one example from the top of my head:
https://www.variscite.com/product/system-on-module-som/corte...
That said, I can understand why they went back to the old form factor. The new high-density connectors required for CM4 integration are a single-source item, don't have any alternatives [1], and are out of stock everywhere [2]. On the other hand, the older SO-DIMM connectors are a commodity item that can be sourced from a number of manufacturers.
The SO-DIMM connectors are also much more convenient and robust for removing and re-attaching the CM. Raspberry Pi makes it relatively hard to flash the eMMC on the CM in-system if you include any USB periphery on-board. Because of this a production process will generally include the CM being placed at least once onto a dedicated flashing board before being placed onto the final device.
[1] https://forums.raspberrypi.com/viewtopic.php?t=325885&sid=71...
[2] https://octopart.com/search?q=DF40C-100DS-0.4V%2851%29&curre...