Thursday, November 04
Daily News Stuff 4 November 2021
Can't Code For What Edition
Top Story
- Gen Z isn't going to solve your tech skills crisis. (ZDNet)
Our most recent hires have been Gen X because, yeah, Gen Z can't code for shit.
Tech News
- Australia is preparing for the launch of two lunar rovers. (The Conversation)
One in a partnership with Japanese and Canadian companies, for launch in 2024, and one in partnership with NASA, for launch in 2026.
The automated rovers will be tasked with searching the lunar soil for indicators of water. We know it's there, but we don't know much about how it is distributed.
- Minisforum has a new mini-PC based on the AMD 4700S. (WCCFTech)
This chip is a recycled PlayStation 5 processor with failed graphics cores, and... It's not completely terrible. Memory is soldered onto the motherboard though, because the PS5 has no provision for anything else.
- Start11 is out of beta. (Thurrott.com)
You still probably don't want to upgrade to Windows 11, but if you buy a new system new may be stuck with it.
- AMD's Zen 4, Zen 4D, and Zen 5 are on their way. (WCCFTech)
Zen 4 is the big update due at the end of next year, with the new socket, and DDR5 and PCIe 5 support. Zen 4D is an alternate version of Zen 4 that packs 16 cores onto each chiplet rather than the current 8, at the expense of cache size. Which might mean a 32-core chip for standard desktops, but might also find its way only into servers.
Zen 5 is the next big iteration after that, and will come in a mix-and-match configuration - one Zen 5 die and one Zen 4D die. So 8 of the fastest cores available and 16 cores that are merely very good.
Zen 5 is expected at the end of 2023 - a much shorter schedule than Zen 4. There was originally planned to be a Zen 3+ out around now, but most versions of that got cancelled to focus resources on Zen 4 and Zen 5.
- The McRib is now an NFT. (Yahoo)
Can't eat NFTs. Well, I have celiac so I can't eat McRibs either, so for me there's not a whole lot of difference.
Posted by: Pixy Misa at
06:31 PM
| Comments (9)
| Add Comment
| Trackbacks (Suck)
Post contains 392 words, total size 3 kb.
1
MLID, like AdoredTV, is a questionable source at best. So this could be true, but I'd wait for a more reliable claimer.
Posted by: Rick C at Friday, November 05 2021 12:50 AM (Z0GF0)
2
Yeah, everything that comes through WCCFTech has an invisible disclaimer attached. This does tie in with other sources suggesting much higher core counts coming to Epyc, and with the limited data available on Zen 5.
I didn't see anything that set off obvious alarm bells, but take it with a pound of salt.
I didn't see anything that set off obvious alarm bells, but take it with a pound of salt.
Posted by: Pixy Misa at Friday, November 05 2021 01:51 AM (PiXy!)
3
"Yeah, everything that comes through WCCFTech has an invisible disclaimer attached."
Yeah, but this one needs a double disclaimer, because they cite Moore's Law is Dead, which is the WCCF of "leakers".
Yeah, but this one needs a double disclaimer, because they cite Moore's Law is Dead, which is the WCCF of "leakers".
Posted by: Rick C at Friday, November 05 2021 06:36 AM (Z0GF0)
4
Maybe they aren't interested in learning to code because they've seen too many laid off to make room for H-1Bs?
Posted by: Frank at Friday, November 05 2021 09:22 AM (rglbH)
5
I would suppose H1Bs can't code for shit, either, but at least they cut'n'paste from stackexchange at a lower pay rate than zoomers.
Posted by: normal at Friday, November 05 2021 09:29 AM (obo9H)
6
I'm looking forward to the intersection of stackexchange copy-paste and the new Unicode code-masking trick. Seems like a quick way to find out which of your contractors are actually writing their own sloppy code, and which are injecting it from random web sites.
(TL/DR: add commit hooks to reject code that includes any of U+202A, U+202B, U+202C, U+202D, U+202E, U+2066, U+2067, U+2068, U+2069)
-j
(TL/DR: add commit hooks to reject code that includes any of U+202A, U+202B, U+202C, U+202D, U+202E, U+2066, U+2067, U+2068, U+2069)
-j
Posted by: J Greely at Friday, November 05 2021 10:10 AM (ZlYZd)
7
Back in the '90s, a contractor submitted his code for a peer review - with all the test cases looking good....on paper. I did something he probably never ran into before, I ran the code. Or at least tried to, it blew up all over the place and never successfully ran. That peer review was VERY short 
Posted by: Frank at Friday, November 05 2021 03:03 PM (rglbH)
8
Heh. When I was working at the university, two quite technical managers asked to review the code that someone had been working on for quite some time (a multi-platform GUI hypertext system that looked astonishingly like the first web browsers that appeared a few years later). They walked into the office to find reams of paper on their desks, a complete printout of the server and each platform's client.
They dutifully started working their way through the stack, starting with the server code (which they both had worked on before it was handed off to this person). By the time I arrived, they were frazzled and frustrated and greeted me with a shout: "J, you played with Waiter recently; What Does This DO?!?"
I looked at the page they were stumped by, immediately recognizing it. "It's supposed to be an index search with wildcards, written without any library calls, but it's actually dead code that none of the clients uses. The real search is like three pages down."
The programmer in question was part of a package deal where a research professor agreed to come to OSU if they found a job for his wife. She was just assigned to us one day, for lack of anywhere more sensible to put her. After this debacle, she was moved over to the AI team, where she was put in charge of maintaining the code that their primary funding grant depended on. A few months later, I found one of their team camping out in the lab surrounded by caffeine as he frantically worked to undo her damage before they lost their funding.
-j
They dutifully started working their way through the stack, starting with the server code (which they both had worked on before it was handed off to this person). By the time I arrived, they were frazzled and frustrated and greeted me with a shout: "J, you played with Waiter recently; What Does This DO?!?"
I looked at the page they were stumped by, immediately recognizing it. "It's supposed to be an index search with wildcards, written without any library calls, but it's actually dead code that none of the clients uses. The real search is like three pages down."
The programmer in question was part of a package deal where a research professor agreed to come to OSU if they found a job for his wife. She was just assigned to us one day, for lack of anywhere more sensible to put her. After this debacle, she was moved over to the AI team, where she was put in charge of maintaining the code that their primary funding grant depended on. A few months later, I found one of their team camping out in the lab surrounded by caffeine as he frantically worked to undo her damage before they lost their funding.
-j
Posted by: J Greely at Friday, November 05 2021 03:22 PM (ZlYZd)
9
Sounds a lot like one gal who worked on a neighboring team at one location I worked at. She had a coffee cup with the words "Of course women don't look as busy as men, we do it right the first time".
We also had gals who really knew what they were doing, so it averaged out.
Posted by: Frank at Friday, November 05 2021 04:29 PM (rglbH)
56kb generated in CPU 0.0149, elapsed 0.1308 seconds.
58 queries taking 0.1204 seconds, 360 records returned.
Powered by Minx 1.1.6c-pink.
58 queries taking 0.1204 seconds, 360 records returned.
Powered by Minx 1.1.6c-pink.
