Comment Re:Before someone points out... (Score 1) 32
No. Not on this level. The bio-hardware is wayyyyyyy to bad for that to make any sense.
There may be technical applications, but people looking at it? Just wasted effort.
No. Not on this level. The bio-hardware is wayyyyyyy to bad for that to make any sense.
There may be technical applications, but people looking at it? Just wasted effort.
As usual. How pathetic.
Does not make a difference. You cannot use this to improve anything. You are not fast enough.
If you prefer smaller (!) jumps to larger ones, sure, waste your money. But it is a purely esthetic difference.
That is complete insightless nonsensical crap. If you do not have vsync, you will get tearing, _regardless_ of refresh rate.
Yes. And they are right. You can detect the difference via inference. It is even easy: Just drag your mouse in a circle. But you cannot use that speed difference. Not possible.
On the $3000 network cable, that possibility to detect shrinks down to a look at the invoice though.
We had similar numbers in our 3rd year software engineering class at Uni (CS studies). And yes, we also had the effect of language expressiveness (for lack of a better term) and that the loc/day number is weakly or not dependent on language used.
But you cannot go automatically from a low expressiveness language to a high expressiveness one, as that requires insight and understanding. Sure you can put repetitive code fragments into library calls, but the sources we are talking about here should already have that.
So they are about 4 orders of magnitude off. Nice.
It does sound very much like it, agreed. And thereby they ignore basically all wisdom we already have about using LLMs for coding assistance.
That would be bad. Because a coalition like the one that freed Germany from the Nazi regime is not going to happen with nuclear weapons in play. A fascist USA would have to collapse under its own weight. And while that will always happen, it can take a long time.
Worst-case (for Microsoft) and best-case (for the world) would be they try this, are "successful", roll it out to everybody and then some defect that is devastating (remote exploit, hardware destruction, data corruption, etc.) and bad enough that it needs to be fixed relatively fast is found. But they cannot roll back and cannot fix that anymore because the AI code is (as usual for this type of code) basically unmaintainable and it would take them years to fix things and they cannot even get the people with the skills for it.
That 1M loc/month figure per engineer? Boils down to 0.5 seconds per line-of-code. Realistic numbers are 3 to 5 orders of magnitude lower, especially when the input code is not too clear and there are lots of hard to find dependencies. That gross nonsense made me think that they may not actually be honest about trying this. ON the other hand, it is Microsoft. They really may be this incompetent.
As to "old", yes. Too many coders are too clueless to understand that "old == exceptionally well tested and understood" (may still be bad, see MS code, but at least you know). Actual engineers worth the title understand that. I had my key insight for that when I traced a bug in a common Linux utility a long time ago. Turns out the code had been unchanged for almost 10 years and it was actually a gcc bug. That is what mature software looks like.
Of all the word processors I've used the best was MSWord 5.2a for the Mac LC2 & LC3. LibreOffice Writer comes close. Nothing since from MS that I've used comes close.
It's been a while so I can't really direct you. When licensing became very uncertain I backed away and haven't done one in a while.
My best advice is to get a good flight controller kit up front so everything works together without a lot of screwing around. Also to read lots of build logs before you do one. And maybe start with a cheap type to build familiarity.
Also any design where you just have arms connected to a central board tends to be flimsy. I started with a SK450 and it's kind of floppy
They deliberately destroyed our public lands to make sure we all had to pay to live and you're here for it
If apple doesn't want to be in the middle of every transaction, they are welcome to stop doing that. No one is forcing them to gatekeep everything.
SLI causes problems with a lot of titles and also isn't available on Linux any longer. (There is still a feature called SLI in the Linux driver but it is NOT SLI.)
You didn't explain why they would need more than 60Hz.
A LISP programmer knows the value of everything, but the cost of nothing. -- Alan Perlis