giantrobot

My 2011 Mac mini nods sagely. It runs even better today than when it was new because I put an SSD in it a few years ago.

I own a Sony Magic Link and you are 100% correct. The UI is the worst sort of point-n-click adventure game memes. Not only do you have to guess what visual elements actually do something you need to figure out what functionality they represent. The spacial metaphor is insipid because it takes a lot of taps to get from one "room" to another.

None of this is helped by how slow the Magic Link is. Supposedly the DataRover 840 was much faster but I've never owned one to tell for sure.

The UI of the Newton MessagePad (I own several) is far from perfect but makes much more sense than MagicCap. It also requires fewer taps to reach different functions.

Every once in a while I'll pull out my Magic Link but the insanity of the UI just inspires me to put it back in a box.

Token generation is nominally bandwidth limited. Prefill/prompt processing is nominally compute limited.

For CPU inference on old hardware I don't think q4 offers any benefit over q8 since the AVX unit doesn't support such small floats. I don't even think AVX supports 4-bit int math. IIRC AVX2 does.

Reads off your SSD don't age it prematurely.

Both North Korea and Israel are notable (compared to space) for their dense oxygen rich atmospheres, accessible large reservoirs of water, access to solvents/lubricants, and a generally "shirt sleeves" environment for workers. Asteroids, the Moon, and deep space don't really have these things. So you're not just bootstrapping an industrial facility at those locations but all the basic infrastructure that industry requires to exist.

With space manufacturing the first widget out of a factory costs trillions of dollars. There's also few if any raw materials that are far more abundant in space than here on Earth (at orders of magnitude smaller cost).

It is and it doesn't. You only get into disk writes if the system starts paging out to disk.

I have a retail copy of Photoshop 7 I bought decades ago with hard earned money. A full copy of Photoshop then was hundreds of dollars. I still use it to this day.

I learned Photoshop well enough back in the day to do cool stuff and make a few bucks on the side doing graphic design. I'm not a professional graphic designer at all so it never made sense for me to get on the version treadmill. Even today it does everything I need and more.

Adobe got paid for the work they did. I've gotten incredible value out of my purchase since I've used it so long and the cost has amortized over decades.

Subscriptions imply change. I don't necessarily want change. If I buy something that suits my needs...it already suits my needs. Adding more shit isn't necessarily a boon. Changing the behavior or look with a new update is infuriating. This is especially true of complex software that takes time to learn.

For a developer selling a subscription the strategy of locking up data is entirely rational. They'll keep getting money no matter what they do because they're holding my data hostage. It's a model I have no interest in.

Even better is when they decide to shut down or get bought out. If the subscription shuts down there's no amount of money I can pay to get it back.

I don't know what you mean, I mean I made a million dollars with only a small hundred million dollar loan from my dad! This sort of entrepreneurship is available to anyone in America.¹

¹ Terms and conditions apply.

It is astonishing to me just how bad end user software has gotten over the past twenty years. It's just goofballs. It's not like that point was when software started requiring more resources, the "What Andy giveth Bill taketh" aphorism was coined in the 90s after all. Software gets complicated over time. Every bug fix or added input validation or whatever adds code. That's fine and appropriate. Convenience and integrity add code/complexity. That's not what I find so bothersome.

I find 2006 an interesting horizon. It was then that the average home computer massively jumped in processing power compared to the software running on it. For the vast majority of users the computer spent more idle time waiting for input than users waited for output. Of course for some users there's never enough power but for most all of their tasks were effectively instant. Even heavyweight (for the time) stacks like Java ran incredibly well. He'll even Emacs could run well!

Then the curve seemed to invert. Hardware kept getting more capable with even faster CPUs, more cores (on common consumer machines, and more RAM. As the article points out a "lightweight" Linux DE with native apps really flies on such hardware.

But more development seemed to move to the web. More JavaScript required more powerful JS engines and those are up more and more memory. More shit loaded into the DOM means more and more objects on the heap with more pointer chasing.

Modern stacks are really only fast because modern computers brute force their way through them. A simple CRUD type task that fly as a VB or Delphi (neither stack performance kings) app on Windows 2000 now requires a 2GHz dual core CPU with 4GB of RAM as a baseline thanks to it now being a web app.

Using a twenty old machine with native applications and the CLI feels to me like a super computer compared to the computers I first used (Apple IIs). A Core 2 Duo is a stupidly powerful CPU for most tasks. If you can get by with a command line workflow even a Core 2 Duo is crazy fast.

/soapbox

Except no one will be able to afford those laptops. The hyperscalers will buy up literally all the RAM and high power GPUs. So you'll be stuck with 8GB of unified memory on anything affordable. A laptop of better capability will cost 50% more than today (at least). That won't be your biggest problem because some C-suite goon with stars in their eyes laid you off because of AI.

The current administration has repeatedly demonstrated they do not feel constrained by laws or the Constitution.

Which requires them to explicitly ask your age outside the bounds of qualification for a job (over 18 etc). Which ends up opening them to age discrimination lawsuits.

You'd be right if ISPs didn't do mass DPI and modification of pages to inject their own ads and trackers. They also routinely hijack and intercept DNS responses. They absolutely cannot be trusted carrying unencrypted traffic.

People are bad with computers because "Designers" have been objectively value negative for decades. Instead of paying attention to the reams of scientific data that tells you exactly how to make a UI better, they smoke crack and make things transparent.

I've run into a lot of "UI designers" that were just graphic designers that lucked out getting a job in a software company. They design and compose the UI for static screenshots like they would do a 2D graphic composition. Far too often they'd hand off Photoshop PSDs they would expect a developer to turn into an actual UI. They don't follow a HIG document or respond to any pushback with "you have to know when to break the rules" (they do not know when to break the rules). A good UI designer is worth their weight in gold. In my experience most are worth their weight in very low grade playground sand.

Big whoopsies on that one.

Shh, don't remind them.

Oh, and enabling human traffickers.

Also the genocides.

Flash as an animation tool vs Flash as a web development tool are two very different discussions. There's overlap and they're related but still separate.

Flash was very interesting as an animation tool because it was at its core a vector drawing animation tool. All the scripting components were icing on the cake allowing automation of things that's incredibly tedious in traditional animation.

The fast vector animations made Flash very useful for web distribution. That's all well and good. But as it became the go-to for interactive websites the structure of Flash was antithetical to the web. Deep links became meaningless and content became locked behind Flash. Adobe was also a terrible steward of Flash and only put effort into Flash for Windows. Every other platform was an also-ran for them and the Flash experience was terrible.

The security of Flash was a bad joke on top of it all.

As an animation tool and delivery vector for interactive content beyond the abilities of browsers of the era it was useful. As the front end of the web it was an awful mess.

I set up an isolated network on my LAN with its own WAP to play with my old devices that don't support WPA. I don't leave it on all the time and the network segmentation limits any blast damage. Works well since I have so much old crap with early WiFi.

All those thinking tokens wasted on being an asshole wasted a lot of electricity.

Nah, it'll never catch on. We don't have the technology.

Including but not limited to dual gigabit Ethernet (on a PCIe bus), NVMe slots, and a capable power supply is included in the box.

I lucked out having bought a few N95 mini PCs a few years ago. They were even cheaper then with 16GB of RAM out of the box. To me they're vastly superior to the various RPis I replaced.

I sold off my Pi 4s and never bothered with the 5s. I kept my mix of older Pis for projects that need GPIO and of those my Pi Zeros are the ones that really get used.

Online services most assuredly billed by the minute. He'll, AOL had a huge marketing campaign offering "free" minutes for new customers. You might also have charges from the phone company but those were independent of the online service charges. It wasn't until the late 90s the major online services went to flat rate "unlimited" plans.

For me and most people I knew at the time, VHS didn't have a noticeable quality loss over broadcast unless you were watching LP/EP recordings.

Many TVs people already had in the 80s didn't have RCA connections so VCRs were connected via twin lead to F connector adapters. They had the same noise as the antenna or cable input. So your commercial tapes usually looked about as good as broadcast. If you actually read the instructions with your VCR to set the timing correctly recorded broadcasts in SP mode also tended to look pretty good.

In absolute terms the VHS video was worse than the original broadcast but on the TVs we had it was hard to notice.

This definitely changed through the 90s. Larger and brighter tubes made the deficiencies of VHS more noticeable. Moving to cable TV from antenna was also very noticeable and made VHS quality more apparent.

If you happened to see a LaserDisc video as a comparison to VHS then the quality difference was stark. As much as VHS and DVD by the late 90s and early 00s. However I think that direct comparison was out of reach for most people.

While DDR3 is cheaper it's still tripled or quadrupled in price over the past two years. I just bought a pair or sticks for an old Mac Pro and it was 4x what i paid just a few years ago.

This assume Claude's price doesn't change. Which isn't a great assumption considering inference providers are moving to usage based billing. Also the VC money isn't going to last indefinitely. Current inference providers are being subsidized with VC money at this point.

"e.g." == exempli gratia == to give an example

"i.e." == id es == that is

SmartMedia was a pain in the ass because the 5v and 3.3v cards weren't immediately obvious. You could spend a lot of money on a card only to have it physically fit but not work in a device or reader.

With CompactFlash and MMC (and SD) if they physically fit they generally worked.

You can try 'sudo mdutil -i off /Volumes/microSDCardName' which will disable Spotlight indexing on it. While a Spotlight index of an empty disk shouldn't cause overheating...there's bugs somewhere in the chain. To say nothing of quality issues on the cards themselves. I often use an app called Disk Arbitrator[0] to mount disks read-only or block automounts when I'm dealing with SD cards because of overheating issues.

[0] https://github.com/aburgh/Disk-Arbitrator

I've had similar issues with microSD in a variety of adapters. I think the core issue is most microSD cards just aren't built, in terms of thermals, for sustained writes. In most devices they're either read-heavy loads or burst-y writes. When you stick them in a warm laptop and do a lot of writes they overheat and start throwing errors.

For many cards their drive controller might advertise and support higher UHS speeds the Flash memory is likely the cheapest silicon that can just barely pass acceptance tests. When I encounter cards that fail sustained writes I've had good luck using pv's (pipe viewer) rate limit. I stick it between dd invocations. This has worked well when writing OS images for Raspberry Pis onto cheap microSD cards. They're fine in the Pis but would fail trying to write OS images.