I remember John Cohn, an IBM Computer Engineer, was on some TV show called The Colony and built one https://www.youtube.com/watch?v=bkH6mFlfH3o and I seem to remember it getting much further than this clip.
Block mode terminals are somewhat similar to the forms based UIs we later got like HTML.
The 3290 was like a 4 split tmux session, it is such a beautiful device (https://ifdesign.com/en/winner-ranking/project/ibm-3290-info...). Perfect keyboard, plasma contrast and the right text color for long sessions. My understanding is they outlived themselves with long service in air traffic control (supplementing vector/raster displays), financial markets, and software development. Once upon a time people seemed to really care about doing things well.
IBM terminals were the original stateless clients. :)
I had a PS/2 Portable 70 for a while. It was a luggable Microchannel PC with a orange plasma display. Going through airports with it was a hassle as it didn't have batteries - I would have to find an outlet to power it on for airport security inspection.
It does turn out that 3270 streams and web forms are similar. Transmit a form, wait for the user to fill in the fields on the form, then receive the fields when the user submits it. It is how a mainframe in the 1970s/80s can support 100s of terminals, despite being small by modern standards.
It's a portfolio, the guy did things that shipped in millions of units.
It was a good design, in critical environments small goose-neck lights are typically used, even if the buttons and indicators are backlit. I still prefer it to backlit keyboards which, with decreasing thickness, increasingly compromise the quality of the keyboard (I have tested this with later chicklet models running through all available keyboards for a few era ThinkPads). Having a small amount of ambient light is easier on the eyes, especially the yellow-shift on the X40, and you can run the screen at minimal brightness to increase battery life and whatever health benefits this all has.
In the US, the brand was not as prominent as elsewhere in the burgeoning PC industry but the AT&T 6300 series were OEM systems. Built like a brick shithouse and not too pricey for a reto PC.
AT&T also had very strong design in their product lines. The terminals that descended from the Blit, with their square (1:1, not 4:5) CRTs was a thing of beauty. It crushed me to know how few remain and were not responsibly recycled.
As a career, we need to be better at preserving our history.
I am curious about the inverse, using the dataset layer, to implement some higher level things like objects for an S3 compatible storage or pages directly for an RDBMS. I seem to remember hearing rumblings about that but it is hard to dredge up.
Main issue with opening it further is lack of DMU-level userland API, especially given how syscall heavy it could get (and iouring might be locked out due to politics)
There was some work on this presented at one of the OpenZFS summits. However it never got submitted. Not sure if it remains a private feature or if they hit some roadblocks.
In theory it should be a pretty good match considering internally ZFS is an object store.
For RDBMS pages on object storage - you might be thinking of Neon.tech. They built a custom page server for PostgreSQL that stores pages directly on S3.
Man. I don't actually know anyone who vapes. I see it in public sometimes and just assumed people refilled them - maybe they do. Seeing him hold some up, seeing all that plastic, metal, electronics, all that Work (Joules) expended, in something that you just dump after a day is nuts. I can't think of anything else like that. Maybe plastic water bottles but they don't have even half the materials or complexity? Maybe I under-estimate how much is put into regular cigarettes or beer & cans.
Refillable vapes used to be the standard around a decade ago, back when a liter of vape base (without nicotine) cost 30€ at max. Disposable vapes pretty much didn't exist. Now the same liter of vape base (still without nicotine) is a "tobacco product" and costs 400€+ due to taxes thanks to decade-long lobbying efforts by big tobacco, turning refillable vapes into a massive niche product due to single-use vapes costing the same or less, without any of the hassle of mixing your own liquids or having to refill them.
Yes, 100ml from the same store I bought a liter from for 40€ in 2015 now cost 56€. There is currently a tax of 0.32€ per ml on liquid, no matter if you're buying the base or the finished product.
I guess it's time to not only make the vaping liquids yourself, but the bases too :P
Quick Googling suggests I'd pay less than 30 euro for a liter of VG. What about for the nicotine concentrate? I'd pay about 20 euros for 100ml of 20mg/ml concentrate.
A little calibrating correction: A vape should last more than a day unless you're a very heavy user. Around three days with a '700 puffs' one maybe, and a week wouldn't be unheard of.
He put a fuse on every individual cell and on the overall unit, so I would say he was reasonably cautious (although he deployed a bunch of high-voltage exposed wires at the end of the video, but we can assume that was just a tech demo).
fuses only help for overcurrent scenarios. if they cell overdischarges due to a mechanical fault, or internally shorts, the fuses wont do anything. any then if it internally shorts at an SOC > like 20-30%, it'll vent and cascade into other cells.
There a lot of discussion here https://retrocomputing.stackexchange.com/questions/7412/why-... but nothing seems conclusive.. I would wager the last answer, "IBM was using 400Hz", to be most directly causal reason. The motor-generator configuration might provide galvanic isolation and some immunity to spikes and transients as well?
Smaller transformers and capacitors in all the linear power supplies.
400Hz is still common in aircraft. Distribution losses are higher, but you're going across the room, not across the country.
There are very survivable fiber cables designed for stage and A/V setups for instance, and even "real" military grade ones. But the common thin LZSH stuff is surprisingly resilient in my experience, so long as it didn't kink the OP would probably have been fine with a "standard" cable. In any case I would always try to place fiber in a place where it can be re-pulled.. conduit, tray, or a plenum.
I bought a big spool of 6 strand Corning stuff a long time ago for various projects, the cost and diameter don't increase much to add some protection lines even if you never imagine using them they can save you a re-pull if you bugger something up in construction.
The contractors were probably dubious of that resiliency given their lack of experience. I recently ran fibre in my house and have to say I was pleasantly surprised that fiber patch cables (unarmored) can survive a good pull through a duct.
They can survive a 500m pull (in 100m stages, or the friction is too high) in mud and rain, through active water line.
Honestly, fibre, even unarmoured with just a standard Kevlar & HDPE sleeve is hardy stuff. When I first started mucking around with it a few years ago I was like “don’t breathe on it too hard”, now I’m like “tie the fibre in a knot on the bullbar and pull it with the truck”.
> Honestly, fibre, even unarmoured with just a standard Kevlar & HDPE sleeve is hardy stuff.
To be fair, it also got a lot better in the last 20–30 years. In particular, we now have bend-insensitive fiber for the last mile (G.657.A1/G.657.A2) and in general, we just figured out how to make it more robust.
You can kink the shit out of fibre and it’s fine. Like, I’ve accidentally managed ~15mm diameter loops while pulling and then proceeded to yank on them. The Kevlar takes the brunt. Only time I broke a fibre was when it was me and two other guys pulling on it as hard as possible - and instead of moving it went “ping”.
I forget if this was one of the products that used the 801/AMERICA CPU, I don't think it's a COTS microprocessor.
reply