ThinkPad 560E
pappp.net
pappp.net
https://news.ycombinator.com/item?id=9438461
>Ted Selker [1] is the amazing guy who invented and refined the Trackpoint [2] or "Joy Button" as he called it (but IBM refused to call it). He put years of research and development into the product, and I'm happy he's finally written up and published the story. [3]
>[...]
>Once I was sitting in a coffee shop in Mountain View hacking on my Thinkpad, and Ted and his wife Ellen rolled in, sat down, and started chatting. Ted noticed that my Thinkpad's Joy Button was all worn down, and he was mortified and quickly excused himself to go out to the car. Ellen rolled her eyes and shrugged, explaining that he was always like that. Then he came back with a big bag of red Joy Buttons, and replaced my worn-out one right there in the coffee shop, and gave me a few extras as spares!
>He's a brilliant inventor, and a really nice guy, who apparently always carries around a big bag of spare Joy Buttons in case anybody needs one.
Dell and HP already offer TrackPoint-equivalents on their Latitudes/Precisions and EliteBooks (respectively) and have been doing so for more than a decade now, so it seems the patents are definitely expired. I feel like the IBM/Lenovo version is higher-quality, though.
> Since ThinkPad computers have a nub that is responsive to pressure in a direction, and there is a patent for this,[citation needed] other companies have made it so a person has to move the finger a large distance to cause the nub to rock from side to side in a much less efficient way.
The Dell ones are a bit better, but still nowhere near the same quality as a proper TrackPoint.
The older sandpaper TrackPoint caps can scratch your display surface. Though some of the softer ones will also mark the display surface, but not as badly.
I have a small stockpile of caps of different kinds, including some originals (and in nonstandard colors), usually end up using the soft shallow concave or shallow convex ones.
Even if you mix up the colors, you can't escape the distinctive branding. Maybe several years ago, during an I-don't-wear-logos phase, I was using a ThinkPad T60 with black TrackPoint cap, and with black labelmaker tape carefully cut to obscure the branding. One day, I was sitting on the grass in front of the public library. Some group of kindergarteners is cutting across the grass, and one of them glances at my ThinkPad with the blacked-out branding, and immediately says, to no one in particular, "thassa IBM!" I would've gotten away with being branding-free, if it weren't for those meddling kids!
BTW, thank you for the writeup on the 560E. It's a good little laptop, and a lot of open source work was done on it, sitting in parks and cafes. :)
Even when you have plenty of room for a mouse the trackpoint is just better for switching. I think my ideal desktop setup would have a nice mechanical keyboard with trackpoint and a mouse-- the trackpoint isn't ideal for drawing but it's better for most other things.
There are the bluetooth lenovo keyboards, -- their trackpoint is good, but I'd only describe the keyboard as good 'for a laptop keyboard'.
[My idea of a good keyboard, trackpoint aside, is the 122 key model-f terminal keyboard I use.]
(I know there's a company that puts new chips in old chassis, but it'd be great to see new chassis from scratch as well).
The matebook x pro seems more like just another modern ultrabook.
On the other hand the pace of evolution has slowed down and upgrading stuff is less of a necessity today. So most people no longer feel the need to pay for flexibility, they'll go for other things like design, thinness, weight, etc. Phones probably showed the way, they became the ubiquitous computing device and for anyone born in this era they simply set the standard fro all other devices.
I miss my ThinkPad 600x for example but not necessarily because I could (and did) tear it down and build it back over and over again. But also because it seems today we're sacrificing a bit too much for the sake of design. Like impossible to use keyboards, lack of ports, everything glued, etc. because anything else will make the device a bit thicker or heavier.
Instead of making new laptops from scratch, they create modifications to existing ThinkPads.
For example, the X62: https://www.notebookcheck.net/X62-Laptop-Review.211598.0.htm...
Here's their website: https://51nb.com/portal.php
And here are some good starting points for more info:
http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/4B800F78-0F7...
The monochrome displays, in particular, were lovely, and, in spite of its outdated CPU (25MHz 040 IIRC), one of these machines, running NeXTSTEP 3.3, was my primary workstation for terminal-centric work for the next couple years.
While I never did get much into Objective-C beyond the basics, as a big PostScript fan since the red/green/blue book days, I did have a bit of fun with Display PostScript (and NeWS on the SPARCstation 1 the NeXTstation ultimately shared a desk with, although, quite unlike the NeXTstation, the Sun was nearly unusably slow for day-to-day GUI use).
Just this week I got a second one running with a whopping 256 MB of RAM, and put OS/2 (ArcaOS) on it. It’s quite a snappy machine. Loads of fun.
https://snafuhall.com/home-lan-1-the-beginning.html#cracktop
I set it up to boot from the ACPI signal and run with lid closed so it boots from the wall switch.
I dont use that laptop at all these days so I put it to use there.
If you can't find replacement batteries, and you're willing to do the research to make sure you do it safely, you might be able to put new cells in an old battery.
Since (IIRC) you'll probably need PCMCIA for WiFi anyway, rather than getting past mini-PCIe allowlisting, there's less motivation to bring up Coreboot on it. But you could try. https://www.neilvandyke.org/coreboot/
The 560E was a fine machine, and probably close to the last best ThinkPad. A few ThinkPad models later, when I moved to the T60, I didn't like the T60 as much, including the keyboard and fit&finish. Still, today I transplant original T60 keyboards into somewhat newer ThinkPads (since, in the generation after the T60, the keyboard backing started flexing too much, and also, the key action otherwise doesn't seem as reliable). Another reason to find original T60 keyboards is that a lot of the new replacement keyboard parts now are very low-quality counterfeits, e.g., bad key action, and requiring multiple presses of a TrackPoint button to get one to register, on a new unit. (So bad, that cleaning 20 years of strangers' crumbs and dandruff out of an original keyboard starts to seem like even more of an attractive idea.)
Didn’t Steve Jobs carry around a ThinkPad running OpenStep (later Rhapsody?) the first few years he was back at Apple?
I'll admit that even Jean-Louis Gassée (Former Mac head, Be founder) later said NeXT+Steve was the right choice (see: https://9to5mac.com/2011/11/11/gassee-thank-god-apple-chose-... ) - though he said it for personal reasons based on his contempt for Apple's management.
Some not-carefully-composed points of argument that Be might have been the right choice, because I like doing it for sport:
BeOS was truly a ground-up creation, OpenStep was a Display PostScript environment largely licensed from Adobe running on a hacked-up Mach with some BSD pieces pulled inside the kernel boundary to make performance acceptable (this is XNU, the kernel Apple still uses in everything). That had two points for Be - Apple had to replace the whole display layer, partly for licensing cost reasons and partly because developers weren't interested in dealing with the "Yellow Box" OpenStep API or being stuck in the "Blue Box" legacy compatibility jail. Equally or more important, BeOS was vastly more performant on the same hardware - doing the "the UI stays responsive while I'm playing a dozen media files back on this shitty Pentum MMX" stunt was always a blast around the turn of the century. BeOS was also naively SMP with a nifty and mostly pleasant thread model that has some advantages (and annoyances) over the dominant event loop way of things, but always remained responsive, while Apple inherited the spinning beach ball from NeXT.
For another big technical point that history actually agrees with me on, OpenStep used a passable UFS implementation (that Apple reverted to HFS+, which was already getting long in the tooth in the late 90s, apparently largely because Adobe wouldn't get their shit together to make their products run on a case-sensitive FS...), but BeFS was lovely in ways we're finally getting elsewhere 20+ years later - exactly so in fact, Apple eventually hired Dominic Giampaolo who designed BeFS to design APFS, and it really shows.
BeOS was also already suited for semi-embedded applications, both in various commercial multimedia gadgets, and also the BeIA internet appliance stuff. Since the bulk of modern NextStep-family devices are phones and tablets, there's no reason to think a Gasse/Be Apple wouldn't have thrived in this environment (I know, Palm borrowed Be parts for Palm OS Cobalt and didn't do well, but maybe they were already doomed by that time).
Also, Be was _profoundly_ personal, looking toward a "everyone has their own WebServer running their personal services" kind of future, rather than the "Everyone back into the ~~mainframe~~ major cloud provider" future we got, and I can't help but thinking having one of the big industry players not in the roach-motel platform game would have been good for society.
Jobs did in fact use a ThinkPad running OpenStep/Rhapsody builds around the time of the NeXT-buys-Apple-with-Apple's-Money situation. It was a 560 (not E), there is a NextComputers forum thread about it where Marc J. Driftmeyer who set up the machine Steve had at Macworld '97 talks about it http://www.nextcomputers.org/forums/index.php?topic=457.0 .
Reminds me of the old joke I heard years before I went to work at Kaleida, which then became my day-to-day reality:
Q: What do you get when you cross Apple and IBM?
A: IBM.
The degree to which Apple in the 90s was dysfunctional is pretty legendary, they started trying to replace the creaking 68k, cooperative multitasking hacked in after the fact classic Mac OS before 1988 and were still shipping what were supposed to be short-term life support releases until 2001 (...and really until 2002 because 10.0 on available hardware was miserable). They kept losing their replacement plans to development hell. A/UX was a nifty Mac/UNIX hybrid that came out in '88, and it was going to be the future, but was expensive (real UNIX license), deeply tied to 68k hardware, and the 4.0 PowerPC port never came out because it collapsed under the weight of "Build a combined OSF platform out of A/UX and (IBM's) AIX that takes over from both and rules the whole computing market" ambitions. That joint project also included Taligent, an OOP-platform-thing that is basically the benchmark for failed software development. There was also Copland, which was a separate in-house microkernel-based OS that was supposed to be MacOS 8, developed between '94-96... that the Platinum theme was the only survivor from.
I think part of the reason they _did_ go with NeXT is OpenStep looked a lot like the Pink/Taligent platform plan, but not bungled.
...I have a draft of a few-page humorous telling of the "How Apple and IBM spent 15 years failing to write OSes" story but I keep tweaking it and need to dig up my copy of _Apple: The Inside Story of Intrigue, Egomania, and Business Blunders_ to add a few references, and find some better reference material for the OS/2 fiasco between IBM and Microsoft that is one of the core "all this 90s tech industry dysfunction fits together" points most of the tellings miss.
OpenStep was built on top of a then rather battle hardened Objective-C toolchain and "IDE" which had been in development since the 80s, and was an open standard that would target both the Mach OS, Solaris, and Windows NT.
Mach 2 always shipped with a full BSD kernel running as a server -- the modifications were such that BSD would request resources like memory and process time from the mach kernel. This is why for many years Mach based systems kicked the butts of other BSDs. While CSRG's BSD kept its awful virtual memory system from VAX that crapped out after 30-40 MB of address space, Mach's VM system was scalable and performant.
(Mach 3 and above were "more proper", and indeed Darwin's combination of OSF/1's Mach derivative and 4.4BSD-lite's kernel/userland was for performance reasons.)
Mach also gave NeXT a threading model, something absent from BSD. It is exploited by the objective-C framework and the driver layer natively.
https://src.paperfox.es/next.jpg
Because of its BSD roots, NeXT shipped with support for multiple users (something Haiku still lacks real support for to this day, with everything "Be" related stuck under a single user, although the underlying "not-unix-but-POSIX" layer does support it), and a mature BSD sockets library, along with its own (horrific, but there!) interpretation of domain services, NetInfo.
Be shipped with a pseudo-unix environment, with a fair amount of POSIX represented, and a bash shell and GNU utilities preinstalled, albeit lacking any concept of users or permissions...
And to me that's the kicker -- both OSes intentionally had "UNIX layers", Be's was just home-grown and lacked features compared to OpenStep's "4.4BSD but with patches and significantly enhanced resource handling" environment, with much greater software compatibility as a result.
NeXT had the advantage of being an older piece of software, with a much greater depth of "backwards compatibility" while also having a fairly well designed native framework with a well-represented stack of native software. If I was given some (ridiculous) mission of building <x> desktop OS from either BeOS R5's kernel/low level userland or some late release of Open/Rhapsody/Darwin, it'd be a no-brainer to me.
(as a final thought, it's curious you'd mention running a personal web server as an example, when the first web server was developed on NeXT)
(I also ignored Be's filesystem because it is, frankly, superior, however I believe it's not enough by itself to change the merit of the entire OS, since its value either depends on developer buy-in or how much utility it gives to the OS shell, and OS X implemented similar features within Finder using indexers and databases. Ugly, but if it works...)
It was actually my last daily-driver Windows machine. I was working mostly in Office apps at the time, and Win95/98 was just GARBAGE on a laptop. You couldn't reliably sleep. It crashed all the time. Eventually, it drove me to buy a Mac, based on the experiences of one of my business partners. This was pre-OSX, so the OS was only about 30% more stable than Win95, but sleep worked EVERY TIME, and the whole system was more responsive.
Then the dot-com crash happened, and OS X happened, and I went back to coding as a solo act.
But I sure have fond memories of that 560Z.
Now I regret that I got rid of it when it didn't boot. :(
Interesting! A technique for reviving (well, enough to boot and run!) apparently dead laptop batteries... I did not know this was possible!
I’ve managed to get a few old mobile phones back to life using this method.
Edit: as a safety note this is not recommended! Please read some of the comments below!
Lithium battery charge circuits generally have a fully charged voltage (about 4.2v), a fully discharged voltage (about 3.2v), and a 'do not attempt to recharge below this' voltage (about 2.4v).
The reason circuits are designed not to recharge a cell below the 2.4v threshold is because when you recharge below that voltage, gasses are typically generated which eventually (many months later usually) burst the plastic envelope of the battery, and the battery electrolyte then evaporates as a highly flammable gas, killing the battery and potentially starting a fire if there is a nearby ignition source.
TL:DR:. 'dead' batteries refuse to charge for safety reasons, and tricks to bypass that safety mechanism might lead to a fire a few months down the line.
[1] Forgot it was there and put some stuff down on it - there's now about a 20° degree bend in the middle of the long side that goes about half way across the back. Somehow, the screen didn't break.
I’m trying to understand why anyone would lie about not having a hard disk. Is there some rule on the auction website that hard disks must have been removed or wiped?
It's quite often that such finds come through multiple hands, and by the time it hits eBay/whatever the seller has no idea what they are selling except maybe for what's written on labels and can be decided in 5 minutes from looking at the wares :)
That's how I bought my (former now) AlphaStation 255/233, I noticed the case being sold as "old server case for use as HTPC" :)
To this day I'm not sure I've used a laptop that has as nice a form factor. Especially wonderful was the tactile feel of the keyboard.