Hacker News Reader

I had the unique fortune of having one at home when I was in my very early adolescence (born 1981) because apparently my mother had one on loan from the institution she worked for as a statistician. It was perhaps one of the most influential pieces of technology in my life (along with my later BeBox running BeOS).

The machine was truly unique and the look and feel of the GUI still represents an almost global maximum in my aesthetic sensibilities. The programming model undergirds my expectations of what a ‘modern’ development environment should be like.

A truly formative, almost ineffably mythical, experience.

As a NeXT developer who started the NeXT User Group at my college, I agree they were lovely machines, way ahead of their time.

But I want to be quite clear that as a business, NeXT failed. They started out wanting to be an integrated hardware/software company, like Apple. But Jobsian arrogance entirely undercut that; it was a machine developed (and priced) for people like Jobs, so it never found market acceptance outside of a few market niches like finance, where people had money to burn.

Eventually they gave up on hardware and shifted to being an operating system on other people's hardware: Sun, HP, and eventually, a narrow set of Intel-based PCs. This kept them going a little longer, and made things easier for their existing customers doing in-house development, but they were still failing more slowly.

They then pivoted toward being developer tools for Windows machines, except they could only run on the more-expensive Windows NT. This too was headed to eventual failure.

Apple did screw up writing a modern operating system enough that they decided to look at buying one, and I know Be and NeXT were candidates. But I don't think the deciding factor was technical. What Apple's board wanted was Jobs, so I think of the NeXT acquisition as more of an acquihire with some bonus technology.

Yes, Jobs and his investors eventually turned a modest profit on the sale. But Apple quickly shut down all of the NeXT stuff, leaving NeXT developers like me high and dry, which is a strong sign it was not a viable business. So I look at the sale as Jobs's price for rejoining Apple, his way of not admitting failure.

Didn’t quite a lot of the NeXTSTOP code form the base of OS X? AFAIK many functions in macOS and iOS still have the prefix “NS”, which is part of that heritage.

Sure. That was the "some bonus technology" part. The IP bought from NeXT in 1996 eventually turned into OS X, launched in 2001. At WWDC 1997, They promised us NeXT developers that they wouldn't abandon us. But it turned out that clients don't wait around 3-4 years to find out what's going to happen an OS with no future beyond hazy promises.

I think at that point they thought Rhapsody was going to be enough.

We had one of the pizza box NeXTstations at the UCC[0], at the University of Western Australia, and at one point we also had an IBM workstation (from memory it was PowerPC?) running OpenSTEP. That was in the late 90s/early 2000s.

At that stage there was lots of interesting hardware for us to collect; we had a BeBox, three SGI workstations including an O2, an Alpha workstation and a couple of Alpha based servers.

I dropped in at the UCC not that long ago and sadly, the days of cool interesting workstations seem to be over. It's just PCs now. I guess the energy of design and innovation isn't focused on the desktop anymore, but those days were really cool :)

[0] https://ucc.asn.au

UW had a lab of them that I got to use for a math class once or twice. I used Mathematica on a NeXT and on a Macintosh II, and it was clear NeXT was just better at everything. If you had told me they used the same processor family, I probably wouldn't have believed you. It felt like as much a step up from the Macintosh as the Macintosh did from the Apple II.

It was essentially a slow monochrome early Mac OS X release, in the days when the Macintosh had System 6 or System 7. It really was more than 10 years ahead of its time.

> If you had told me they used the same processor family, I probably wouldn't have believed you.

I don't know how real it was, but one of NeXT's marketing points at the time was the power of all the supporting logic they put around the main CPU. There was a bunch of DMA support and a MC56001 signal processor from the beginning, and later on, it was possible to buy video boards from NeXT with onboard i860 processors.

One area where they did fall down, at least initially, was in the Magneto-Optical primary storage. The original NeXT cube shipped with one of these _instead of_ a hard disk. I think the idea was that you could take a single MO disk around from workstation to workstation and carry your environment around with you. The downside is that the access time on the MO drive was something around 80ms or more in an era when a hard disk was 20ms or so. Killed performance, as far as I understand, and the next generation machine was a more traditional hard-disk/floppy combination.

I think the idea was that you could take a single MO disk around from workstation to workstation and carry your environment around with you.

This philosophy lived on into the Apple years. The reason the original iPods had Firewire connectors on them was so that you could eventually go from Mac to Mac and just boot off of your iPod and continue your work.

It's why a lot of people who worked with Xserve servers always carried iPods with them — the servers had Firewire ports right on the front and you could plug your iPod in and access your files.

Now that I think about it, it's also interesting that it's shared with Jef Raskin's Canon Cat. (Which had to be a frustration with Raskin, given that Jobs took over his original Macintosh project and then arranged to have Canon kill the Cat, in relation to Canon's funding of NeXT.)

Back in 2003 Apple touted a feature called “Home on iPod” being petr of the then up-coming OS X Panther that would have supposedly allowed one to carry one’s home directory around on an iPod and mount it on any system one could authenticate against.

It was pulled before it was released, however, and never heard from again. As far as I know, no justification was ever given.

https://www.macrumors.com/2003/10/08/pulled-panther-feature-...

I was at Motorola Tokyo Design Ops, back in the day and the General Magic people and the old Mac people and the Next people were frequently coming through.

I learned the 56K on the NeXT (and it saved my butt when I had to emulate a slow small MCU (together with an FPGA) to build a real-time emulator so our customer (Canon) could complete development in parallel with the US mothership getting their act together for shipping verified silicon). But I didn't spend much time with the i860. I should have. And, over course, their main CPU (68040) was better than the 68020s running on my desktop workstation.

That Magnetic Optic thing, though, it turned a pretty slick design, and a very slick SDK, and an "ok" implementation of Display Postscript into something that just felt like a sluggish brick. And, of course, back in the day, Jobs hated networking because it was "just an umbilical cord back to the mothership".

> back in the day, Jobs hated networking because it was "just an umbilical cord back to the mothership"

I didn't realize this... but it completely fits both the person and the time.

For those that might be unaware, back in the 70's through maybe even the early 90's, there was a belief that personal computing represented a form of personal liberation from centralized control/power.

The idea was that a computer you could own and manage yourself represented freedom from the establishment (which took the for, of centralized minicomputers, mainframes, and the like). Of course, 30 years later, we now know how that all turned out, with the idea of the network being a connection back to a central mothership sounding pretty darn accurate.

If you ever get to Seattle, you gotta check out the Living Computer Museum. You won't regret it.

https://livingcomputers.org/

I have seen the Alto - even used one at a museum, what's amazing is how not slow it feels.

> One might make the argument that NeXT's acquisition by Apple is something that almost every tech startup in 2020 could only dream of.

One might also make the argument that this is basically the most extreme case of "acquihiring" ever.

Given the number of Next personnel who were moved into the top leadership positions at Apple, the joke at the time was that Next acquired Apple for -400 million dollars.

I've seen something similar happen on a smaller scale before, when working at a payment terminal firm that bought a smaller firm.

At the time we referred to it as a "takeunder"

We used NeXTstations (and later OPENSTEP on x86) at one of my first jobs (equities and options trading floor in a bank).

Was very impressed by them. As I was already using FreeBSD, Mac OS X was like the perfect combination when it came out.

I _almost_ convinced my parents to buy me a cube as a HS graduation present.

They were all for it until they saw the price?

I seem to remember it being basically a deal of co-signing a loan and helping with a deposit and considering it part of my first year uni tuition. But it really was way too expensive.

Ended up getting a 486 50mhz and running very early (like pre 1.0) Linux on it. That scratched a bit of the itch :-)

I tried (and failed).

Once I got to college, the basement of my dorm had a computer shop that specialized in x86 NeXTStep hardware. I tried to find a way to make that work too, but the costs were just too high. You didn't need just a 486DX2/66, it had to have EISA, 32MB RAM, lots of SCSI disk, and the like. It was easily a $6-7K proposition.

When I first started Mac OS X development, and later iOS, interface Builder and Objective C were a revelation. To this day, “NS” (for NeXTSTEP) prefixed classes can still be found in the mac frameworks. The company may have ‘failed’, but the ideas succeeded. (And given how much Steve’s companies were an extension of himself, ‘failed’ is subjective here — in the end, it was through NeXT that Jobs won back Apple.)

As far as I know, "NS" actually stands for "NeXT and Sun", and the original prefix before OPENSTEP was "NX".

Had to do some googling, which led me full-circle back to HN :)

https://news.ycombinator.com/item?id=15977053

Interesting, the explanation that it stands for "NeXT Software, Inc" that came up in the parent thread seems plausible.

I always loved the clever phrasing: "NeXT bought Apple for a negative $430M and 1.5M of Apple Stock Shares in 1997."

My graduation thesis was porting a particle visualization framework written by my supervisor on a NeXT Cube, on Objective-C / OpenGL, into Windows using C++ / MFC / OpenGL.

The decision was taken because the department wasn't seeing a future to keep those NeXT workstations around, as this was shortly before Apple's acquisition.

Those Renderman manuals were great, though.

They were rare even in their day. I was visiting lots of customer sites at the time and I only ever came across one, at a bank in London.

Xerox Alto was launched the year before I was born and to this day I have used an original NeXT Computer during 2 years ;)