I used to sell NeXT software and hardware when working for a company called OpenSource (they actually owned the opensource.com domain name before RedHat got it).
At that time "open" meant interopable rather than the current "open source software" meaning. For instance, the DBkit and EOF database software could talk to Sybase or Oracle or any other database for which there was a driver.
As a result we used NeXTstations for all our tasks, and really, that was the best computing environment I have used. As another poster says, we really haven't "budged the needle" since Nextstep v3.3 of the OS.
Later while working for another company I got the ability to talk to Keith Ohlfs who designed the UI and personally thanked him for it.
My understanding is the use of Open was used before this by Symbolics and possibly other lisp machine manufactuers. I'd argue too that their systems were more open, where the user could read the system software and patch it in while it was running.
The foundations of the open source movement too stem from these communities (the GPL) not from NeXt. It would appear to me, like other Unix manufacturers, NeXt claimed openness when mostly it was marketing-speak for "based on Unix".
The word "open" at that time, when used in conjunction with Unix, meant something along the lines of "POSIX" compliant. Even earlier uses were more along the lines of "uses TCPIP and supports NFS" IIRC, though that is before my time.
Open as used above, was started by the users, not the OS vendors, in reaction to APIs that weren't portable between different versions of Unix - the vendors were trying to create lock-in and the users didn't like it. Even Microsoft NT supported POSIX (not sure how well it worked).
Definitely the open source movement came from outside NeXT... Gnu was already around and used a lot, in fact the GCC compiler, ported for Next was the supported C/ObjC/C++ compiler.
The most interesting part of that article was that they produce prototype boards on the same assembly line as production. No waiting for boards to come in. If they want a board made simply queue it and it's done.
They were agile before it was cool to be agile.
EDIT: Also, I had a NeXT Cube and NeXT Turbostation with laser printer. They were fantastic. We've barely budged the needle since then.
Well... that is the difference between PR and practicality. PR says "can" like they do it every day.
While they could run prototype boards down the line, it is not quite as easy as the video makes it sound and it is unlikely they would do that lightly.
To run a prototype board down the line they would have to:
* Get the prototype PCBs fabricated (in 1990, this took a lot more time than today where you can get a multi-layer board fabbed in days). Note also that the NeXT boards were probably a lot more than 4 layers, which adds time and expense, especially 20 years ago. The good news is all the other items probably would take less time to complete than the PCB fab, so this would be the pacing item for the schedule.
* Get a new soldermask made to match the board.
* Procure the necessary non-production parts. With luck, can get small quantities (or samples) quickly from distributors.
* Reprogram the pick-n-place machines to put the parts in the right places.
* Load the pick-n-place machines with the right (new non-production) parts.
* Possibly have to tune the soldering ovens to get the temperature/time profile right (should be OK for small part changes).
Once everything is in-house and programmed, you can simply queue it and it is done.
It's also likely that they did it on the same line because the automated equipment to build the stuff was so hideously expensive that they had to build it on the same line.
Completely agree with this. I also had a Cube, and then a NeXTStation Turbo Color (which I still own), and while Mac OS X has certainly evolved, basically all the important things we have today, and the elegance of design, were already present in NeXTStep 3.x. If I were required by law to stop using OS X tomorrow, I'd go back to NeXTStep.
If they had started to gain traction that factory would have been a significant competitive advantage for them. Further, I think that it could have helped them lower prices even more think, sub $1500 workstations.
Their hardware was a resounding flop and building a state-of-the-art production facility for something for which there was no demand is in retrospect clearly a mistake.
Also, given that NeXT was bought for about 15% of Apple's market cap (after laying off most of their employees), it's disingenuous to imply that Apple == NeXT. What saved Apple was the iPod, not OS X and Cocoa.
NeXT, effectively, acquired Apple for negative $429 million (and 1.5 million shares of Apple stock): As everyone knows by now, what began as an Apple purchase of NeXT has been turned inside out and become a NeXT takeover.
edit: I want to add a couple points to my followup.
1. NeXT's hardware was a resounding flop. Their state-of-the-art factory turned out to be for naught. This is absolutely true.
2. That said, this mistake did not prevent NeXT from co-opting Apple and becoming the third largest company (by market cap) in the world.
3. Apple == NeXT: Apple was, pun intended, rotten. Steve Jobs and his most trusted lieutenants assumed key positions of power throughout Apple. The only highly visible and critical Apple employee I can identify who pre-dated Steve's return to Apple is Jony Ive. However, Ive was laboring in obscurity before Steve assumed his position as "iCEO" of Apple[1 and 2].
4. Arguably, the iMac saved Apple. In one of the recently-posted WWDC or Macworld keynotes, Steve addresses this point in quite a bit of detail. The iPod made Apple a veritable powerhouse...but not at first[3]. The iPhone made Apple, well, Apple. As it happens, the iPhone (and OS X, and the iPad) is the result of decades of engineering work that started at NeXT[4].
The only highly visible and critical Apple employee I can identify who pre-dated Steve's return to Apple is Jony Ive.
Fred Anderson, the CFO under Amelio, was kept on until the backdating scandal. He played a key role in saving Apple--they were almost cash broke when Amelio took over, and Anderson arranged a key bond issue to keep the company afloat.
Journalistic hyperbole is not equal to fact. Apple acquired NeXT, for $400 million in cash in stock, at a time in which Apple had a market cap of $3.1 billion. I don't know why you seem to think otherwise. Read it from the horse's mouth along with the relevant financial statements:
http://web.archive.org/web/20020208190346/http://product.inf...
Within a period of, I believe, 12 months, Steve Jobs became CEO, installed NeXT execs in almost every key exec role in Apple, and replaced the board of directors.
Although Apple acquired NeXT, in every meaningful way, NeXT consumed Apple.
Takeover for me means change of hands of ownership and power of decision in the company, and whatever your technical or managerial arguments are I think the resulting company was owned by Apple's stockholders, Steve Jobs probably among then after the acquisition, not by NeXT investors.
For the 2 people that downvoted I would love to know what you consider a takeover. I'm pretty sure that in financial term takeover means change of ownership.
BTW, I'm quit, HN is just becoming a place where Steve Jobs and Rails must not be criticized, even when they must be. I will put some random string as my password and never return here, it was a good time but I think the community lost something in these years.
Here's a low quality video of the factory from back in the day http://www.youtube.com/watch?v=jhfUKEu7sJ0 (The copy of the video is low quality that is, the original production values were excellent)
The best I could find on the fate of this factory was from Wikipedia [1].
"NeXT withdrew from the hardware business in 1993 and the company was renamed NeXT Software Inc; subsequently 300 of the 540 staff employees were laid off. NeXT negotiated to sell the hardware business including the Fremont factory to Canon. Canon later pulled out of the deal. Work on the PowerPC machines was stopped along with all hardware production."
The choice of storage technology always seemed to me to be the reason it didn't take off. Kind of a bummer that it didn't work out (some would say it did, à la Apple).
Of course it worked out, half the people here run the latest version of NeXTStep on their phones and pads, another half run it on their desktops, and NeXT is considered the most successful contemporary computer company and is dominating the industry in innovation, all possible because of the AppKit and FoundationKit and the various NS kit APIs that came since then. They even acquired their closest rival, rebranded according to their trademark, and then deep sixed the rivals products, replacing them with their own.
I was in elementary school and a guy came in for "Career Day" to speak with the five of us who were interested in computing. He brought a NeXT computer (I have no idea which one), with some rectangular cartridge (must have been the magneto-optical media). He explained why it was such a fantastic machine, and proceeded to spend the allotted 30 minutes attempting to boot it.
I have no idea who this guy was or what he did for a living, but it was my very first experience watching a demo go horribly wrong. I have to say, that event influenced me in over-preparing for any sort of demo. I don't remember the details, but I remembered that it was NeXT, and I remembered that it didn't boot. (OK, I'll admit that I partly remembered the name because the capital XT reminded me of my clunky IBM PC XT clone I had at home).
I remember well going to a NeXT exhibit at Michigan State. I was blown away by the beauty of the machine but thought the disk storage and lack of color graphics were deal killers.
The sales guy working the booth was quick to let me know that I could order a machine with a hard drive and that I didn't have to take the optical disk. In fact he said everyone was ordering them that way. But he grew silent when asked when asked about the availability of color monitors.
Sure wish I had known when the universities NeXT cubes went through salvage because I would have certaintly bought one - they were absolutely beautiful.
There were NeXT slabs that were available from salvage circa 1998. They were likely the machines in the computing lab in Wells Hall, next to the language lab. There were also a half dozen color turbo NeXT stations in Holmes Hall.
Those machines were some of the nicest to work on Mathematica with. We also had installed the ISCA BBS client on those machines, before the remote telnet access was turned off.
What's sad was how expensive Openstep was once Next was solely a software company. I had a Nextstation and decided I wanted a copy of OpenStep 4 for my PC. Next was asking for something like 800 dollars for a copy.
"In Mr. Jobs's view, the factory testifies to the fact that the United States can still compete as both a low-cost and a world-class manufacturer when it sets its mind to the task."
Due to wage/cost inflation in China and the coming rise of the Yuan (along with the weakness of the dollar), the US is about to become the cheapest place for high end manufacturing in the world.
Modulo regulatory approval, of course. Most of my family works in various manufacturing industries (high end foundry, steel mills, etc.) and you're looking at years to get even the most basic permits through unless you are in one of the states that allow you to "purchase" fast-track approval. Companies end up buying the old, dead husks of manufacturing plants that went out of business not because they want any of the machines or people but because their permits are often still valid and in some cases transferable to your other locations.
Even if you wanted to expand to meet increasing demand, the opportunity will be gone before you can get approval to build out for it (even in relatively business-friendly states like IN). And the company-purchase strategy has to be taken carefully, both because business capital is hard to come by today and because manufacturing operates with margins so small that there's no room for error.
Further, you have the risk of OSHA shutting you down or penalizing you to unprofitability. As I understand it, safety regulations are much like traffic laws -- if a police officer wants to pull you over while you're driving, they can almost definitely find something to charge you with, even if it's just their interpretation -- and the same is true of your OSHA inspector.
Wages/productivity play a large portion of being a good place to do high end manufacturing, but are still a relatively small part of your ability to profitably grow and run a manufacturing business.
Plus IP and counterfeit/security issues, communications issues, and such. Provided I didn't need to be colocated with Chinese parts suppliers, and that my production process wasn't dirty enough to have major environmental issues in the US, the US seems like a top choice.
The good news is that the U.S. is decades ahead of most countries in terms of automated manufacture. Probably only Japan and South Korea (maybe Germany too) can comfortably compare. The videos linked to in this same topic show a completely automated surface mount process to build the motherboard of a computer. Something that's still not regularly used in most countries because it's hard and expensive. Easier to just have low paid workers solder the stuff on by hand.
Well, that could be due a convergence between automation and labor-intensive manufacturing (which can't be automated easily) moving overseas to fine a cheaper work force.
Also, I think the parent post was referring to the fact that Steve Jobs is now CEO of Apple which IIRC does most of their manufacturing overseas (SE Asia). If SJ still believed that quote, then he would have been trying to 'bring it home' so to speak.
"Also, I think the parent post was referring to the fact that Steve Jobs is now CEO of Apple which IIRC does most of their manufacturing overseas (SE Asia)."
Does Apple have any manufacturing operations currently? I would be surprised if they did.
If by manufacturing, you mean chemicals, pharmaceuticals, processed agricultural products, pulp&paper, and heavy industrial equipment. Electronics and consumer products manufacturing is dead in the US.
Dell and HP manufacture almost nothing in the US. Many/most of their products aren't even designed in-house. They contract with Asian design houses and slap their own bezel on the result.
You're right about Intel. They are one of the few significant electronics companies left in the US. But they aren't representative of the category.
Yet both the iPad nor the iPhone were manufactured in China. Hence, I'm curious why Steve Jobs reversed an opinion he previously worked so hard to support, as evidenced by the article.
I'm assuming that you have relatively recent experience in order to make this kind of statement. If you are happy with the cost, quality, capability and flexibility of your US-based contract manufacturing solution, please name names as there are many people who would be interested.
The 1st-gen NeXT keyboards use Alps key-switches, like the Apple keyboards of the late-80s. The one I have is rock-solid and one of the finest keyboards I've used.
It also had some clever innovations:
- It is very compact. It has the same width as a traditional PC keyboard, but without the top row of function keys. That means it's only 5 rows high.
- Consequently, the Escape key is next to the "1" key, where the backtick/tilde key would normally be. However, the tilde is an important character in Unix, so if you hit Shift+Escape as if to type a tilde, you'll still get a tilde character. It's a nice touch.
- The Control key is next to the "A". Command and Option keys are on both sides of the space bar.
- There is no Caps Lock key. To engage Caps Lock, press Command+Shift; green LEDs on both Shift keys light up to indicate Caps Lock.
- There are no home/end/page up/page down keys. They are replaced with Power, Volume up/down, and Brightness up/down. All NextSTEP applications support the traditional Emacs shortcuts for Home/End/Page Up/Page Down etc., and that tradition lives on in OS X.
- The hardware (cube, monitor, or printer) has no buttons or switches of any kind. Power, volume, and brightness can only be controlled from the keyboard (like a modern laptop). That means you can't turn the machine on or off without the keyboard, and the keyboard connects to the monitor!
The OS X keyboard technically still has a Caps Lock key; but you're free to remap it to any of the other modifier keys (or none of them) in the keyboard prefs. Being a Vim user I don't use Control much, so I put it to ⌘ to call up Alfred more easily.
I actually prefer the feel and sound of the NeXT keyboard over the model M. Though I much prefer the layout of the model M to the NeXT. Coding was a pain on the NeXT keyboard - no pageup/dn home/end keys. But it was fantastic for writing papers and working with mathematica, frame, etc.
IIRC the vertical bar (or was it the tilde) was in a totally nonstandard and awkward place too... didn't seem very smart considering it was still a UNIX machine.
The navigation keys were not such a big deal once you realized that all text fields accepted basic Emacs keystrokes for cursor movement.
Unless you like emacs. No Caps Lock taking up precious space, Control/meta/super keys in great position, and the emacs keybindings work not only in emacs, but system wide.
My uncle did some electrical work for NeXT when they were first setting up, and he said he knew they were going to fail. They'd just finished putting in a very expensive new hardwood floor, and the guy in charge of design came in and said "it's going the wrong way- rip it out and do it right."
It's interesting how that story is colored differently, depending on whether one considers NeXT (and/or its perfectionist attitude) to be a spectacular success.
Ironic to look back on now, considering that OS X and the Macs that a lot of us use are the direct descendants of NeXTStep and the NeXT hardware.
I owned a few slabs back in the day, and even found a picture of when I used one as my main desktop at an ISP in the late 90s: http://www.mrbill.net/next.jpg
If I recall correctly, id built much of the software on NeXT systems for as long as it was reasonably possible. I have a feeling Carmack loves his current work on iOS.
Being what I call a "computer hobbyist", I have used quite a few operating systems over the years, and one of my all time favorites is Nextstep (along with BeOS). In fact, I still prefer the overall look of Nextstep over Mac OS X, even now, many years later.
Back in 1996, I bought a used Nextstation from a company called Spherical Solutions, run by a man named Sam Goldberger (IIRC). I loved that computer, but sadly, it was killed by a good old Florida lightening strike in 1998.
If anyone is interested in an analysis of the rise and fall of NeXT, I highly recommend Randall Stross' "Steve Jobs & the NeXT Big Thing", which was published in 1993.
Considering Apple's incredible comeback, it's kind of strange to read something that is so critical of Jobs, but the book really helps inform about his thought processes on product design, marketing, and sales.
:( I once owned 2 next cubes with matching printers. Gave them away many years ago when I started downsizing. I miss them more than some ex-girlfriends :)
In what way? Entrepreneurs are still failing and then eventually hitting home runs. Steve's full name is still Steven P. Jobs. And the US can still complete as both a low-cost and a world-class manufacturer.
"And the US can still complete as both a low-cost and a world-class manufacturer."
Assuming that you're referring to the type of electronics manufacturing that NeXT was doing at its plant, can you provide any evidence to support this assertion?
Most of that is almost certainly, in part, for tax related reasons. Intel is profitable; profitable US companies want to keep those profits outside the US (see: Google).
At that time "open" meant interopable rather than the current "open source software" meaning. For instance, the DBkit and EOF database software could talk to Sybase or Oracle or any other database for which there was a driver.
As a result we used NeXTstations for all our tasks, and really, that was the best computing environment I have used. As another poster says, we really haven't "budged the needle" since Nextstep v3.3 of the OS.
Later while working for another company I got the ability to talk to Keith Ohlfs who designed the UI and personally thanked him for it.