Hacker News new | more | comments | ask | show | jobs | submit login
How I found one of the earliest browsers on a NeXTcube (pupeno.com)
137 points by wslh on Dec 25, 2015 | hide | past | web | favorite | 69 comments

When I joined NeXT one of my first assignments was to evaluate a new piece of software that we were considering releasing. It was software that would automatically prompt you to sign up for an Internet connection when your new NeXT machine first booted. It communicated with Virginia to get the new machine an IP address. I ran it a bunch of times and it usually worked - but it would sometimes fail. And when it failed, we had no way to debug the problems on remote systems outside our control. And since there were monetary charges involved, this got very messy very quickly.

Based on my strong report, Steve decided that NeXT boxes wouldn't automatically sign you up for an Internet connection - since we couldn't debug or support the customer when things went wrong.

Candidly, I've looked back on that report and wondered if I set back the entire industry. We weren't looking at browsers - but we were looking at access to all communication possibilities across the Internet as well as the access to University and Government resources (ftp and remote mounts.)

Or maybe you prevented a bunch of people from having a first bad experience with the internet.

Or maybe they provided the conditions ripe for Sky Dayton, of Earthlink, to negotiate with Steve Jobs to have Earthlink bundled on the iMacs when he went back to Apple ..

Apparently Steve Jobs narrowly missed Berners Lee demonstrating his browser for inclusion with an upcoming version of next step to catch a plane.

I constantly think about how them meeting that day may have moved the mainstream adoption of the web up by 5 years.

Harumph. I was an editor for NeXTWorld magazine [1] at the time TimBL invented the web. "WWWNeXTStepEditor version 0.12" was the first web browser I used.

Mass adoption of the web was held back by lack of ISPs. It may seem strange now, but in 1991 there was no way to connect to the Internet unless you were affiliated with a university or a big R&D company. in 1991, TimBl's WWW wasn't any more relevant to NeXT than Gopher or Brewster Kahle's WAIS or, for that matter, CompuServe.

Some of us at NeXTWorld, including John Perry Barlow, Simson Garfinkel, and Seth Ross saw the potential of the WorldWideWeb, but we never wrote about it at NeXTWorld because most NeXT users couldn't connect to the Internet. I don't know when Jobs first saw the WorldWideWeb but he had many opportunities and it never was important until he got excited about WebOjects at Apple. By that time, Marc Andreessen had released Mosaic, the first web browser for Windows, and ISPs were sprouting up to enable consumer connections to the Internet.

NeXT computers would have needed to provide a modem, and PPP for TCP/IP over dialup, and NeXT-provided connection points (through a 3rd party partner?) to make the web useful to NeXT users outside of academia and government.

[1] http://www.fortuityconsulting.com/danielkehoe/earlyweb.html

The prototype I evaluated at NeXT was supposed to work with any modem (remember Hayes compatibility?) and didn't need PPP, since we had arranged a proprietary handshake with the organization that gave out IP addresses (I've forgotten the name. At the time I believe there was just one entity allowed to give out IP addresses.) We discussed requiring a specific modem - but decided that we didn't want that to be a requirement for getting an IP address.

Ultimately, it came down to "unsupportability" by NeXT. Several people other than me told Steve that things would go wrong connecting to the Internet, and when they did we would be mostly helpless to support them. At the same time, the organization in Virginia didn't want to support the software either - the software was a "black box" mostly provided by us on their machines. So we killed the project. I never heard anybody mention browsers.

It's absolutely fascinating to hear this story. I'd never heard it before. That was a historically significant moment. Do you recall if the software was internally developed at NeXT or acquired from a third party?

I was told that we had developed most of the software in conjunction with the group that gave out IP addresses - but that we ended up doing almost all the work ourselves and setting up their machines at their site. (They were skeptical about the whole project, and not enthusiastic about the automatic registration of new IPs. When I spoke to them it was obvious that they wanted to ignore the whole project.) My impression was that our engineers who had done the work weren't around Redwood City (or perhaps they had quit - we weren't doing well and lots of people were leaving at the time.) So I was figuring out the code on my own and doing the debugging - nobody was taking ownership of the code. It was an overly complex mess since it both registered for an IP and then did billing. Very fragile.

> and it never was important until he got excited about WebOjects at Apple.

Not sure exactly what you mean by that since WebObjects was created at NeXT way before Steve was back at Apple. In fact before the acquisition Steve was the main WebObjects promoter so way before being back at Apple. True at Apple, Steve did continue to promote WebObjects internally for a couple years but afterwards he moved on.

Ah, you're absolutely right. I forgot WebObjects was a NeXT project. Now I'm wondering when Steve Jobs first saw the WorldWideWeb.

What enabled mainstream web adoption was mainstream computers (Intel 486) and modems (14.4k/28.8k) getting fast enough to make something like the web useful. Jobs wasn't in a position to speed that up. They got affordable in 1993. (TBL defined the cornerstones of the WWW in 1989.)

In 1989 Jobs was still selling (very few) Next computers for an entry price of $10k.

The legislation that allowed someone to get their own connection to the internet (essentially to become an ISP) without having a government sponsor didn't exist until 1995.

But commercial ISPs with full access to the Internet existed in the USA since 1992. The first one was The World. Details at http://www.indra.com/homepages/spike/isp.html

Yeah, there was no such legislation - or if there was, it wasn't being followed by those who had Internet access in the early 90's.

Untrue. You could get commercial internet connections from UUNet and others much earlier (1992?)

Could you say more about how you think that would have happened? I was a NeXT developer at the time, and I'm having trouble imagining it.

At the time, NextStep was a struggling niche OS. Jobs didn't get any real power until he returned to the helm of Apple in '97, and NextStep technology didn't hit the mainstream until 2001 with the release of the consumer version of OS X.

In contrast, Netscape's first product was released in late '94, and it was popular enough that they IPO'd in 1995.

I can imagine an alternate universe where Jobs has a "Xerox" moment and immediately pivots NeXT's software efforts to developing a cross-platform graphical browser.

The "cross-platform" part is the most optimistic, to be sure. But if Jobs believed that NeXT's real value was the software (as OS X, and especially iOS, bear out), then you can imagine a strategy that used a web browser as a Trojan Horse for a platform that could overthrow the currently dominant Windows monopoly. That was basically Netscape's strategy a few years later, after all.

Regular people couldn't afford Next computers. They used PC clones (running mostly 16-bit Windows at the time, which is why Netscape was successful -- their browser ran on 16-bit Windows).

Which is why the parent mentions "cross platform" repeatedly.


The problem with the NextStep stuff at the time was that it required high-end hardware. Even years later when they started to make cross-platform development tools, they were limited to niche markets where you new your users would pay up for plenty of RAM. That was fine for, e.g., financial trading.

So I think they would have had to start basically from scratch, sullying their hands by writing for the sad world of 16-bit Windows 3.1. If they had done it, though, it could have been incredible.

On the other hand, they could well have been killed by the same issue that killed Netscape: the web, being definitionally open, doesn't provide a lot of defensible advantage to early movers.

All true. I think this is good evidence we live in the best of all possible worlds: A Jobs/NeXT WWW browser a couple of years ahead of Netscape probably would have fared even worse against Microsoft's inevitable counterattack. For all we know, we could have ended up in a "Blackbird" (aka MSN) dominated universe, and without a revitalized Apple to give us Unix on the desktop.

Still, I do find it totally plausible that Jobs could have been inspired to redirect NeXT's struggling business to the nascent web. Some of those old WebObjects interviews that get posted to HN periodically demonstrate the he "got" the web and its potential as an application delivery platform much earlier than most.

NEXTSTEP was cross platform, so not really such a far fetched idea.

NeXTSTEP only became "cross platform" (meaning: ran on x86 hardware) as a desperate alternative to their failing proprietary hardware business. Seeng as how at that point the OS itself was NeXT's business, it's hard to imagine Jobs supporting software that ran on competing OSes.

Even today the only real cross-platform product Apple makes is iTunes, and that only exists so they can sell iPods (and now iPhones) to windows users. Facetime, ApplePay, iBooks... all would be better as cross platform apps, but Apple isn't interested because it doesn't help them sell more macs/iphones.

And that decision almost never happened:

While it seems a given that the iPod was to be made compatible with Windows, Jobs was very resistant to the idea. At one point he said that Windows users would get to use the iPod "over my dead body". After continued convincing, Jobs gave up:

"Screw it," he said at one meeting where they showed him the analysis. "I'm sick of listening to you assholes. Go do whatever the hell you want."


Think about what iPod sales would have been if iTunes never made it to the PC platform. Would there have been enough profit to fund the development of the iPhone? Would we be where we are today?

Well Android may still have happened (but perhaps looked more like Blackberry in its initial form). And Nokia may be around selling Maemo based phoned, though i wonder if what sunk them was Samsung (and later Chinese companies) undermining their feature phone sales in Africa etc.

> it's hard to imagine Jobs supporting software that ran on competing OSes.

Perhaps hard to imagine, but they actually did it. OPENSTEP Enterprise allowed you to use the NeXT development tools to build apps for 32-bit Windows platforms. That came toward the end of the NeXT saga, so you could argue it was a move of last resort. But it clearly wasn't out of the question.

until 10.5 or 10.6, OS X still included the classic NextStep and Windows GUI images. http://www.theregister.co.uk/2008/09/02/mac_images/

But maybe, NextSTEP would have been more relevant if it was an OS that allowed you to use the web. But then, the web back then wasn't that great and I think the tipping point was ISPs offering commercial internet access in 94 or 95.

er, the web was invented there...

I wonder if we'll ever see the day Apple makes a windows machine

If we do, it wouldn't be Apple but in name, just like with SGI's NT machines.

Those were the most depressing SGI machines I ever saw. The ugliest too.

this is not correct. please be more careful not to assert opinions as facts.

But Nextstep was a very expensive OS. I looked into it and they wanted somewhere in the neighborhood of $900.00 for a copy. I eventually bought a used NextStation from Sam Goldberger's Spherical Solutions.

Sure, but that's only because they were selling it into niche markets willing to pay a lot for its advantages. If they had been targeting deployment to consumer platforms, they would have priced it appropriately.

I know it almost happened because Tim Berners Lee said it did: https://www.w3.org/blog/2011/10/steve-jobs/

"We almost met once.

There was a get-together of NeXT developers in France, and we set up demos at tables around the room.

Robert Cailliau and I set up the WorldWideWeb.app on one table.

Steve arrived, and started making his way around the tables chatting with each project

He didn’t get to us before he had to leave."

Isn't that incredible?! Even if you aren't a fan of Steve Jobs you have to admit he would have instantly understood the power of what Berners Lee was showing him.

I think about that story a lot. Gives me goosebumps.

Sorry, I'm not questioning that they almost met. I'm just not seeing what difference it would have made to the Web.

NeXT, for all its technical merit, was a repeated commercial failure. It failed as hardware, it failed as an OS for other people's hardware, and then it failed as developer tools of other people's OSes. The Jobs of 2007, at the helm of a successful company, perhaps could take an obscure technolog7 and get it widely adopted. The Jobs of 1992 had no leverage. The NeXT was full of amazing tech but he couldn't get anybody but a few koolaid-drinkers, myself included, to use it.

So if he'd adopted the WWW into the NeXT OS a year or two earlier than everybody else, I think it would have been another cool technology on the NeXT platform that everybody ignored. Mosaic and Navigator made a difference not really for what they were, but because they were available to a large audience of Windows users.

If anything, I think the later history of Apple suggests that Jobs's talent wasn't in cutting-edge technology, but in taking established technology, making it user friendly, and marketing the hell out of it. The iPod wasn't the first music player and the iPhone wasn't the first smartphone. They were just the first ones built for and marketed to the broad consumer market. There was never really a consumer market for browsers, which is part of what killed Netscape.

>NeXT, for all its technical merit, was a repeated commercial failure. It failed as hardware, it failed as an OS for other people's hardware, and then it failed as developer tools of other people's OSes.

Well, we seem to forget that OS X 10.0 is basically NeXT with a new skin. So it might have failed as an "OS for other people's hardware" meaning PCs, but it did phenomenally well with Apple's hardware.

And even the "failed" NeXT company got sold for a cool half billion, an order of magnitude better than e.g. BeOS ($11 million).

And NeXT's itself market performance was OK enough to allow it to live and reach several versions all these years up to 1997 until it was bought.

They failed to live up to any of their hype or any of their market goals. It didn't revolutionize education. It didn't sell much hardware, despite multiple hardware versions. It didn't gain significant market share as an OS. It didn't gain significant market share in Windows developer tools. When Apple bought them, the NeXT market was negligible enough that they basically stopped making anything for it.

They definitely had some success in niche markets, I think they turned a modest profit one year, and I don't think their investors lost any money. But it's not like they made a ton for the level of risk they took.

Yes, Jobs managed to make lemonade out of his multiple failures. But NeXT was never a commercial success.

Watch Cringley's interview of Jobs in which Jobs specifies precisely how much of the Xerox Parc technology that he saw but didn't grasp the importance of including networking and object oriented programming.

Steve Jobs was for a long time seen as someone with brilliance who tripped up on his own misguided and dogmatic world view, not the current picture of him as the ultimate visionary who sees and knows all.

  > Even if  you aren't a fan  of Steve Jobs you  have to
  > admit he would have instantly understood the power of
  > what Berners Lee was showing him.
I see no reason to think Steve Jobs was omniscient. Do you really think he could instantly an accurately assess the future importance of a technology?

Seems to have worked out for him all the countless other times. For examples see Apple Computer, Pixar, NeXt...

Berners Lee himself writes about the possibility decades later. Decades later! The inventor of the web browser! Remembering and writing about something that almost happened... but didn't. He's clearly given some thought to the matter.

It's not like you remembering the time a co-worker ate vanilla ice cream at lunch time last week. This is Tim Berners Lee reminiscing about an opportunity to show Steve Jobs his wares.

That's a bit of selective remembering.

Jobs wanted Pixar to be a hardware company. Not a terrible idea as NVidia has shown us, but definitely the wrong time, and he did not foresee the brand that Pixar Animation Studios would become.

NeXT was a 2nd attempt at the overpriced workstation that was Lisa. Both were market failures.

The kind of market failure that got sold for half a billion.

It wasn't really NeXT that they purchased; it was Jobs.

Sure, they got an OS, and they got some good people. But it was basically an acquihire. NeXT had approximately no commercial success. If Apple had only wanted a next-generation OS to compensate for their failure to develop one internally, they could have bought Be, and for a much lower price. There were no other suitors to acquire NeXT; without Apple's acquisition, they surely would have died out within a couple of years, just like Be did.

And I'll note that the OS wasn't that great a deal for Apple; it took them more than 4 years to get a consumer-grade release of it together, far longer than they were predicting at the time. I'd again put that down to the power of Jobs's reality distortion field, not any kind of market success.

>It wasn't really NeXT that they purchased; it was Jobs.

Easy to say in retrospect. But at the time Jobs was the guy who was once ousted from Apple and hadn't done that great of a commercial job with NeXT (besides the technology they created). So there wasn't much ground at the time for Jobs to be considered as some kind of "miracle man" that will save Apple.

So, while we now can say "they purchased Jobs", I think it was mostly the allure of a turn-key next-generation OS that the needed sorely at that point that made them go into it and for half a billion.

Let's not forget that Apple had just killed Copland, and was simultaneously discussing to buy BeOS.

>And I'll note that the OS wasn't that great a deal for Apple; it took them more than 4 years to get a consumer-grade release of it together, far longer than they were predicting at the time.

Again that's in retrospect. At the moment they didn't think "we're really buying Jobs, who'll do miracles and make the company #1 in the world". They thought: "We need an OS, and this NeXT thing looks like it can be turned into the next OS version soon".

>I'd again put that down to the power of Jobs's reality distortion field, not any kind of market success.

But they weren't looking for market success. After all they didn't want to buy an existing userbase but an OS to turn into "OS 9 successor" to give their (Apple's) users and attract new ones.

So, what they wanted was a good platform for what they do. And given Apple's enterprise, dtp and graphics focus in the 90s, NeXT was pretty damn close.

I don't think he was omniscient but there is an interview he did in Wired I believe, around '95, where he laid out the next fifteen years in technology and pretty much nailed it. He wasn't a god. That much is obvious. But he was really smart and the term "visionary" is reasonably earned.


If this was a read-write version, then it's easy to mentally box it into a network-backed word processor (since that's what it would have been most similar to).

The power of the web took quite a few leaps in terms of a distributed effort making it into something new each time. When cast in the original scientific philosophy, it's not clear that this would seem any more important than the prodigies, compuserves, and aols that already existed.

What the heck? The "web" had one of the fastest adoptions I can imagine. People started using it before data communications speeds were truly ready for it.

You copied that binary right? I think from a historical perspective that would be important to preserve...

Looking at the file listing in the terminal in the photo (0_16, 1_0, 2_01), it looks like it's the exact same files as can be found here http://browsers.evolt.org/browsers/archive/worldwideweb/NeXT...

Yep, and it turns out according to the README it's the editor, which is still cool.

> WorldWideWeb_0.16.tar.Z is the only wysiwyg hypertext editor for the NeXT.

No, I had no way of extracting any information from that machine. And also, the last time I had access to that machine, it wouldn't turn on. The reason was actually simple and solvable so I was just sad I couldn't play again with it and relieved that nobody could break it (software wise).

>> WorldWideWeb_0.1.0.tar

Aren't they these? There's a missing leading 0.


The leading "0." is kind of the point of the whole article ;)

The article is misleading, sort of, probably not intentionally. But if you look at the screenshots of the files, they are actually _1.0 and _2.0, but also there's a _0.16 – which is the earliest one he got running. That's available in links up the thread – still a very cool thing to find in the wild.

I'm the author of the original article and it's not my intention to be misleading. This happened to me a few years ago and the few pictures I got are actually quite blurry. My friends, though geeks, were sort of feed up with me and that machine and wanting to document it. I distinctly remember the version numbers I played with starting with a 0 and I remember 0.1 or 0_1 and 0.2 or 0_2 and that the first one I tried, the earliest, didn't work.

Oh, how I wish I had that computer.

What makes us think that Tim Berners-Lee isn't meticulous about keeping good archival backups of his early work?

I am not the author, please contact him to insist about the preservation. (a stupid downvoter was here).

What's with all the secrecy?

That machine, under certain circumstances, can be publicly accessed without much trouble. I shouldn't have access to it. I was careful, I didn't break anything, but someone better than me at this should have done this research.

I'm afraid of malicious or less skilled people playing with the machine and breaking it (software-wise).

The author seems to think that somebody would want this machine so badly as to break into somebody's house?

No, there's no need to break. This machine, under certain circumstances, can be publicly accessible without any problems.

Wondering the exact same thing

I clarified it in some comments.

That is cool! It's also a bit disconcerting that so much computing history is disappearing, and will not be accessible even in the lifetimes of those who experienced it. I imagine there will be digital archaeologists in the future who spend careers finding this kind of stuff and attempting to simulate how it must have worked back in the long ago times...

and black marketeers, and hoarders, and vandals...

Sir Tim Berners-Lee is actually quite reachable; he participates in numerous Semantic Web mailing lists these days.

I don't want to send him a public personal message and the one I sent didn't generate any reply.

Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact