This was super interesting, and it would seem Bill Gates absolutely predicted the effect that not Java, but ultimately interactive web apps in a broad sense would come to have by making it less important to have one operating system or another.
Interesting that the reply mostly implied that there wasn't much reason to worry, and that the future and dominance of web apps was still uncertain. But we are reading this email exchange now I think largely because of how well placed Bill Gates's worries were. I wonder how many technologies he and other executives lost sleep over that did not amount to anything in the long run.
The reply was on point and serious . Mainframe still exists and generates billions in revenue. Windows/ Microsoft lost their foothold in a new market. However Microsoft is also transforming into a cloud first company.
Also the reply gives reasonable time estimates: 20 years. I think this is about right as web apps clear dominance wouldn't be until roughly the mid 2010s anyway
github just removed git: urls from their web ui so it no longer tells you how to make a clone you can push from using standard git
it's been replaced with instructions for using their own gh cli tool (which, however, is also free software)
for example on any github project page, https://github.com/username/projectname, click on the "code ↓" button in the top right; the "local" tab on the left has a "clone" tab with 'https' and 'github cli' subtabs. the git@github.com: url you can use with standard git has been removed
The heuristic for Github though is that the vast majority of people visiting a repository page who aren't logged in to Github on that browser don't have an authorized private key. It makes complete sense to present only options for anonymous checkouts to anonymous users.
Long ago they used to push SSH as the "right" way to pull repositories (I still almost exclusively use SSH), but their change to recommending HTTPS instead by default is by now quite old: it's been the case for many years.
Other related things have changed more recently: Github doesn't allow git:// urls anymore (but HTTPS still allows anonymous checkouts), and when doing authenticated HTTPS, you now need to provide a token instead of your password.
The git:// protocol, to my recollection, doesn't support authentication at all. If you've ever been able to push to git:// URLs on Github, it's news to me.
i wasn't talking about git:// urls and i think that removing them was probably a good idea, for the reasons you describe, but also because they didn't have any way to stymie mitm attacks
but it seems like my original comment was incorrect in a couple of different important ways, and i appreciate people having pointed that out
Given the number of users who are novices to both git and ssh and all the rest, and also given that the UI doesn't really have room for "teaching" (it's already a popup, so adding another popup would be pretty ugly), I think it is a fairly reasonable change.
Clearly, more people were confused by the original than outraged about the current implementation... I don't even recall the change being made.
This may have something to do with one of my other pet hates - projects with submodules using git:// urls. If you don't use SSH keys for github you can't do a "git submodule update --init" without first changing the submodule urls in .git/config to https:// style.
> embrace, extend, and extinguish is alive and well
???
you can clone Github projects and move them on every other git repository you want
if Github removed the support for git, Github would be dead in a week.
Not only because git is literally in their name, but because every programmer tool that deals with code (editors, IDEs, etc.) has embedded support for git, the protocol, not git the Github version of the protocol.
Remove interoperability and you killed the golden goose.
if www browsers had removed support for ftp, browsers would have been dead in a week, in 01998
but because they waited until last year, removing that interoperability did not significantly affect their dominance as a way of accessing the internet
> wrt nat i would say that nat created a lot of issues with ftp
let me rephrase it.
FTP behind NAT had a lot of issues.
But the protocol was never designed to be accessed by a WEB browser, that were limited to read only access.
The best way to access FTP is through an FTP client.
I don't necessarily agree with the new wave of browser security at all costs, like HTTPS everywhere, but FTP clients are far from being hard to find.
> there is still a substantial information that is only accessible through ftp
can you point me to an example?
I haven't encountered that problem since the end of the 90s. The only instance I can recall is upgrading the firmware of home routers/modems using TFTP.
Funny how all my "nerdy" friends love Gates now. I guess 100B$+ is enough to afford PR team that can convince public you are no longer "more evil than satan himself".
Well, Gates is absolutely terrible at PR and doesn't tick any of the boxes of being chic, unless it's some kind of so uncool that he's cool kind of thing.
Despite this and more importantly despite his reputation in business and the often laughable products MS produced, his tremendous philanthropy really changed my perspective on him.
In a world where there are mega-billionaires, he's my least unfavourite. He's doing good things with his money like eradicating malaria and funding bleeding edge climate tech. He's pretty outstanding wrt how he spends his money.
Strong disagree. I started software dev in the late 90's and among everyone I knew in tech MS was uniformly hated, even the DOJ sued them. They did everything they could to stomp out open internet standards. Untold amounts of money, business, etc. were lost by all of us fighting this scourge. The fact that anyone alive thinks anything remotely positive of him could only be due to PR. IMO even if he gave away every cent, I'd just consider that restitution for all of the damage he's done to the world. Then he'd still owe us.
Eh, I would say he puts a lot of effort intro creating image of benevolent nerd philantropist. Point in case: https://www.youtube.com/watch?v=0hnZJIRd6Qw There must have been a team of people dedicated to carefully crafting this video, creating props, there's even a dozen or so actors there (each in the background for just a few seconds). It looks like someone put a lot of effort into this and it surely consumed thousands of dollars (my uneducated guess is somewhere in 10-100k$ range). To what end other than working on his personal brand?
Speaking of malaria, after googling for 3 minutes it seems he spent around 2B$ on it. Does it really matter? Yes and no. Maybe it had impact on people's lives, but given Gates' net value is 100B$+, giving away a billion or two does not matter. It will not affect his quality of life in any significant way. He could give away a few dozen billions more and still will spend rest of his life living in luxury that I could only dream off. Sure, better to fight malaria than buy another yacht, but still, it's easy decision to make when you have that much money.
> Is it not enough for you that he went out of his way to improve other people's lives?
How did he went out of his way?
> How do you know? Have you ever been in that position?
I don't think snark is necessary here; probably there are no 100B$+ users on HN.
But it's not hard to extrapolate from where am I. I have comfortable tech job, I could easily give away few thousands bucks to some charitable cause just to show off (and my quality of life wouldn't change). Would that automatically make me a good and trustworthy person? I think the answer here is no. I also guess if for some reason my net worth went up a few orders of magnitude, I could also give few orders of magnitude more. Would then I be a good and trustworthy person?
> I have comfortable tech job, I could easily give away few thousands bucks to some charitable cause just to show off (and my quality of life wouldn't change).
Yes, you could. Do you?
> Would that automatically make me a good and trustworthy person? I think the answer here is no.
It wouldn't. But also because there are no good and bad people. There are good and bad actions. I use Linux and absolutely hate what Bill Gates was doing wrt the whole free software ecosystem. I still commend him for his charity work.
> I also guess if for some reason my net worth went up a few orders of magnitude, I could also give few orders of magnitude more.
Can't fault him on hiring a crew to craft his videos. He has better things to do with his time. Wouldn't you outsource that if you were in his position?
My 2 mins of googling came up with more like 3.6B$ on malaria, but even 2B$ is a shit-ton of money. I certainly haven't donated 2% of my net value to malaria (and it's only on the order of 1/10000 of his!).
Malaria rates are down around 30% since 2010, thanks in part to Gates. He's likely saved thousands to hundreds of thousands of lives.
Malaria is just one part of his philanthropy. I'd love to see the numbers for Jeff Bezos or Elon Musk. It would be great to all billionaires follow suit with this kind of "easy decision".
2, 3 or 5 billion don't matter if you still have tens of billions left.
I'm not saying he shouldn't outsource his videos. What I'm saying is that this looks more like a professionally made commercial (advertising his personal brand). I'm pretty sure it cost O(hours) of his time (consider even getting to the set - you can see this is a professional set, so probably in some studio; I also doubt this was done in one take given how there are other actors around including the skating Santa Clauses). All in all, it really feels artificial and dishonest. What kind of person would even come up with idea of creating such a video in the first place?
There are somewhat diminishing returns spending money on philanthropic causes: if he were to spend 30 rather than 3 billion the result wouldn't be ten times better. I'm happy he's trying to make the world a better place while still keeping an eye on the money being used efficiently.
No, we've just realized that they're just less evil than all the other current behemoths.
Sun tried to be "cool" but was really just a money loss machine and now Java and MySQL is the property of we-will-audit-you-Oracle and Mr I-will-sue-people-when-I-need-another-yacht-Ellison.
G had a motto of "do no evil", that motto is long gone. I've always disliked their handling of the Android platform (the handset manufucaturer/carrier dominance over application stores back in the J2ME days was horrible, Google's quest for marketshare let that go on for a few more years and allowed for some bad fragmentation for developers).
And this was just the start, their "AI-handles-support-so-we-don't-need-to-talk-to-humans" combined with accounts being suspended willy-nilly just leaves people stranded (anybody remember how an indie guy making an Stadia exclusive game was shut out of his Google account?) and speaking of stadia, what's Googles cancelled product body-count up to now?
Apple has had some pros on the privacy side but they've stumbled even there, and apart from that they've been quite a heavy-handed player with a clear my-way-or-the-highway attitude (vs Epic,Spotify,etc).
Amazon busting unions, fleecing people on AWS and copying popular products on the marketplace needs no more mentions.
The Ubers, Spotifys,etc of the world might've made things cheaper and more streamlined but there's been at a cost of distinct monetary pressure on drivers,musicians,etc.
Facebook facilitating genocides (Rohynga), election fraud (US/UK) and general lax privacy (Cambridge Analytica).
MS might have gotten worse in terms of privacy and ads(start menu ads,wtf?), but in general they've mostly kept to making money and actually supporting developers and opensource.
NPM/Github costs might be fractional for MS but would've probably landed the rest of us another SourceForge (remember the spyware they added to make money? That was probably what got GH off the ground more than anything) because we've not yet sorted a way for developers and distribution to get paid for our "free" contributions.
If Microsoft hadn't existed, everyone would be bitching just the same about whichever tech titan would have formed in their absence and erecting tombstones of their founders. It's not as if any of the other wannabes jockeying for power in that era would have had any hesitation in using their dominance anti-competitively either if they had gotten it. Even a certain scrappy don't-be-evil business that hackers were cheerleading at the time didn't manage to resist the lure of leveraging their market dominance.
The question is whether an omnidominant platform like Microsoft's would have arisen in the first place. While the current Android/Apple situation isn't ideal, it's certainly better than only one of the two. I don't think it's a forgone conclusion that a "Microsoft-esque" monopoly would have happened.
For whatever it worths, oracle has been spectacularly and surprisingly good at shepherding Java since the takeover, they held onto almost the whole core team from Sun (which is not a common thing with takeovers) and the platform is advancing faster then ever (and not in a haphazard way).
Mercifully, I don't have to worry about Larry Ellison every day of the week. He's not a 'oh god what has he done now' sort.
The good thing about Larry is as shitty as his bag of tricks is, it doesn't change much. What people hated about Microsoft is that they innovated on fucking other companies over and putting them out of business.
There would not have been 'a different Microsoft'. It was Gates and Ballmer and friends that did that. Paul Allen got sick and then split with Microsoft to do pretty much completely different things for the rest of his life. Not that people have good things to say in Seattle about his behavior as landlord.
> As you and know very well, this sort of widespread interest can become a self fullfilling phenomenon, because programmer attention creates programs. Some of these will be successful and that only fuels more participation in the phenomenon.
I hope this is what's happening with Rust. :)
I also thought this paragraph just afterwards had an interesting point:
> In the limit, they can make the web totally OS agnostic - but there will still be other things that motivate one platform versus another.
This looks like it happened; but did/do people actually credit Sun for this? On the client side it was Netscape that broke MS' monopoly with Internet Explorer, but on the server side Java was and is still very popular with the business folks, so... does Nathan have a point here?
Internet Explorer dominated for years. Microsoft killed Netscape not the other way around.
Mozilla helped with the downfall of IE but it was really Google that put the final nail in the coffin.
As for backend, Microsoft had some systems running IIS and Apache can run on Windows Server too, but UNIX and Linux was always the dominant platform for web hosting. Java didn’t change anything there.
Javas impact on Windows can be felt more with .NET, which started life as a direct competitor to Java, after Microsoft was taken to court for their own proprietary implementation of Java (if I recall that part correctly). There was a time when you’d have to select which Java engine you wanted your browser to use and Microsoft Java (or whatever it was called) and Sun Java weren’t fully compatible with each other.
> which Java engine you wanted your browser to use and Microsoft Java (or whatever it was called)
It was called JScript. It was released after JavaScript.
I am a peaceful man; never been in a fight with anybody, and oppose the death penalty. The exception is that I think all the people responsible for this clusterfuck of names should be shot.
TBF, when JScript was born, JavaScript wasn't public nor free, it was Netscape's
Javascript came out in 1995, JScript in 1996, the ECMAScript standardization began in 1997 and the first release is from 1998
They reverse engineered the language and named it differently because they had no rights over JavaScript at the time
And good lord they did it, they could have adopted another language as the default one (they had VBScript too) and the story of the front end programming could have been very different.
> They reverse engineered the language and named it differently because they had no rights over JavaScript at the time
The story I heard was that it was the Java trademark that was a bigger problem for Microsft naming it JavaScript. But that's somewhat irrelevant as you're right about the more important point about how JScript was reverse engineered from JavaScript.
> Even MS is not evil all the time
There was a lot of differences Microsoft introduced in JScript and as you said, Netscape submitted Javascript to ECMA for standardisation only a year later yet MS still pushed ahead with IE-specific language features to lock people into IE.
> yet MS still pushed ahead with IE-specific language features
Yeah, they all did though.
I was doing frontend dev back then, if we can call it that, it was web development with some JS actually or DHTML for those who remember.
Scripting elements was very limited at the beginning, you couldn't access the entire DOM, only forms and images, if I remember correctly, and maybe the decoration of links, that's how the hover behaviour was born.
The story is quite more nuanced, there's also the fact that IE was actually a better browser than Netscape after the Netscape 4 fiasco (I was a paying customer of Netscape 3 gold and Netscape 4)
Netscape tried the non-standard layer element accessible through document.layers (all the layout was done with tables back then, so this looked like an improvement) but didn't catch up.
IE 4 had a lot more features, it introduced document.all to access the entire DOM by ID, documents.forms, document.images, document.all.tags() that allowed the access to all tags by tag name, it introduced innerHTML, outerHTML, innerText and outerText properties.
It's easy to understand why developers switched to IE, given also that Windows was the primary platform that users used to browse the web (do we still use the browse verb?) and IE4 was available on the Mac too. IE5 introduced the XMLHttpRequest object so AJAX was born (IE 5 on Mac was even more advanced than the one on Windows)
And yes, MS abused of their dominant position with IE6, but that doesn't invalidate the fact that they actually innovated in the field, before abusing of it.
EDIT: it wasn't called XmlHttpRequest on IE, you had to instantiate an Activex Object (COM component) the MSXML2.XMLHttp. Forefox came up with the XmlHttpRequest name later (XML was still going strong at the time). You could actually instantiate any Activex component available on the client machine from IE >= 5, even Word if you wanted to.
I remember developing intranets for Windows networks where we used Word automation from the browser using Activex.
To think about it now it was a terrible idea, but it looked like a miracle back then. You could instantiate the ADODB COM component and run queries from the browser to the SQL server.
Yeah MS was hot on COM back then. Even JScript was a COM object IIRC.
Agreed Netscape also introduced their own incompatibilities too, but Netscape didn’t have a desktop monopoly. Though I do agree that IE4 was actually a really good product. If I recall correctly even IE3 loaded pages quicker than Netscape Navigator 3.
I really wanted to like Netscape Communicator but it was hard to overlook just how sluggish it was compared to IE.
As for front end development, I mostly an sysadmin back then but I did have a few websites live (my first being publish in 1994). I even won in the JavaScript category of Planet Source Code (remember that place?) one month for a recreation of the Windows desktop using JS. But I much preferred writing desktop software to web pages (and I still don’t really enjoy frontend dev all that much even now)
For those that have been living outside Windows, MS has been hot on COM since OLE 2.0 days.
Not only are many .NET bindings for Win32, largely based on top of COM, WinDev has doubled down on COM since they rebooted Longhorn ideas into Windows Vista, and WinRT is nothing more than COM with another set of additional interfaces (TL;DR version, it is a bit more actually).
> Microsoft was taken to court for their own proprietary implementation of Java
That's correct, they tried hard to fracture Java by making different implementation with their own classes and packages, runtime behaved a bit different too. While whole sales pitch for Java was to write and compile once, and run reliably anywhere (which was the case for every C/C++ app especially UI).
Embrace part done, extinguish attempted. Everybody was very glad seeing them get burn, courts did their job and MS rolled back any Java participation. The rest is history (and present)
> While whole sales pitch for Java was to write and compile once, and run reliably anywhere
which was never true, BTW
There are few things in the story that are simply corporations VS corporations and not good vs evil.
Sun wanted to completely control Java and they sued MS for not completely adhering to the standard they imposed. It wasn't simply a different implementation, it was a Windows only implementation, the fastest Java implementation on Windows plus some Windows only extension.
Which, honestly, wasn't that bad, considered that Java applications on Windows started to be a viable option almost 10 years later. Those who remember the first editions of Eclipse know what I mean. Many java programs were tied to specific OS implementations anyway, especially UI, Swing was always quite out of place on every platform.
Anyway, long story short, Sun sued MS for not respecting their standard, but then Sun almost killed Java because the development stagnated for a very long time, features were not coming, bugs weren't being resolved and they mostly made a big ball of corporate ready mud.
Their inability to handle the success made .Net a much more viable platform on Windows first and *nix much later, which otherwise would not exist.
Visual J++ was probably the best Java IDE at the time, probably second only to IBM Visual Age Java for OS/2, that was born for SmallTalk and had a lot of the SmallTalk vibes.
I believe, Javas impact on Windows can be felt most with Android. Android's development began in 2003 first as OS for digital cameras but they pivoted in 2004 to smart phones. At the same time, in 2003, Windows Mobile launched, which goes back to 2000 (Windows Pocket PC) and 1996 (Windows CE). Windows Smartphones were actually a thing used in practice before Android and iOS. Personally, I had a T-Mobile MDA II, for example. Impressive foresight by Bill Gates. "Only the paranoid survive" (Andy Grove) comes to mind.
I think it’s a bit of a stretch to say Android has had a bigger impact on Windows then .NET, but each to their own.
It is also worth noting that Android is Google’s own implementation of Java and has nothing to do with Microsoft’s code (J++) nor Sun/Oracles either. So regardless of opinions about “impact”, I’m not really convinced Android is even in the scope of this conversation.
The last release of Navigator was in the 90s. There were still Netscape branded browsers after that but it wasn’t called Navigator. However the market share Netscape had at that point was virtually nothing compared to Microsoft/IE and nearly all sites were written to support IE. For all intents and purposes Microsoft had won the browser wars.
> On the client side it was Netscape that broke MS' monopoly with Internet Explorer
What? Microsoft completely obliterated Netscape, which was dominant before IE. In the early 2000s IE had like ~95% market share. Later Mozilla Firefox turned that around, but only after they made the steaming pile of shit that was Netscape vaguely decent, which took a few years.
Yes, Netscape took them to court, proved monopoly by then it was too late. Microsoft had already took over. Bundled IE with Windows (it was first an add on) and added a deal with Apple to include it in their OS. It was Chrome that tore it down. Which was a good thing and a bad thing.
> First, the excitement is overblown - at least from a business perspective. At the moment Java is expanding into a vacuum. It allows you to make cool web pages, and that is a very attractive thing for people. It gives programmers something new to learn, book people something new to sell books on, software tool companies a way to issue new development tools etc.
This is a pretty interesting take, and it feels relevant to this day with respect to both frontend and backend frameworks. I always thought of frameworks as just fads where people simply enjoy to try new ideas and are just having fun. I've always agreed with the overall sentiment that, for example, no one will be going out of business just because their frontend isn't React, but it's interesting to consider that people also profit off of these "fads" by selling books and tools, resulting in a sort of perpetual motion machine.
> no one will be going out of business just because their frontend isn't React
Yes.
But one is more likely to go out of business if their frontend is still using JQuery. It's harder to maintain and iterate, harder to recruite good engineers.
I would ask for more money if the company hiring me uses JQuery and doesn't use TypeScript/React. And if the company somehow hires me, I would be less productive because I can't use TypeScript and a modern framework.
What if they use typescript/Vue? I work for a fortune 100 company that uses Vue. We have found that people familiar with react can pick up Vue very quickly. That hardly seems like a reason to ask for more money.
My experience with True React Frontenders is opposite, except for more money nuance. I don’t even feel that “no one will be going out of business” vibe, really (but that must depend on a type of a business).
> it's interesting to consider that people also profit off of these "fads" by selling books and tools, resulting in a sort of perpetual motion machine.
> It allows you to make cool web pages, and that is a very attractive thing for people
This is the most misinformed take on what Java is used for I have seen on HN. Today Java is used by FAANG and other tech shops. It's employed in the enterprise, science and finance. What it is almost never used for is creating "cool web pages"
Besides the point that it is 1996 and probably talking about applets, wouldn't spring boot backends fall into the modern interpretation of "mak[ing] cool web pages"? Sure, java is used for more, but the number of job ads led me to believe it is quite a substantial part of java development.
FAANG is basically just creating cool web apps, no? Point taken on science though. Also I think Java was dog slow at the time. Eventually hardware caught up.
I remember applets being used for roll-over buttons on web pages and always knowing that an applet was being used because Internet Explorer 4 hung when loading the page, waiting for the JVM to fire up (with much disk churning).
> New things are NEVER a threat to the old world as soon as people say. Look at the mainframe vs PC. It has taken us TWO DECADES and even after all that IBM still has billions in mainframe revenue. Cool new technology always expands rapidly into NEW areas (where there is a vacuum). Pundits always say that this is going to kill the old businesses - eventually that happens but not anywhere near as soon as they say.
There's a tendency to overestimate things, but not so sure about the underestimate part. Unless I missed something I don't think there are many people running NoSQL databases on Itanium processors with GNU Hurd powering their XSLT/Xforms website.
(NoSQL is still useful, but it didn't "replace SQL" as promised, and the others have completely failed in spite of promising much).
It's always hard to predict the future; often the first iteration of something new isn't very good or even downright bad, and it's not easy to predict if it will ever be any good. I remember trying Rust somewhere in the early 2010s "before it was cool" and I hit some bug in the installation (would eat up all my memory). Then I tried the new version that fixed it and I ran in to problems actually compiling programs due to another bug. This doesn't reflect the status of Rust today at all, but back then it was essentially impossible for me to tell if it would always remain the buggy mess it was then, or if it would be able to put itself together and become stable. There's been plenty of buggy things that never became stable (Itanium, Hurd).
I think it's more that technologists tend to underestimate the social factors, edge cases, general "stickiness" of this stuff. We see the destination and think "ah, well the technical problems have been solved, we can all start using the new thing right now and it's clearly superior, as everyone can see, so surely that will happen very quickly"
When in reality there's,
- People who aren't convinced yet
- People who don't want to be convinced (because of comfort, hubris, or livelihood)
- Human processes and skills that have to be adapted to a new way of doing things
- Material cost of making those transitions, which has to be justified to those making decisions
Nathan Myhrvold answer is extremely impressive. There is a lot of extremely accurate forecasting in this email from the cheapening of the value of the OS to his extremely interesting idea about storing files on the internet with the File API which is more or less Dropbox before Dropbox.
Reading that you realise how wrong a choice it was to put Ballmer in charge and how much it cost Microsoft.
Filfre.net / The Digital Antiquarian has a two-part (and ongoing) series of articles on this era in regards to web browsers and the perceived threat of Java to Microsoft (specifically, the idea that was floating around that users could have access to all the software tools they need via a browser and JAVA: no need for MS tools or software, etc).
> The new Java applications are NOT credible threats to traditional PC software any time soon
Well, Nathan was right there. But even with Java as (depending on where you look) 2nd, 3rd or 4th popular programming language, how many of those applications are being run on Windows hardware? Developed maybe, but deployed? Ultimately Microsoft didn't get stung too badly by this, since they embraced Linux on the server but it's weird to see them talking about how it's "expanding into a vaccuum" so early, batting around what to do about it when we know now that ultimately Windows lost that battle.
> Continue to "embrace and extend" - both at the level of new Java tools (like J++), and our broader browser strategy.
Kind of an "Oh my god, he admit it!" moment, it's wild to remember how open they were on the EEE strategy.
And nowadays they are an OpenJDK contributor, were part of the companies doing the ARM port, and develop Java support for VSCode in collaboration with Red-Hat.
Meanwhile the chocolate factory has created their own version of ".NET" for phones.
Oh man, VR and metaverse-style concepts were big in the 90s even if the technology obviously wasn't quite there yet. Active Worlds predated Second Life by several years and surprisingly enough is still online to this day.
VRML was a technology that was way ahead of its time. Hardware of the time just could not deliver a good experience with it. Even high end SGI hardware (the land of VRML's birth) struggled with it. By the time commodity hardware got to the point it could deliver, interest and support of VRML was long expired.
Contrast VRML to a game like Quake. One of the key optimizations in Quake was a very computationally expensive pre-processing of the game map to determine surface visibility. The map environment was also a simplified set or surfaces rather than a giant tree of arbitrary primitives.
VRML had almost the opposite Design leaving the client to do a ton of work just to construct the usable representation of the virtual environment and then figure out what to draw. It also didn't help the client software often had terrible UIs and were implemented as browser plugins.
I recall watching on TV a VR demo with sensor gloves connected to a Mac where the demonstrator would grab and manipulate objects in 3D. That was in the mid-80s.
I hate how "metaverse" became such a mainstream word in the last couple of years but 99% of the population has never even heard of Snow Crash, which originated the term.
My first successful company sold EJBs. It was a nice hype; simple to get into and go from sitting in your underwear in the attic to 10s of millions of revenue and contracts and sitting at Fortune 1000 corps with the cto. And of course learning (me as a university student and geek still) rapidly that enterprise software is not about software but about process, support, sla, contracts, consultancy (we went from 400 devs to 400 IT consultants in an afternoon ugh), continuity, XML and communication.
>that enterprise software is not about software but about process,
True that ! NOTHING moves as slow as a B2B software contract, it can take 6+months and even sometimes years to go from proof-of-concept and golf-talks to actual first invoice being paid to you the provider.
Basically data objects loaded via orm, with code to repersist to database when updated.
... except there was a long road of transactions, distributed architecture issues, etc still to hammer out, and these were the days of a gig of ram maybe and no ssds, and 100 base t if you were lucky. And jit was a wip.
So it was a app server with its own dedicated mumbo jumbo for everything to dazzle CEOs.
You are talking about JPA entities, javabeans are nothing more than objects managed by an application server (or nowadays without them as well with spring boot), whose instances can be queried and loaded dynamically. It is basically a huge map of strings to java objects, and objects range from database connections (which were then configured outside of the program, making configuration decoupled, e.g. at deployment one could even change connection pooling, database vendor, etc) to remote objects on another computer to ldap to.. well, anything.
Nathan Myhrvold's long answer was more interesting. He was talking about the future of the Internet and what would make money.
Now 25 years later, watch what's actually made the most money, and did his reply cover those. I can think of hosted services and search.
Even Microsoft had one of the very first maps online in the 90s [ terraserver ] but they failed to realize how large this business eventually will turn into.
I loved playing with Terraserver back in the day. However it was built to be a demo rather than a product. Their GIS products were Streets & Trips and then MapPoint which then rolled into an online version that went through various MSN/Live/Bing rebrands.
I don't think that Microsoft failed to realize how large a business mapping would become. I think the problem was mostly these products existed in the Ballmer era where Microsoft expanded into seemingly every niche without clear goals besides using up all the oxygen in the room.
Folks underestimate the important role Java had in saving open systems and Unix, and then enabling Linux on the server side. Around 1996, Windows already was the platform of choice for most new projects on the server side, after totally dominating the client side (desktop).
Oracle had already acquired Sun by the time I was in the technology industry. The respondent states that Sun is so set on giving things away. Can someone comment on the culture of Sun at that time?
Sun seems to have left a legacy in the open source community. What was the business logic on giving away the software for free? I know it's very common today, but Sun may have been a leader in open source. Any books recommendations would be appreciated too.
I started in tech in the late 90s. Here's what I recall:
1. Sun originally meant Stanford University Network and the university shared the open-source ethos of Unix along with UC Berkeley.
2. Although Microsoft was already dominant on the desktop by then, the Internet was still up for grabs.
3. And even though JavaScript shipped with Netscape (meaning almost from the beginning of the public Internet), it was still severely underpowered.
4. Java, in the form of "applets" (mini apps running on the browser), could add dynamism to HTML (CSS didn't exist yet) and so define a new software paradigm for the Internet.
5. Sun, like Apple, made money on hardware.
All five factors incentivized Sun to give Java away. Java applets, although a cool demo for the time, never really went anywhere, but Java did go on to compete with Microsoft for the (enterprise) server.
Giving things away is not the same as open source. Java was closed source until 2006. Sun released the Java runtime and class library for no charge but not the source. The class library's interfaces were public as was the JVM specification so there were open source Java implementations like Kaffe, gcj, and the GNU classpath before Sun open sourced Java proper.
Originally Sun released the JRE (on PC/Mac/etc) at no charge but charged for licenses elsewhere. With the release of Java 1.2 (J2SE) Sun charged for releases of J2ME on devices and J2EE products like application servers.
Sun wanted to get the Java runtime spread as widely as possible to make it the platform to develop stuff on. The idea was to get stuff written in Java and then deployment was just embedding a class file (later a jar) in a web page.
I can recommend the book “Steve Jobs and the NeXT big thing”. The Author compares NeXT a lot with Sun including their business strategies. I have never thought I will learn so much about the early days of Sun by reading this book.
In a way they got to be interwined via the OpenStep and Objective-C collaboration projects, that ironically are in the genesis of Java and also Java EE, hence why Java has so many Objective-C touch points.
Having listened to "oxide and friends" and "on the metal" podcasts.. next and sun were nothing alike.
The above podcasts are hosted by Bryan cantrill and often has guests from that era, though the newer episodes try not to reminisce so much, because it became a bit much at some point
Totally agree, which is the reason why I recommend the book. The author explains the difference in a great detail and by that you learn so much about Sun. …and NeXT or course :).
IMO it was a strategic move to commoditize their complement. My impression during that era was that Scott McNealy saw Microsoft (whose strength was in software) as Sun Microsystems' (whose strength was in hardware) most important competitor and took steps to stem their influence. Java was one example of a giveaway intended to create a Sun-controlled moat to protect their business. The purchase of StarOffice and its open-sourcing as OpenOffice, which eventually begat LibreOffice, seemed to be an attempt to put a dent in Microsoft's MSOffice revenue.
If you sell hardware the last thing you want to do is commoditize hardware, which is what a portable software strategy does. What you want is hardware lockin. Java was the opposite of that. Sun eventually had to be sold, but Microsoft didn't.
Sun intended Java to be an uber-platform, to get it running on everything from tiny embedded systems to their giant E10k systems. They'd make money selling JVM licenses in various verticals.
Sun bought into Netscape's vision of the web as the delivery platform for software. Java could run on the server side and then as applets in the browser on the client. The expansion of Java would then expand its use in places Sun wanted to charge for licenses.
Java wouldn't create a moat around their hardware. It was intended to eliminate Microsoft's moat around the Windows platform. If ISVs could write software in Java that could be delivered by a browser on any OS, the actual OS stopped being important. As long as there was a browser and JVM available on a platform it could run that software.
Sun's business model was selling hardware: beefy workstations that you do Serious Computing™ with, unlike those wimpy home computers like PCs.[1] Later on, when x86 actually became decent and competitive they shifted to server hardware, but still: monetary units for hardware, not software.
And oh yeah, with this hardware you also got this SunOS Unix thingy. SunOS was very much core to Sun's success especially in the 80s (not the least because Bill Joy was already a "big name" even before Sun on account of creating BSD and vi), but it wasn't their core business. From their perspective they weren't really "giving things away for free" because they weren't selling it. Interoperability also gave Sun advantages as their SunOS/Solaris machines could communicate with other systems so it made the barrier to adoption/migration lower. People talk about this WSL thing on Windows now, but for Windows NT and 2000 there was SFU – "Services for Unix" – which pretty much did the same thing for Microsoft: make it easier to use Windows with other systems.
Microsoft in the 90s on the other hand was very much in the business of selling software. I'm not sure if they were making any hardware in the 90s (maybe some mice and keyboards?), but if they did it was a little side-thing and not their core business.
Sun was never really an "open source company" in the way we would understand that now: parts were open, many parts were not. They did eventually release OpenSolaris but that was much later. This wasn't entirely Sun's choice: the legal situation at the time was complex, and even OpenSolaris wasn't 100% open due to legal reasons (see the BSDi lawsuit, or the SCO lawsuit that is hopefully finally dead once and for all).
It's similar to Apple today; you can still download the source for Darwin, Swift, Safari, clang, and some other things. What does Apple have to gain by keeping these closed? It's not their core business: no one "pays" for Safari or even macOS; they pay for iPhones and AirBooks. There is still a lot of "secret sauce" at Apple – as there was at Sun – and you could argue if that really protects Apple's profits or not – but there's a wide agreement that "some openness" doesn't really hurt their bottom line.
It's not a coincidence that Microsoft has become "more open" as their business model moved away from almost exclusively selling Windows, Office, and other software suites to selling Azure Cloud, Office365 subscriptions, X-Box, and other things.
In short: Sun never really "gave away" things, they just had a very different business model than Microsoft. On one hand you could say that the author of the email doesn't really seem to appreciate that, but you could also say that Sun failed to foresee that their core products and business model would slowly become obsolete. This is easy to argue with the benefit of hindsight, but less obvious in 1996: they had $7.1 billion revenue/$476 million net income in 1996 (Microsoft had $8.6 billion revenue and $2.2 billion income in the same year; less overhead if you distribute only software) and in 1997 this increased to $8.6 billion/$762 million. The profits continued to increase until the dot-com crash, which they recovered from only to be done in by the financial crisis.
In fact, something that newer generations might miss out is that Sun was the first UNIX vendor to create the separation of user and developer SKUs, which other UNIX vendors were rather quick to adopt.
It was this split of the UNIX SDK into a separate commercial package that eventually fuelled the GCC contributions, until then largely ignored.
A friendly reminder that what CEOs of large corporation focus on is often against the best interest of the user. I had the same feeling reading one of Andy Grove's books. He treated everything I perceived as positive for the customer (such as commoditization of hardware, rise of thin clients etc.) as a threat.
If history taught me anything it is that this attitude is not history.
Well, big companies are just paperclip optimizers without any moral feelings. Moral looking actions are just calculated profit-maximizings (e.g. putting out a rainbow flag will likely have better profits long term than not doing so in case of user-facing companies).
If your current OS is the market leader why wouldn't you lose sleep over it. They are running a business and have to worry about the competition. Progress on competing products might be good for consumers but not for the employees or shareholders of the company.
If you are afraid of progress then you aren't the market leader because you aren't leading. You just happen to be the biggest company currently on the market.
Why not expand on the things where portable software is weak, like making a well-integrated system? But they missed it and let Apple fill that space.
Why not accept the future you predicted and be the first to arrive there? But web development only considered Microsoft an annoyance for a long time due to Internet Explorer.
Fear is the first step in taking steps to compete. Its an acknowledgement of the threat from a better product. If any company doesn't have that they will fail. This can be seen during the ballmer era where he totally ignored the threat posed by iPhone and ended up conceding entire mobile os market share to Apple and Google.
Whenever questions like why are they afraid of progress, it's for common good etc come up (which I feel are based on morality) , it makes sense to imagine yourself in their shoes and see what you would do.
This did end up happening though (although not so much due to Java as to JavaScript etc.).
It was important for Microsoft to anticipate this and pivot from relying too much on Windows on the desktop. And they tried pivoting left and right with the Xbox, Azure, .NET, their early tablets (which Steve Jobs famously discounted because they needed a stylus) and so on.
In fact, one of the reasons Windows is getting so messed up is that everyone that cares about their career is now on those business units, with Azure playing the role of "Cloud OS".
Leaving Windows to younger generations without background on why things are the way they are, and feel like they need to rewrite everything to leave their mark.
The point is that he is worrying here not about improving their own things to compete, but about barrier for entry being low which means fast progress. Gatekeeping is bad, business owners or not.
Strategy to nullify competition instead of competing is disgusting and damaging to progress and market itself.
Is there a strategy to nullify competition in the mails? They simply stated that they are worried about the coming competition, which is completely normal for any business.
Given how anti-competitive MS was in practice, I totally read it all in a negative way. Especially when they mention "embrace and extend" in the same discussion.
He didnt figure out the concept of online subscription and ads. It shows he wasnt that revolutionary. Just lucky, born at the right time to the right family and lived in right neighborhood/country.
It’s interesting how the tone and culture of corporate email hasn’t changed in (almost) 30 years. Change Java for some other technology and I feel like I could have gotten that email in my inbox today.
Surprised this point not realized or discussed more in this thread. Android is famously claimed by linux fans but for all practical purposes its a Java device which factually made this technology among the most impactful on the planet.
I was using Java on Symbian devices already, Blackberry was JavaSE based, and if it wasn't for how Google ripped off Sun, I would probably be using a SavaJe OS device.
> Cool new technology always expands rapidly into NEW areas (where there is a vacuum).
We often hear startups with mission to kill and goliath company slow moving slow. The email shows how it is actually impossible to do that and you have to create a market to grow.
I am fascinated no one has commented on the last point of Myhrvolds reply, which was prescient and seemingly ignored. It was almost like a playbook for what MacOS did.
The tldr:
We must not lose Windows on Desktop. To do so:
1) have the best audio and video tools and multimedia experience
2) we should be bug-free and boot fast
3) we should solve desktop storage with transparent cloud storage
It’s funny this omen explains exactly one way they took their ears off the ball and lost the desktop, their core business. This email wasn’t just about the web as an os, but also the end user desktop experience.
I had never heard of Nathan Myhrvold before, but, his lucid descriptions of the nature of technology has me convinced he's a genius.
Incredibile career he's had. Went to the best schools, published in top scientific journals in multiple foods. And, he took a leave of absence as Microsoft CTO (around when these emails were written) to go to chef school in France. He even has cookbooks!
How does it prevent one to put caps manually? I too have disabled mose auto-things on my smartphone yet I use caps. It is not a technical proble, it is about decency and respect.
Has anyone ever used the cross platform capability of java? Java is cross platform in the way c is cross platform, you can compile to many target platforms. But aren't pretty much all java applications intended to run on one specific platform only? Developers like the idea of cross platform but they pretty much never use it. I think javascript is the only language that is truly used in a cross platform way.
Java + Swing was doing cross-platform desktop apps long before Electron was a thing; my editor of choice for years (before moving to Sublime, then Atom and VS Code) was a Swing-based cross-platform editor called JEdit. And the JetBrains IDEs are still Java/Swing applications.
(There were other options out there, too— for example, SWT’s the widget toolkit used by Eclipse which feels a bit more native on some platforms than Swing.)
Minecraft. Macs were always missing games, and Minecraft was "the only game" that ran on MacBook without Bootcamp in late 2010s. Other than that, I'm sure have seen it happen but not sure...
Java is the only language I have seen reliably developed from multiple OSs as a team. Even JS fails at this, as plenty of its packages are wrappers over native tools, which more often than not fail to run/run the same way on different OSs. Java on the other hand is basically 100% pure, dependencies/tooling doing FFI is very rare
Exactly, that's the "c is cross platform" way. What java offers is that you compile once and you can run the same binaries on any platform (hence freely switching OS as Bill Gates feared). But who uses that in real life?
Because compiling on windows and deploying on linux/unix/etc, that was posssible 20 years before java in c/c++.
> Because compiling on windows and deploying on linux/unix/etc, that was posssible 20 years before java in c/c++.
"20 years before Java"? That's 1975 we're talking about.
Was it possible to build a program on Windows 95/NT and then deploy that to a Unix system? Probably. But given that today this can still be quite a pain to cross-compile C code I bet it wasn't easy. Besides, "it compiles" doesn't mean "it works" or "it works on this other platform" (quite likely with a different CPU architecture), especially not in C. And few people were writing any tests in the 90s.
Java abstracted a lot of details like endianness, platform-specific APIs, etc. away.
> who uses that in real life?
There's quite a few cross-platform Java desktop apps; the IntelliJ IDEs are a good example (although they've since developed Kotlin, it's still the same Java VM). DBeaver is another example, openoffice.org used quite a bit of Java back in the day (although I think that's mostly phased out now) – I'm sure there are more.
And of course in the 90s you had Java Web Applets.
Java is a bit less popular for this than it was, but I've run plenty of Java programs on my FreeBSD machine back in the day, and I'm pretty sure "FreeBSD" wasn't really something the developers really thought of when they wrote it (I can't remember which off-hand, it was a long time ago but I have "fond" memories of dealing with the manual installation bruhhaha because of the license at the time).
It's not like POSIX was universally adopted the day it was released; that took a long time. And not all systems were Unix/POSIX in the first place (e.g. Windows, VMS, the IBM s/3.., and on the consumer side Amiga, Mac OS, BeOS, etc).
There are a few examples of where this is done.
One that comes to my mind would be JOSM: https://josm.openstreetmap.de/
You download one JAR file for any platform you intend to run the application on. There are installers for convenience but the JAR file works just as-is.
We use it all the time, the JARs compiled on Windows get deployed on the cloud without changes, instead of microservices one gets modules (binaries) that consumers cannot mess with unless they are into binary hacking, compiled just once.
Nowadays we do the same with .NET code as well, when using .NET Core.
Yes! Especially in the 90s. The multi-platform story for most things was SO BAD for SO LONG. Now we have better toolchains and we can deploy to most systems easily, but that's a historical aberration. We spent a lot longer struggling in the dark.
That's not to say that Java provided a "good" cross-platform experience by our current standards - but the fact that it provided it at all was huge!
All of my Java backend applications can run seamlessly on Windows and Linux. My small desktop apps also run on Linux and Windows just the same. (with small exceptions)
I can reuse my Java code on Android, but since approach to UX on mobile is completely different, the UI front needs to be separate anyway. But I once wrote a game with LibGDX that essentially had the same core codebase.
C is not cross-platform; you need to compile for each target platform. Java compiles to byte-code which runs on a Java VM, of which there is an implementation on different platforms; you only compile your program once and it will run anywhere there is a JVM. For console/server/webapps applications, this is pretty easy, but for GUI applications it's a bit trickier as each OS has a different look, which is more difficult to abstract. So some applications might have the core running in Java, and a native UI wrapper that provides a native look and feel. But they can be completely Java based (and it's gotten a lot better at dealing with multiple GUI system over time).
While they run on a different VM (Dalvik), Android programs are (mostly) written in Java, and provide a cross-platform function - namely different CPUs (Android can come in ARM, ARM64, x86, x64, MIPS), and all run the same binary (except if there's native extensions... which negates all of that)
Interesting that the reply mostly implied that there wasn't much reason to worry, and that the future and dominance of web apps was still uncertain. But we are reading this email exchange now I think largely because of how well placed Bill Gates's worries were. I wonder how many technologies he and other executives lost sleep over that did not amount to anything in the long run.