Slashdot Mirror
How Moore's Law Saved Us From the Gopher Web
Urchin writes "In the early 1990s, the World Wide Web was a power-hungry monster unpopular with network administrators, says Robert Topolski, chief technologist of the Open Technology Initiative. They preferred the sleek text-only Gopher protocol. Had they been able to use data filtering technology to prioritize gopher traffic Topolski thinks the World Wide Web might not have survived. But it took computers another decade or so to be powerful enough to give administrators that option, and by that time the Web was already enormously popular." My geek imagination is now all atwitter imagining an alternate gopher-driven universe.
early web wasn't worse than gopher or ftp.
I'm pressing ESC twice to access this damn BBS.
or try their hardest at least.
Even if the Web had been stunted by throttling, the demand for multimedia content would have eventually driven the rise of the Web or at least a super-Gopher.
Javascript + Nintendo DSi = DSiCade
Even if Gopher had dominated due to filtering (a premise I don't agree with), multimedia capabilities would have eventually been added to the protocol out of demand. We'd have the same web we have today.
If Gopher had won we would have had more a focus on content than presentation. I hardly think this is a bad thing.
XML is like violence. If it doesn't solve the problem, use more.
Oh, not the little, brown, furry rodents.
Thats what is really stopping me from getting an iPhone, because I can't access badger-net.
People think that if Person X hadn't been around we might not have Technology Y. Okay, this is based on the idea that somehow Person X has some unique ability and only Person X can create Technology Y. Hate to break it to you, but you're not special. Neither is Person X. Second, the reason we have Technology Y is because we needed it. If those needs haven't gone away, then the pressure to fill that void remains -- and somebody else will come along and fill it eventually. Now you're right that maybe Betamax might have beaten VHS if not for a disturbance in the force, or it would have been HD-DVD instead of Bluray, or whatever... But we'd still have high density optical media. Gopher would have died simply because it didn't meet the needs of the population. Maybe it wouldn't be HTTP that replaced it five, or ten years later, but something like it would have been created.
#fuckbeta #iamslashdot #dicemustdie
I remember Gopher. Used it a bit, when I first got online. The WWW was to Gopher as Web 2.0 is to WWW. Really. The web was a natural progression of improvement from Gopher. It was wasn't called Gopher 2.0, much like Windows 95 wasn't called Windows 4.0. It was a new version, and somebody though it woudl be good to give it a new name.
"In the early 1990s, the World Wide Web was a power-hungry monster unpopular with network administrators"
As I write this, Firefox is using 300mb of ram and 100% of one core, so not much has changed since then.
Wasn't it really at heart a search engine? In a Gopher world there would be no Google. And it sounds like what it does 'go-fer' instead of a marketing name.
Fifty years of Yippie! 1968-2018
if gopher had won we would have ended up extending it to be able to embed references to ftp or tftp hosted files, and tftp would have been an important part of the internet instead of a rarely used protocol.
the only difference users would see would be that the text of a page would load first and URLS would all start with gopher://
Snowden and Manning are heroes.
Imagine what the Web could be like today, if it had not been for Microsoft's anti-web-standards Internet Explorer and Macromedia's CPU-wasting Flash.
Every time I have to sit through a bunch of crappy Flash or out of control javascript, I find myself wishing I could get a decent gopher feed.
They condemned the web forever!
Gopher was JUST FINE!
NO SIG
My gopher pr0n!
NO SIG
I can't help but hear Kenny Loggins singing. It's alright. When I saw the title.
Your time spent imagineering alternate universes clearly shows that you have too much time on your hands.
This article doesn't take into account the idea of porn driving many technological innovations. Gopher-web might've lasted longer if admins could throttle the WWW, but Gopher isn't much for porn..
Early 90's computers...486DX2? Pentium 90? That's enough to route much more traffic than any of the nodes at that time could even conceive of, and can do QoS to boot.
Doesn't Gopher run on port 70, making it easy to prioritize over port 80 traffic? It would seem (although I could be completely wrong) that the biggest holdback wasn't hardware, just that QoS hadn't really been brought to fruition in time.
Oh, you're not stuck, you're just unable to let go of the onion rings.
I doubt gopher would have met the needs of the internet as well as the web, and would have been sufficient. The combination of HTTP and HTML has been proven to be enormously successful. Gopher would have needed some major work to make it as flexible as HTML. Web would have probably have replaced Gopher in any case. The design of the web is more practical and better thought out than gopher.
Gopher web is a series of tubes^D^D^Dnnels.
I don't see tftp as ever having been an important part of the internet over long-haul connections. Tftp would have been what it was intended to be and is, a very straightforward protocol that can be implemented with incredibly tiny footprint with little risk of getting it wrong. Notably:
-TCP is *much* better at reliable communication without penalty. TFTP is intentionally dumb, send a block, ack a block, send a block, ack a block. Again, easy to write, horrible performance. In TCP we have adjusting window sizes and partial acknowledgements and all sorts of features where acks are not required as often and data retransmit on fail is more granular. You can implement a UDP based protocol with some features, but it would no longer be remotely like tftp.
-TFTP has a 16-bit block number field. That makes for some tiny filesizes unless you have ludicrous block size. If you have ludicrous block size, any single packet drop would require retransmit of the entire thing. This could have been increased, but not without breaking compatibility and effectively making a new protocol.
XML is like violence. If it doesn't solve the problem, use more.
Full story available at gopher://gopher.slashdot.org
Some of us don't like where the bloated, cpu and bandwidth wasting internet is heading. A world where gopher survived and flourished doesn't sound all that bad.
---- Booth was a patriot ----
Am I the only one who found this quote from the article ironic:
"Although it now stretches to fewer than 100 sites there is still fun to be had in the text-only web, providing your web browser still supports Gopher."
I think the author of this article either doesn't know what he's talking about or got confused. Gopher sites are not part of the web at all, and definitely aren't part of some "text-only" web. Maybe text-only gopher, but definitely not the web.
Phil
I've wondered this often, and often looked up articles about it. However, I'm still stumped. Yes, I understand that it is a "network of networks" - but how does it work? What the hell is a DNS server? I get how my home network works (vaguely) - IP addresses/subnets are assigned, and the computers communicate with the router and vice versa. However, once you extrapolate this to the web, I'm lost. Could a helpful slashdotter please give some sort of explanation? The Wikipedia article is kind of over my head in some spots, and completely unhelpful in others. It'd be real helpful to read an explanation from a real person.
So in other words, Net Neutrality saved the web?
to make Gopher rule? Aha! Proof that network neutrality builds a better network!
OTOH...
I sort of liked Gopher, and now it's gone.
OMG! Network Neutrality killed Gopher!
after seeing it on a secretary's desktop at NASA in the early 90s. My comment was very close to "Yeah, but I can already get all that with gopher; I don't think it will take off." Now, in my defense, just six months later I predicted that in a few years you would see panel trucks with web addresses instead of 800 numbers. The couple of people I told that to looked at my like _I_ was crazy. Damn, I wish I would have put my retirement savings behind that thought.
Is it just my observation, or are there way too many stupid people in the world?
In 1993, I heard about this "world wide web" thing and tried it. I think it was "www", some really awful command line tool. gopher worked far better at the time and was much easier to use. Then Mosaic came out and changed the world. I think you could have done a multimedia gopher along the same lines though; it maybe wouldn't have been quite as flexible, but as far as bandwidth consumption goes, media is media...
In the pre WWW era, Gopher ruled because there wasn't a better alternative. The big complaint was the lack of layout control and flexibility for expansion. Lynx came about as a fusion of the Gopher network protocol and a hypertext interface. Eventually Lynx adopted HTTP and HTML as additional methods and became an extremely popular Web browser. (More users than any of the individual Mosaic browsers.) There was a strong demand for better layout and flexibility and regardless of what network administrators wanted, these features would have evolved.
My geek imagination is now all atwitter imagining an alternate gopher-driven universe.
That sounds a lot like Minitel.
Space game using normal deck of cards: http://BattleCards.org
in the 'early nineties' 1990-1995, there really was not enough power to do what you're talking about. well maybe on a ridiculous setup, but nothing practical in the production world.
as someone who has installed Novell's multi-protocol-router back in the day on something equivalent to a P90, it wasn't terribly hard to overtax the thing. packet scheduling and shaping weren't anywhere near possible. now that was basically a 'software router'
dedicated routers do their jobs better than installing software on a general purpose machine, however in that era the router hardware lagged behind what you could get in a desktop or server. By quite a bit. I don't know about now, I haven't done that sort of stuff in forever, I'm a code monkey now.
People have been playing around with prioritization forever, you could give some clients priority over others in Token Ring 16. I don't recall if Token Ring 4mbit had it. But that was not easy, and then you're basically using the entire network as a router between PCs (kinda sorta). So computationally it wasn't taxing, but your slow local area network just got slower. To the best of my knowledge there was no way to prioritize packets, just client machines (although I didn't work with that much anyways so I'm only 87% sure)
The important thing is that I had a token ring tied to my belt, which was the style at the time. Couldn't get ethernet cause of WW1. =P
I founded one of the early online journals before the invention of HTML/HTTP. It's the Chicago Journal of Theoretical Computer Science, providing articles in copy-edited LaTeX source, as well as precompiled PS and PDF.
At first, the journal served papers through anonymous FTP.
Then, I crafted a Gopher structure to make browsing easier.
As soon as HTML/HTTP came along, I created the HTML version of the journal. It was much more maintainable than the Gopher version, because the hyperlinks decoupled the document structure from the file-system tree structure just enough. In a few years, I stopped maintaining the Gopher version, because it required an order of magnitude more work than the HTML, and readers all preferred the HTML anyway.
Adding pictures and stuff is rather trivial for the data architecture, although demanding for the network implementation. With a more maintainable structure, Gopher would have added the extras. It was the Hyperlinks that made HTML work better.
HTML also has some serious maintenance problems, but they appear later when the archive gets large, and they can be addressed with things like PHP compiling and content management systems.
From another point of view: Gopher essentially made file trees visible over the network (which is what I thought I wanted at first). HTML/HTTP provides a crude network database model distributed over the network.
Future advances in data architecture (as opposed to the types of data within that architecture) will have to do with other database models, and with other sorts of commitments between distributed servers, and with looser coupling between data ownership and server ownership. E.g., a way to provide reasonable assurance of future access to a particular data item (access includes being able to find it, not just its existence), without depending on a particular server at a particular registered domain name (the Wayback machine ameliorates the problem, but doesn't solve it).
Mike O'Donnell http://people.cs.uchicago.edu/~odonnell/
There were hypertext Gopher clients out right around the time mosaic was in beta.
Netmanage had a nice one, I still remember using it to find super models on an Italian university server.
point and click for everything...
As I look at another frenetic ad, not much different from Punch the Monkey of yore, another empty box with a puzzle piece in the middle of it, and 25 connections to ad servers I've never heard of as a result of a single connection to the web site I actually want to go to, a text-only gopher web sounds pretty good.
Archie and gopher were so much better than trolling through ftp archive sites just from word of email. I want my archie back!!!! Oh wait, now google automatically does a waaay better job of doing what archie provided. And those hyperlink thingies make the bulletin board reference to another ftp site automatically download the file for me. Gosh, it's amazing how that automatic sort of thing have ever displaced the amazingly cumbersome, high maintenance archie/gopher/ftp side of things.
-John Van Voorhis
It's surprising that no one here on slashdot has pointed out that a major difference between the html and gopher was that gopher services had to get a licence from the University of Minnesota while http servers could be constructed without a licence.
Free open software with free open standards is what got the web going.
No, not little, brown, furry rodents...
Massive, golden, beautiful beasts!!
As most of you know (or maybe don't know), it's called gopher because it was developed at the University of Minnesota... http://en.wikipedia.org/wiki/Gopher_(protocol)
...or at least a super-Gopher.
I think I saw that guy in Caddyshack...
"Not an actor, but he plays one on TV."
...and have access to a pre-OS X MacOS, TurboGopher VR is a must-see. The screenshot at the bottom of this page (http://www.tidbits.com/iskm/iskm3html/pt4/ch24/ch24.html) doesn't do it justice.
Suffice to say, it is probably the only Gopher client that will ever have a key mapped to a "jump" action that is interpreted literally.
"...always new atoms but always doing the same dance, remembering what the dance was yesterday." -Richard Feynman
The difference between HTTP and Gopher has NOTHING WHATSOEVER to do with ability to serve multimedia content, nor with bandwidth. HTTP, or really HTML, just allows more diverse linking patterns than Gopher's hierarchical format. But there's nothing non-graphical or content specific about gopher. I RAN graphical Gopher clients perfectly happily (well, including early Web browsers that supported that protocol.
Buy Text Processing in Python
"My geek imagination is now all atwitter imagining an alternate gopher-driven universe."
It looks a lot like this one except everyone has a pointy goatee, including the gopher-driven internet.
Insert witty sig here.
Let's respond with that old open source chestnut. If you don't like how it works, fix it yourself!
Or he could use Opera.
It's really very simple, but it's a small technical detail that seems minor at first sight.
When you connect to a Gopher server it waits for you to send a string (just like HTML) this string is the name of the item you want. HTML is almost the same, except you put a command (GET, POST...) in front of it. This is not important because you can easily run an HTTP site with just GETs.
The difference is that Gopher immediately gives you the file, whereas http gives you a status code and file type then the file.
Why is this important? Well it's not, if the world is perfect, but in the real world things change so the "selector" that was valid yesterday and pointed to an image might not be valid today. But the server doesn't know it used to point to an image so it goes and returns the main site page which isn't an image. At this point the client throws a tantrum.
This means that the only links you can reasonably do are to your own site or to the root of someone else's, no deeplinking, no "intergophers". OTOH, because http returns the status and type there are lots of options if things move, but none of them are a client program crash (normally!). This also applies to links you make, ie no bookmarks, so you don't remember where that important document was on the gopher server, but you a (safe) bookmark directly to the http location.
HTTP wins, html comes along for the ride, Ooops!.
The author, essentially, details how if service providers were to choose the technologies we run then we wouldn't have had the uber successful WWW.
To me, this is a strong argument for net neutrality. If the ISPs are allowed to prioritise types of traffic then we may not benefit from the next generation of Internet service. Simply due to the short-sightedness of ISPs. Not through malice, but through market forces. A service would have to become popular for ISPs to encourage or throttle it, however it would be difficult for the service to become popular in the first place.
But is it haiku-compliant?
http://xkcd.com/554/
Just in case.
Gopher was doomed indirectly by Moore's Law, but what killed Gopher was LICENSING.
While processor speed was doubling, MODEM speed was not increasing as rapidly. In fact, acoustic modem speed in commercial devices has since peaked and is not perceptibly increasing.
Gopher existed during a brief period when the modem speeds were climbing from 2400 baud to 19,200. At these speeds, even small image transfers (then GIF and the earliest JPGs) took several seconds. The average Gopher page, at about 1,500 bytes, could load "instantly" at 2400 baud. The average web page, at 15,000 bytes, took several seconds at 2400 baud.
Even at 19,200, large (150K) images took several seconds to load. This delay, while tolerable, discouraged web pages that auto-loaded more than one or two images.
Only when modem speeds increased to 56K did a page loaded with a modest amount of graphics download at an acceptable pace.
Gopher was designed as a campus-wide information system, and only became an Internet phenomenon when University of Minnesota politics drove its development onto the Internet. After three years of design, the U of MN CWIS committee had a foot-thick set of requirements and specifications, with out one page of code. Exasperated with this process, Mark McCahill, Paul Lindner, and Farhad Anklesaria produced a working CWIS prototype, Gopher, between the April and May meetings of the CWIS committee, and demonstrated it at the next meeting. The committee, predictably, had a conniption and further development of Gopher was prohibited. Mark put Gopher up on an FTP site and colleagues at other institutions were invited to continue development, and it took off from there. So Gopher was never intended as a world-wide Internet protocol.
While Gopher was effective at presenting the Internet as a file structure, the Web was always going to be more popular since it presents the Internet like a magazine. The limiting factor was the bandwidth necessary to communicate the images (and other media).
But what killed Gopher was licensing. Just as Gopher was never designed as anything more than a campus-wide information system to help students and faculty connect with campus resources or register for class, so was Gopher programming never intended to become the full-time occupation of the developers. All of us had numerous other responsibilities, not the least of which was walk-in and telephone computer consulting for the 100,000-person U of Mn campuses. While the U of Mn was quite happy to accept the praise and publicity that came from Internet Gopher, no funds were ever actually directed to the project.
In an effort to make it possible for Gopher to be supported full time, a radical new idea was floated. What if we LICENSED the SOFTWARE? For... MONEY?
If this seems conventional to you, now, it certainly was not, then. Back in 1993, most Internet domain names ended in .edu or .mil. A TLD of ".com" was considered crass and improper, I kid you not.
The ruling philosophy was "The Internet is a public medium build by colleges and the military for the use of the public, and using it for profit is crass commercialism." I know, a lot has changed, eh?
So when the Gopher team suggested that institutions PAY for the privilege of using Gopher, Gopher servers vanished overnight. Just the suggestion that money be involved offended some system administrators sensibilites enough to make them drop it; others feared that lawsuits were imminent.
Meanwhile, at just that time, modem speeds increased to 56K, and a brand new FREE, unlicensed server called a WEB server was arriving on the scene. And of course it could show pretty pictures etc. etc. Without funds to allocate the Gopher team's resources full-time, our other duties took us away from Gopher support just as the Web was taking off. The spotlight moved on from Gopher, for multiple reasons, but it moved because of money.
The one place where Gopher SHOULD have been revived, but was not, was on cell phone
Bob Alberti, CISSP, Internet Gopher, RFC 1436. Writer of early MUD, Scepter of Goth, 1983. http://tinyurl.com/Mitlanyal
It helps to be funded to give it away. "Web browsers" came from CERN and the National Center for Atmospheric Research, both tax-funded organizations. The competition wasn't subsidized. There was Xanadu from Ted Nelson and Autodesk; that suffered from Nelson's fascination with micropayments. There were proprietary graphical clients, like AOL, but that had to be a closed world to make money. There were some big commercial services, like Nexis, Lexis, Mead Data Central, and Western Union InfoMaster, all closed pay systems.
At least five other "hypertext" systems preceded the Web. Intermedia, in the 1980s, was one of the better ones, but it was tied to Apple's version of UNIX (A/UX), which was a turkey later discontinued by Apple.
The early systems needed expensive hardware; that was the main problem. Somebody would have done this once the hardware got cheap enough.
How about the guy trying to get the gopher singing "The Balad of the Green Beret"?
I agree with you, and want to point out a slightly OT analogous (in an opposing sense) and relative ubiquitous meme, which is: "we've killed the mad scientist/erased the research documents/etc. and now THE WORLD IS SAFE, because no one will ever be able to recreate the whatever-bad-technology-was-threatening-the-world".
That meme has always rubbed me wrong, considering that just knowing you can do something is a big part of scientific discovery.
There's only one problem with Topolski's argument: it's completely bogus. In fact, it is revisionist history. Network administrators, at the time, were cheering the release of something more powerful and flexible than Gopher (which UMN had just decided it was going to try to license for money). Here's the truth behind Topolski's nonsense. The reason Topolski is making this tenuous, bogus argument is that he has just been hired by a Washington, DC lobbying group called the New America Foundation. This group is what's known as an "astroturf group." It pretends to be populist, but in fact is funded by big corporate money and promotes agendas that those corporations tell it to promote. In the case of the "New America Foundation," this is quite blatant: the Chairman of the group is Eric Schmidt, the CEO of GoogleClick (Google, which has merged with DoubleClick and is therefore the world's largest invader of Web users' privacy). Schmidt he has funneled more than $1 million of Google's money to the group. The group, in turn, parrots Google's corporate agenda to the letter. As does Topolski. Both Google and Topolski are seeking to regulate the Internet in ways that benefit Google at others' expense. In particular, the legislation which Google favors would force ISPs to raise prices, harm or even destroy competitive Internet service providers (leaving a cable/telco duopoly), and harm all Internet users' quality of service. In short, this is a corporate scam. Don't fall for it.
How does he think today's technology can block images sounds and videos? Sure such binary traffic carries some evil bits, today. But it doesn't have to.
If you think that's easy, try Mac OS X. Find a program, download the disk image, double click to mount it, and double click to run. If you want you can drag the icon to Applications or wherever to "install". Only operating system extensions and big legacy software like Adobe Illustrator use an installation program.
I just spent the last two weeks trying to get some Linux/Unix software to work via apt-get and tarball modes of distribution. Oh let me tell you the joys of hours spent figuring out why a package doesn't want to build, install, or run.
I don't know what you call the Moore's Law of storage capacity, but that's what really makes modern computing better than the 90's. These Mac packages include whatever libraries they need to run -- no shared libraries to mess with where every application wants a different version.
A pioneer gets an idea out a little sooner than somebody else would have. Or, in fact, the famous pioneer often isn't the first to develop the idea but is the one with the timing or connections to make it popular. As the Adam or Eve of the idea, the pioneer gets to define the flavor that will propagate along with the idea.
For technology, HTML came along with its own set of quirks. If it hadn't then something else like Super Gopher would fill the technological niche with its quirks. And those quirks could have a big effect on the later evolution of technology.
If the population of soft fishes grows and a shark is not around as the big predator then some form of turtle or otter will fill the niche. If Tesla and Einstein don't become famous then Edison and Bohr dominate.
"Those were the happy days, I hope they never come again."
Tired of all the isms, don't exploit people as an employer, or a government, mmmmK?
in 1981 I joined a company that had "thicknet" running down the halls of the "experimental" zone in the IT building. We had at least 2 XEROX "Portrait Display" GUI workstations and a file server with an attached laser printer (all of which appeared to be used mainly to create decks of nicely formatted presentation foils; main value of this system IMHO).
I didn't see a Lisa on that site for at least a year, and my understanding of what happened then - at least as a customer - was the Lisa was a $10K version of what Xerox wanted us to pay $18K for per seat. Ironically, that company (and most others at that time in the Fortune 50) went instead with IBM PC's and XT's (which at up to $6K per seat fully configured were cheaper than either option, if obviously inferior in capability at that point in time).
I write this because one aspect of this thread has been to discuss multiple sources of things that came later. Are individuals the key, or are some outcomes inevitable? Raskin is claimed in the preceding entry to be the single source of both the Apple and PARC GUI evolution (.i.e., the "key individual"); but clearly both were interconnected, and PARC - at least to me - had a quasi-commercial product on the market before Apple. Just as the Macintosh was evolved from the Lisa, I would say the "big picture" view - networked Macintoshes connected to a shared postscript printer with file services - was being offered to the market by PARC prior to Apple's commercial delivery of it - my direct experience.
Perhaps Apple could have come up with all the same concepts w/o PARC getting there first from a production standpoint . But to me, BOTH were needed - XEROX with the corporate PoC, and Apple then making it a commercial reality (Lisa to Mac to the Mac family, etc.). This is the point here about the web as well. If Gopher had evolved to look more like the web, it wouldn't have been the Gopher from 1990, but something else derived from Gopher. To me classic Gopher "lost" because Mosaic (at least beta .7 and forward) was better - I and other tech-heads could use Gopher, but anyone (my wife and my children in my own case) could use Mosaic. I switched, along with everyone else; but Gopher had its in this story, as PARC did in the GUI PC story.
Apple eventually beat Xerox in this space in my opinion because they commercialized an advanced workstation concept at a mass-market level (i.e., taking the view of a "PC" vendor versus a "scientific workstation" or "office automation" vendor). I think we can all agree with that as a fact of history; but the foundation for the GUI-based PC and all the things that made it work link back decades. In short, this was not just based on the work and insights of one or two key people. Those people that were key to Apple's success - Jobs, Rankin, the Macintosh Development team, etc. - deserve the credit they have been given. Just don't tell me they weren't also standing on the shoulders of others.
The licensing issue is what killed the possibility of using Gopher for an externally visible service where I was working at the time.
The real problem was that it would have meant asking the boss to do something; licenses require the involvement of a suitable corporate officer actually making a decision. Just putting up a web server on the machine that was hosting our FTP server to serve the content already approved for release in a nicer way took a little technical work and the absence of a decision from above.
The license may have been for the Gopher server code but there was no way we could have justified spending the time to do our own implementation.
We experimented with Gopher internally very briefly but the organisational obstacles killed it very quickly.