Failed Internet Technologies Examples


I was talking to some co-workers recently about what works and what doesn’t as far as platform technologies on the Internet.  Why did something like WAIS and FTP die out but http and RSS succeed?    The answer is not 100% clear to me… So I am coming to you!

I am gather a list of failed Internet technologies and why.  I want to include both MS and non-MS technologies.  Please comment on this post or send me an email. 

thanks for your help!

Update: As many early reader’s commented, I think FTP is a really bad example of what I mean here. Not only is it still in use today, but it had a heyday, when it was mainstream. 
What I by “failed” is ones that never really took off… So think of IE’s Channel’s concept as a good example.


Comments (25)

  1. Alan Dean says:

    I’m not entirely sure that you can call FTP a failure. It is certainly not as widely used as other protocols, but I reckon that most organisations of any size have some usage of FTP somewhere (often business-critical) and many people download files via FTP regularly.

  2. Ryan G says:

    If you’re going to mention WAIS, you should mention the Gopher protocol as well.  I imagine both WAIS and gopher died out because they didn’t have the same richness of WWW.  Granted they could have moved in that direction, but http was sort of built with that in mind already.

  3. Einar G says:

    Agree with Alan. FTP hardly is failure. Even tough it is not that visible to normal user it is used in organisations and then ofthen critcal

  4. anotherlab says:

    I don’t think FTP has died out, it’s just not used to the same extent as HTTP.  It’s a more specialized tool aimed for the power user.

    WAIS faded away because it had out lived it’s usefulness.  It was a text based index database search tool that didn’t really link to anything.  Mosiac and Yahoo pretty much obsoleted WAIS.  Instead of getting terse character text from a Gopher server, you could search across the web and get results for your query in all sorts of formats.

  5. Shawn Oster says:

    A few things jump out at me:

    1. Easy setup on the backend.  There should either be tools or services make it dead simple to setup the technology.  Just like it’s pretty easy to get a RSS feed going just by starting a blog.

    2. Easy consumption.  The tools for the average user to consumer and use the tech need to come along so it becomes easy to play with and discover.

    3. Simplicity in at least one part of the equation.  With RSS it’s pretty easy for an average HTML guy to whip up a feed, even though a quality feed reader takes a bit more.  Not so easy to create a FTP server or client.  People need to be able to tinker with it.

    4. Easy way for non-techs to talk about it.  RSS isn’t actually what’s popular, it’s blogs.  A technology is only as rockstar as what it’s used for.  No one in the consumer segment says "did you subscribe to my RSS feed?" they say, "hey, you subscribed to my blog yet?"

    5. Small, loose spec.  The more loose you keep it the more people will adopt it because it seems less formal and more grassroots.  Look at RSS today, there are still a lot of badly formed feeds but it was pretty easy to get started because you didn’t have to wade through a massive document.  People like to feel they’re getting in on the ground floor with something, helping to pioneer it.  If there is a 50 page spec, 30 examples and talk of the 8 projects in development then people lack the trail-blazer spark and passion.

    I think of NNTP, that didn’t have the same level as RSS because even if a company wanted to have newsgroups there was no super simple way to start one.  I still use newsgroups but it never grew outside of the tech world.

    IRC is/was great but finding a server, connecting, figuring out the right port, THEN discovering the rules of that particular server was really beyond your average consumer.  Clients like mIRC always had UI’s that always seemed only a step above the code.  Plus, how do you set up a IRC server?  IRC could have been awesome to get people connected with tech support but it violated rules #1 AND #2.  I bet it would have been a lot more popular if you could just drop something on your site and have instant IRC chat, with a client that didn’t bother you with ports and what not.

  6. jvierra says:

    Back in the old days we used FTP because it was public domain and Columbia Universtiy was good at providing sorce for all platforms.

    I remember my rtip to upper Manhatten to pick up a hand made tape that included all of teh source for all cuurent platforms.

    This was magic.  A platform-to-platform congruence in an IBM world.

    Today, due to numerous advances in technology and protocols FTP is a last choice for any serious file transfer applications.  Still…FTP survives and has many guises.  There are many tools that implement secure and reliable file transport that use FTP at teh core.

    I don’t know of any serious efforts that have failed.  MOst have just been consumed by newer technologies.

    After all..aren’t web services just a modern implementation of RPC?

    Not technology is ever a wasted effort – it is only superseeded by a newer and, perhaps, better technology.

    Has DOS died out?  I don’t think so.. There are many using teh NT command shell which is basically DOS with some extensions.

    The only Internet technology that I know of that has completely failed is Oracle’s attempt at a thin client run solely via an Oracle database server with no OS.

    Oops -soory.  I guess that wasn’t really a technology but just a bit of bragging.

  7. X.Static says:

    Java applets spring to mind (the concept itself is still alive and kicking, but embedded Java apps are all but dead).

    VB6 Internet Apps (I don’t remember what they were called, but they were dreadful).

    The last one that springs to mind isn’t an internet technology per se, but it is related: Visual Interdev was one of the worst IDE’s ever released by Microsoft.

  8. Jens Gyldenkærne Clausen says:

    I disagree with Shawn Oster’s issue #4 – "RSS isn’t actually what’s popular, it’s blogs"

    News sites has in my opinion a major part in the popularity of rss today. I used to subscribe to different newsletters to get daily news at work – these mails are now long gone in favour of feeds. Most of the news sites do have blogs now, but their adaptation of blogs came much later than the feeds.

    The popularity of RSS is in my opinion only partly related to blogs. RSS gives you the ability to choose among lots of stories just by looking at the headline – in a compact and standardized format. Blogs and news sites are perfect hosts for a feed, but it is the properties of the feed that makes rss a killer application.

  9. Sergio Pereira says:

    I know I’m dodging the original questions but there does not seem to be a clear logic on what takes off and what tanks. Sure, after something succeeds or fails it’s always easy trying to explain the reason, but no one draws those bright conclusions before the launch.

    I for one am happy that there aren’t any recipes for success. It forces creative people to explore more and more, eventually finding solutions that are mind-boggling and highly disruptive. I think that makes our lives that more interesting.

  10. Page Brooks says:

    I don’t know if this is justified, but VRML comes to mind when I think of failed Internet technologies.  Of course, this was 10 years ago. 🙂

  11. Phil says:

    CORBA – Common Object Request Broker Architecture.  This web technology was quickly overshadowed by web services technology.  It had a fatal flaw in that as a messaging structure it’s ‘common’ hub meant that there was a single point of failure.  Not many enterprise applications still use this technology.

  12. Alfred Myers says:

    The main factor, in my opinion, is the content and tooling for both generating and consuming the services provided by the protocol.

    Both web pages (HTML) and RSS (both via HTTP) are very rich on the content offered by their means and have endless tools for both generating and consuming their content.

    Although IE’s Channels failed, its ideas are to some degree very much like the concept of RSS. Probably the main reason for it failing was that was backed up almost by no one but Microsoft.

    So to be successful, a technology must be useful, easy to use, have good tooling and be backed up by the community or industry as a whole.

  13. Henock Zewdie says:

    VB6’s WebClass is one technology that comes to mind immediately.

  14. Dave S. says:

    ColdFusion.  Probably because of .NET’s success.  Some still use it but probably because of dependency.

  15. Jon Kruger says:

    If FTP has died out, what is the replacement?  I’d love it if there were something better!

  16. John Coleman says:

    FTP died?  I don’t think so… it’s very mainstream.  It’s not not something home users user every day.  Nothing has ever come close to replacing it at all.  It’s the backbone of our bulk data transfer.

  17. Kalpesh says:

    Its more of acceptance by users (not only people in IT) that matters.

    People who are into development might know RSS. What about others?

    Also, some tools really simplify the job to consume some technology by end users. e.g. one doesnt need to get the feed from some page & add it to their aggregator. Browsers like IE & FF do it easily for them to discover & add 🙂

    FTP hasnt died out but it isnt visible to other users, even when they are using it. HTTP does the job of FTP  for users easily, but cant replace it as such.

  18. J.Marsch says:

    Anyone remember GOPHER?  I remember piddling around with it back in school, but soon after that web sites started budding and everyone was using Netscape and Trumpet Winsock…

  19. Garry says:

    IMAP EMail – No Idea why it never really took off, seemed like a great idea, downloading EMail and leaving a copy on the server for you to read from another client.

    But have to agree about FTP, I use Biztalk to transform 1000’s messages per day, can easily offer up file delivery over HTTP or webservice, just cannot get any of our clients to buy in to the idea when FTP is so realiable.

  20. Wolf Logan says:

    I think the thing that jumps out at me most from comparing the two lists of technologies is that the failed ones are much more "centralised" than the successful ones.

    Gopher failed not so much because it lacked the richness of HTTP (though it was certainly clunkier), but because any content-level changes to a Gopher server required manual filing and structure updating, which was a centralised activity. It was fine for relatively static collections of heirarchically organised information, but the kind of ad-hoc delegation of content management that was so prevalent in the early days of the web was nearly impossible.

    Ditto WAIS, which was a great search technology — as long as you had some kind of centralised control over the content. Offering indexing of content you didn’t really own was non-trivial, so most WAIS servers were over small private data collections, and there wasn’t a good universal federation mechanism in place. So WAIS essentially starved out when systems like the World Wide Web Worm (an early search indexer) and AltaVista began to appear.

    It seems that the recipe for a successful protocol or technology is to remove the bottlenecks, and make it possible to distribute it widely. If you can easily federate it (without adding to the implementation complexity), then that’s a bonus.

  21. Ed says:

    Generating a list technologies that no one ever used is going to be tricky.

    Gopher was a great way of publishing actual information.  WebDav is better than FTP and IMAP is better than using a dodgy HTML mail reader – they’re just not consumer level solutions I guess.

    CORBA / IIOP failed because it was a bit over ambitious / complex.  WMS/WFS (OpenGIS mapping protocols) are a bit stalled because proprietary systems (Google Earth, Virtual Earth etc.) are generally better written.  (WMS/WFS is what NASA World Wind is based on).  

    I’ve vague recolections of using an ‘archie’ search engine which trawled FTP sites at uni.

  22. How about client-side VBScript as opposed to javascript? Or dare I say, ActiveX?

  23. VRML. Imagine SecondLife today if VRML had stuck around. 🙂

  24. MarcoM says:

    IMAP4 is quite widely used, I would not call it a failure.

    CORBA, IIRC (I might be wrong) is used internally in the Gnome desktop architecture, but not used on the internet anymore.

    For me, the single most backed technology with the biggest failure were Java applets.

  25. Goran says:

    Netscape comes to mind when I think of failed/obselete Internet technologies.  I remember the browser wars where Microsoft came out as the  winner.  Who uses Netscape today?  

    Now it’s Firefox and Internet Explorer.