back to article Internet overlords deny Google's 'dotless' domains dream

The Internet Architecture Board (IAB) has issued a statement in which it all-but-rules-out Google's plan to take over some new top-level domains and offer them in “dotless” configurations that would enable web addresses like “http://search". Google outlined its plans for .search, .app and cloud back in April. The idea of …

COMMENTS

This topic is closed for new posts.
  1. the spectacularly refined chap

    Hardly news

    I remember raising this very issue on these forums a while back and was assured "Oh no, it'll be fine" by those who know no better. The idea that Google (or anyone else) would be allowed to force administrators all over the world to reconfigure their systems simply to allow a few companies even more prestige over a few monopoly domains was always going to be a non-starter.

    The fact this has taken three months to be formally addressed (looking back through my posting history) just show how ill-considered this nonsensical scheme has been from the start.

    1. Ian Yates

      Re: Hardly news

      I think the news it that the IETF are completely incompetent and that at least part of it can foresee the danger that the new TLD land-grab poses to those who aren't paying huge sums of money for it (i.e., us).

      I've nothing against companies suggesting new TLDs, but they should continue to be managed by decentralised, not-for-profit* organisations; and they need to be heavily vetted.

  2. Anonymous Coward
    Anonymous Coward

    @TSRC - I agree that it would not be fine however I don't share your view that it was destined to fail.

    The MPAA & RIAA have given us salutory lessons in how well positioned companies are able to force their agenda on the world at large.

  3. jake Silver badge

    Nobody with a clue ...

    ... ever claimed the gootards had an actual understanding of the underpinnings of TehIntraWebTubes. Totally clueless bunch of upstart kids, the lot of 'em ...

  4. Anonymous Coward
    Anonymous Coward

    Oh please... this is a 10+ year old rehash of AOL era technology

    Those of us old enough and has been in the internet industry for long enough knows that's just a rehash of AOL keywords, I'm sure to the old geezers on IETF and IAB, this was clear on day one, no further discussions is required..

    1. jake Silver badge

      Re: Oh please... this is a 10+ year old rehash of AOL era technology

      Try 20+ years. But AOL "keywords" were (are?) internal to AOL, and didn't reach out to TehIntraWebTubes at large. Kinda like typing "starfish" into Wikipedia's search box.

      1. Anonymous Coward
        Anonymous Coward

        Re: Oh please... this is a 10+ year old rehash of AOL era technology

        I don't think it's 20+ years, pretty sure AOL didn't have keywords before '93.

        In any case, the fundamental situation hasn't changed, big companies wants to own single words and thereby relegating everybody else to 2nd/3rd tier netizen, hence "rehash".

        1. jake Silver badge

          Re: Oh please... this is a 10+ year old rehash of AOL era technology

          AOL's Keywords came out in early 1991, with the release of the "Dark Side" DOS release. Not everybody knew what <ctrl>K was, but it worked.

          See also: PCGEOS (or GEOSWorks).

      2. Anonymous Coward
        Anonymous Coward

        Re: Oh please... this is a 10+ year old rehash of AOL era technology

        And no, it's not like typing "starfish" into wikipedia's search box, it's not just a search matching function because AOL integrated into their browser that people used and is advertised as an alternative to typing in an http address.

        1. jake Silver badge

          Re: Oh please... this is a 10+ year old rehash of AOL era technology

          AOL didn't HAVE a browser, or even Internet access, back then. Neither did (almost) anybody else.

          See: Eternal September.

          jake, Monday, September 7258, 1993

    2. Daniel B.
      Boffin

      Re: Oh please... this is a 10+ year old rehash of AOL era technology

      Not quite that. In fact as the IETF and IAB dudes have pointed out, it won't work as advertised. DNS resolution can indeed search for TLDs ... but the first thing it'll do when presented with a dotless query is to search for $DOTLESS_FQDN.my.default.search.domain.com and return that if it finds it. If you really want to search for a TLD on DNS, the correct way to do this is by appending a dot on the search. So for http://search to work, you'd actually need to type http://search./ for it to work unambiguously!

      AOL had its closed wall garden to implement keywords, the overall internet is not the same.

  5. Ole Juul

    and conversely

    dotless domains introduce potential security vulnerabilities

    I'd be glad to use two dots if it would reduce vulnerabilities.

    1. MachDiamond Silver badge
      Pirate

      Re: and conversely

      I'm often clued in on a phishing expedition when I see addresses formatted "http://mybank.phishingsite.com. I can see very quickly from the syntax that I'm not going to taken to "mybank". The other glaring clue is that I will receive between 3 and 8 "notifications of security information" in a row lined up in my InBox. I'm not sure if two dots will help, but I am worried that with no dots I might be less likely to see the discrepancy.

  6. Anonymous Coward
    Anonymous Coward

    Expectations

    "most users entering single-label names want them to be resolved in a local context"

    So it must be incredibly annoying for most users then that when you type in a single-label name, most browsers assume you wish to do a web search rather than going to the local address. I'm regularly typing something, then having to retype it with the http:// prefix after the browser takes me somewhere different.

    1. Anonymous Coward
      Anonymous Coward

      Re: Expectations

      Seems to be a Chrome thing, I've never had this problem with Firefox.

      Then again, I guess my fingers are preprogrammed to always put the http:// prefix in, since I'm used to doing it from back in the early days of the commercial Internet (~1996; okay I realise some had it a few years before that, but it was pretty new for most in Brisbane then) when you had to do it.

      1. robmobz
        Devil

        Re: Expectations

        "Seems to be a Chrome thing, I've never had this problem with Firefox."

        Chrome or internet explorer (not sire about newer ones but certainly ie6)

    2. DaLo

      Re: Expectations

      Just add a trailing forward slash (/) after the name, easier than typing http://

      So type myserver/

    3. Michael Wojcik Silver badge

      Re: Expectations

      when you type in a single-label name, most browsers assume you wish to do a web search rather than going to the local address

      With Firefox, it's trivial to disable this misfeature in about:config: set keyword.enabled to false. I also set browser.fixup.alternate.enabled to false - that prevents Firefox from trying to add "www." and various gTLDs to hostnames in the address bar.

  7. Anonymous Coward
    Anonymous Coward

    Sod the practicality, rush this through, get it announced.

    People who similarly lack understanding of the technical side of things will fix the economy when http://porn and http://sex etc come up for sale.

  8. Number6

    Short is everything

    Back in the DOS (and even CP/M) days, many popular applications had two-letter names in the filesystem, who remembers wp and ws? I guess Google are trying to shorten what you need to get to them, despite many browsers now offering a preconfigured search box to talk to the search engine of your choice.

    I hope the IAB continues to reject the dotless domain, such things should be considered in the same category as non-routeable IP addresses (192.168.x.y etc), something that people can use in the privacy of their own LAN but which should not be seen in public.

  9. AustinTX

    It's stupid having to pay someone for a dot and some letters

    The stinkin' ICANN doesn't have a divine right to force us to rent a dot and three/four letters to tack onto the end of our web addresses! Anyone can create a free subdomain. Anyone can opt-in to use alternate DNS. Just combine the two, and let people register their own addresses, with or without dots, that work for anyone else using that DNS system. I can do it with my HOSTS file, or my own DNS server. Corporations create short web names for their LANs all the time. A simple browser plugin would suffice for http-only resolutions. If you use the alternate DNS, you could go to http://el.reg.is-cool_man, else it would also work as http://el.reg.is-cool_man.altdns.com for those still locked in ICANN's walled garden.

    Companies like FreeDNS and Google are perfectly poised to add this extended DNS ability, and could achieve critical mass with a single browser update. Though you could reject registration attempts for duplicates of existing ICANN DNS names, I would even suggest it's legal to duplicate them anyway, within the confines of the alternate, opt-in network. FreeDNS's spam and porn-guard service is proof that it's legal to redirect legitimate URLs when using an opt-in alternative. So much for domain speculators and their eternally parked addresses! A nominal fee, and the release of domains unused for x amount of time, would eliminate a good deal of abuse. Not saying there aren't problems with the idea, but every one I've thought of seemed to have a good, solid solution. Instead of renting theregister.co.uk for $30/year, how about http://theregister for $3/forever?

    1. craigj

      Re: It's stupid having to pay someone for a dot and some letters

      Alternative DNS isn't new, its been done before, a company called new.net sold their own "alternative domains" like .xxx - It didn't work out too well for them when Icann launched the same extension.

      I think the last thing you would want is competing DNS systems, never knowing if all of your users will reach your intended website because they may be looking up to a different DNS server. Seems like a Scammers dream!

      1. AustinTX
        Thumb Up

        Re: It's stupid having to pay someone for a dot and some letters

        I didn't know about new.net. Very cool. And I agree with you to a point about the concern for scam websites. So, block registrations of domains that duplicate ICANN DNS. And an alternate DNS network would not place you in a walled garden if it federated/relayed to ICANN DNS upon finding no registrations. Competing DNS systems don't need to cause confusion as long as they simply duplicate ICANN while adding additional services.

        Signing up for a new domain name should be as free and simple as signing up for a new email address! Sure, people registered lots of "look alike" emails for nefarious purposes, but we've coped just fine. Assuming an alternate DNS network achieves critical mass and takes over enough of ICANN's turf, the worst that could happen is ICANN gives up on dot-tld and likewise allows anyone to effectively create their own "TLD" for far less than $185,000.00!! I think either Google or Mozilla could manage this almost instantly. FreeDNS or OpenDNS could provide the backbone for it, but of course Google could use their own large network of DNS servers.

        It seems like much of the problems we have with the Internet today (high price, poor customer service, walled gardening), actually are about a lack of competition. My main issue with ICANN being that they're obviously a monopoly for-profit corporation posing as a caring non-profit. They have too much power and offer too little for the "services" they impose. They're toll-guards, not facilitators now. I think they're doing enough harm now, that feeling some heat would be good for them. The idea of competing/federated DNS networks scares me less than what I think ICANN might do to maintain their grip on our cohones.

  10. IGnatius T Foobar
    FAIL

    And furthermore...

    And furthermore ... if IETF/IAB really want to stop this sort of nonsense, they should immediately enact a global ban on Windows 8.1, which illegally mixes local-computer search with Bing search.

    1. Anonymous Coward
      Anonymous Coward

      Re: And furthermore...

      upvote despite a questionable use of "illegally" - "irresponsibly", perhaps, but is it actually illegal? Given most politicians speaking about Internet issues seem unable to actually spell URI, not convinced there is actually statute on the issue ...

    2. Anonymous Coward
      Thumb Down

      Re: And furthermore...

      Ban? Excuse me, but federating local, network, and internet/web search has been going on for at least fifteen years looking at my database search & business intelligence engineering folders. That's just on Microsoft platforms, which also has had the option to federate with Bing for at least the last two years. Federated search has been a feature of MS Help documentation for some time as you'd know if you ever used help in the last ten years, or used Visual Studio documentation, for instance. Furthermore, federated search has been available for at least a decade on Linux. (Probably longer but I don't have the older files to back that up.)

      Moving on to the enterprise, their systems have long used connectors their enterprise buses or content management systems to blend content. If it's soooo. illegal, or improper, why has it been in use for so long? Methinks thou protest to much.

This topic is closed for new posts.

Other stories you might like