Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror
×
The Internet Internet Explorer IT

Opera to Stop Spoofing User Agent as IE 360

Anonymous reader writes "The Opera browser will stop spoofing its User Agent (UA) as Internet Explorer. Currently Opera, by default, spoofs its UA to identify itself as Internet Explorer. This is seen, by some, as a move that will bring up Opera's usage stats a bit higher, and will hopefully make webmasters, who develop IE centric sites, more aware of Opera."
This discussion has been archived. No new comments can be posted.

Opera to Stop Spoofing User Agent as IE

Comments Filter:
  • Screwed both ways (Score:2, Insightful)

    by intmainvoid ( 109559 )
    They're dammed if they do (users getting blocked from sites they would otherwise be perfectly able to access with Opera) and dammed if they don't (on the usage stats).

    Can't they just stick the word "Opera" somewhere in the user agent string, but still make like they're IE?
    • Not necessarily. Just make it a user option. If the site works with opera, its reported that opera is looking. If it blocks them, the user checks a "pretend to be IE" box, and bam, it works.
      • It currently is an option. In fact, it's an option in quick-prefences, so with 2 mouse-clicks you can change your identity to Opera, Mozilla, or IE.

        They're talking about the "default" option, which is set when you install. After all, Joe Sixpack probably has no idea what that option would do for him.

        But as others have said, they're losing both ways. I've been to sites that won't allow me to access their forms if I'm ID'd as Opera, but ID'd as IE and it's ok.
    • Re:Screwed both ways (Score:2, Informative)

      by karmatic ( 776420 )
      Mozilla/4.0 (compatible; MSIE 5.0; Windows 2000) Opera 5.12 [en]

      That's what they currently do.
    • This doesn't seem to be hurting Firefox or Safari, though; I browse with both, and can't recall when a site wouldn't render properly. As far as sites that employ IE-specific technology, are concerned, these sites won't render properly anyway, unless Opera embeds an actual IE rendering layer, like the new Netscape is trying to do.

      I, personally, wonder what has taken them so long. Just make it a user-changeable preference, and be done with it. If they want to get clever about it, perhaps make it so that Op
      • Try this one with mozilla: http://www.casinocity.nl/ [casinocity.nl] (casino site, you have to be 18+, the insurance company I used to use as an bad example fixed their site)

        Anyway, that kind of stupid test is why the pretending to be IE is there. When they just redirect, you do not know if it is acting correct or not.

        Anyway: Trying to convince companies & developers to develop good compatible sites is always good!
        • And when they do that, the key here is you go to their competition.

          Changing your ID string and supporting these idiot businesses is insane. Folks have been convincing them for years that their strategy was correct.

        • Try this one with mozilla: http://www.casinocity.nl/ [casinocity.nl] (casino site, you have to be 18+, the insurance company I used to use as an bad example fixed their site)

          Here's the error page returned by Firefox: "Your browser is to old to visit this site." Notice the high quality of spelling ;). Also notice the lack of document type definition and paragraph tags in the page source.

          Yeah, I usually hate grammar natzis, but there's a difference between a Slashdot comment and something that's supposed to be commerc

      • This doesn't seem to be hurting Firefox or Safari, though; I browse with both, and can't recall when a site wouldn't render properly.

        Parts of British Airways, Travlocity, and CheapTickets won't render properly in Safari.
    • Re:Screwed both ways (Score:5, Informative)

      by swright ( 202401 ) on Sunday July 31, 2005 @02:35PM (#13208833) Homepage
      gah, they *already do*!

      They've always had Opera and the version in the useragent string - they just have the MSIE bit in there as well.

      this fools the lame IE-only stuff, but lets any sensible software detect that really it is Opera.

      more info here: http://www.opera.com/support/search/supsearch.dml? index=570 [opera.com]
      • "They've always had Opera and the version in the useragent string - they just have the MSIE bit in there as well."

        We just have to hope that Opera doesn't become popular, otherwise you'll have to have "Mozilla/5.0 (compatible, MSIE 7) (Opera) TheActualBrowserName" in a user-agent to get pages served to you...
    • by dougmc ( 70836 ) <dougmc+slashdot@frenzied.us> on Sunday July 31, 2005 @02:37PM (#13208850) Homepage
      and dammed if they don't (on the usage stats).
      Of course, this assumes that it actually matters how many people use Opera, and that they be accurately counted. I suspect that it only matters for bragging rights, but I'm sure that others will say that `if enough people use Opera, we'll support it'. (Except that if they did their site correctly, it would work on any browser already.)

      Opera (the company) has always whined that they weren't being properly counted because of they defaulted to pretending to be IE, so it'll be good to finally remove this whine. (Of course, they can still whine about it, as they'll say it's people using older versions, or people who have changed it manually, so maybe nothing will change.)

      • by VoidWraith ( 797276 ) <void_wraith&hotmail,com> on Sunday July 31, 2005 @02:54PM (#13208953)
        Actually, if they did their site correctly, it would work in everything but IE.
      • Whining? (Score:5, Insightful)

        by hkmwbz ( 531650 ) on Sunday July 31, 2005 @03:33PM (#13209175) Journal
        "Opera (the company) has always whined that they weren't being properly counted because of they defaulted to pretending to be IE, so it'll be good to finally remove this whine. (Of course, they can still whine about it, as they'll say it's people using older versions, or people who have changed it manually, so maybe nothing will change.)"
        I don't get it.

        In what way is pointing out the fact that sites often fail to detect Opera because it spoofs as IE by default whining?

        What do you mean by "whining" anyway?

        Is it whining if your browser is being discriminated against, and you make a point of that? Were the black slaves in the US "whining" when they wanted freedom? Were those who wanted to abolish slavery "whining"? Yeah, I'm purposedly exaggerating slightly, but surely you get my point.

        I don't get the hostility towards Opera. The company pays several people to work with web standards in the W3C. The guy who invented CSS works for the company. Even as tiny as Opera is it has still defined what a modern browser is supposed to do. A lot of the "innovations" in Firefox and IE7 were introduced by Opera. Heck, the company even officially opposes software patents [ffii.org], so it's not even trying to prevent free software from just doing whatever Opera can do (or at least trying). Stuff Mozilla representatives are bragging about in Minimo, such as Small Screen Rendering, spatial navigation, and other things Minimo is supposedly going to revolutionize the mobile browser market with, were invented by Opera, and have been available to users of mobile phones with Opera on them for ages.

        Why the constant derogatory comments about Opera on Slashdot? I mean, the first paragraph you wrote was informative, but then you just had to add that second paragraph to make sure that you showed everyone how you really think Opera is lame, "so please don't mod me down for saying something remotely positive about Opera"?

        • Re:Whining? (Score:3, Insightful)

          by dougmc ( 70836 )

          In what way is pointing out the fact that sites often fail to detect Opera because it spoofs as IE by default whining?

          Opera (the company) and Opera users have often claimed that the numbers of Opera users were being drastically under-reported because of this spoofing.

          And often it took the form of whining ... `This report is so unfair to Opera users ... there's so many more of us than your web server logs show, etc.'

          Is it whining if your browser is being discriminated against, and you make a point

          • Re:Whining? (Score:3, Insightful)

            by masklinn ( 823351 )
            If the Opera developers wanted Opera to be counted properly, it wouldn't spoof it's user agent by default. {snip} IE doesn't
            Wrong, the "Mozilla/4.0" at the beginning of the IE UA is nothing more than spoofing Netscape's old UAS, and adding (compatible, mybrowsername) doesn't make it any better.
            The Mozilla Foundation is the only one supposed to use the "Mozilla/x.y" UAS, anyone else using it is spoofing, case closed.
            • Wrong

              Fair enough. Obviously IE users must be under-represented as well, since IE spoofs it's UA string as well.

              and adding (compatible, mybrowsername) doesn't make it any better.

              Well, everybodys looks for it, so it works out just fine, even if it's not technically right.

              Of course, I'm being sarcastic here. Opera and Opera users may not think they're being counted properly, but according to this page [student.uu.se] most of the website statistics services already count them correctly, spoofed or not.

    • Re:Screwed both ways (Score:5, Interesting)

      by Linus Torvaalds ( 876626 ) on Sunday July 31, 2005 @05:30PM (#13209786)
      It's irrelevant for the usage stats anyway. HTTP logs are a completely unreliable way of measuring browser usage.

      Example: if they change their browser to violate the HTTP protocol when hitting the back button, so that it does the same thing as Internet Explorer does, then they will show up in logs a lot more. Now how does that equate to higher usage? It doesn't. But the stupid people who think you can measure browser usage by looking at logs will think that a load of people have suddenly switched to Opera.

      Observing HTTP traffic is so unreliable, you might as well make up market share statistics. Ignore people who think they can tell you how popular a browser is without conducting a proper survey.
  • A better idea... (Score:5, Insightful)

    by mendaliv ( 898932 ) on Sunday July 31, 2005 @02:32PM (#13208805)
    I would think a better way to combat the "sites that target opera users" problem would be to have a big button next to "refresh" that says "if the page looks weird click here!"

    In that case, the page would refresh and the browser would lie to the webserver about what browser it is for the remainder of that session on that domain.

    Best of luck to Opera though. Hopefully there aren't so many sites that will screw the browser over.
    • That is such a great but simple idea. Why don't you email Opera about it! I'm being sincere here as well. "if the page looks weird click here" is brilliant and such a simple question for the joe sixpacks out there. You should email them.
      • Opera 8 has a "Report this site" (or something like that) function in the Help menu. From what I've heard (I've never tried it), it essentially makes the browser ID as regular MSIE 6.0 - not the "MSIE 6.0 but it's really Opera 8.02" that it defaults to. I DO know that it sends the site to Opera for inclusion in a list of sites to auto-apply the MSIE UA to. IIRC, this behavior takes place even if Opera's running in "Identify as Opera" mode.
      • But it doesn't seem to help when I use it on it.slashdot.org... the colors are still weird looking, no matter how many times I click that button!
    • Re:A better idea... (Score:3, Interesting)

      by Anonymous Coward
      What would be nice is if the search engines (google, yahoo, etc) accessed the webpages once as IE, once as Opera, and once as "Search engine", and if the static contents differ too much between the three results, ignore that page totally. Perhaps non-trivial but it'll go a long way into making the web less browser-dependent.
    • Re:A better idea... (Score:5, Interesting)

      by QuietLagoon ( 813062 ) on Sunday July 31, 2005 @04:23PM (#13209477)
      Hopefully there aren't so many sites that will screw the browser over.

      www.cvs.com is one of the more major sites that block Opera. You receive an error page stating, "At this time, our site does not support the Opera browser. We hope to remedy this in the near future.".

      If you write to the webmaster about it, you receive a canned reply that says they are planning to have Opera support very soon. Unfortunately, cvs.com has been giving that same canned reply for about four years.

    • Well, it doesn't do exactly that, but there is a "Report a site problem" under help that sends some details about borked sites to Opera.
  • Not likely (Score:5, Insightful)

    by NutscrapeSucks ( 446616 ) on Sunday July 31, 2005 @02:33PM (#13208807)
    Our stats package can supposedly detect Opera's spoofed UA, and I'm still seeing numbers like 0.2%.

    Despite my username, right now IE5/Macintosh is the bane of my existance as it is still over the magic 1.0% line.
    • "Our stats package can supposedly detect Opera's spoofed UA"
      Supposedly?
      "and I'm still seeing numbers like 0.2%."
      Doesn't that tell you that the "supposedly" above might be wrong? Most people agree that (even with today's flawed browser stats), Opera has at least closer to one per cent globally.
      • Re:Not likely (Score:3, Informative)

        Supposedly as in they claim to detect it, but I haven't actually verified it myself.

        Doesn't that tell you that the "supposedly" above might be wrong? Most people agree that (even with today's flawed browser stats), Opera has at least closer to one per cent globally.

        Actually it reinforces my opinion that "flawed stats" are an excuse that allows Opera Fans over-estimate their marketshare by dismissing any emprical evidence that runs counter to their assumptions (just what you did). Quite frankly, Opera's issu
      • it's closer to 0.04% on the servers i administer.

        we're actually seeing about 5 times as many Lynx users as Opera users, and we're definitely counting the "spoofed" user-agent with "Opera" at the end as Opera, not MSIE.

    • IE5/Mac is actually a lot better than the windows version, think of it more like an older version of Mozilla.. It has CSS support which is far in advance of IE6 for windows, but still somewhat behind modern versions of Mozilla or Opera..
      • Not at all in my experience. Our CSS works fine with IE6/Win, Firefox, and Safari, but on IE5/Mac it's often screwed. IE6 positioning support is pretty complete (as long as you use the correct DOCTYPE), it just has some annoying bugs. IE5/Mac appears to be both incomplete and even more buggy.
      • Once Apple released safari, microsoft stopped supporting IE on Mac. If they can't dominate, they take their toys and go home.

        Many of my clients though associate WWW with IE so they still continue to use it. What Apple needs to do is recycle their "Browse the internet" app and stick that in the dock.
  • Er (Score:4, Insightful)

    by shreevatsa ( 845645 ) <<shreevatsa.slashdot> <at> <gmail.com>> on Sunday July 31, 2005 @02:33PM (#13208809)
    This is seen, by some, as a move that will bring up Opera's usage stats a bit higher
    Bring up the usage stats, or maybe, thanks to the websites that don't even serve you if you're not using IE, bring down the usage itself? (Hopefully not the latter!)


    ...will hopefully make webmasters, who develop IE centric sites, more aware of Opera.
    More aware of the standards, you mean.
    Anyway, Opera has much fewer users than Firefox, so I think any difference that Opera makes will be much less than what Firefox would.... still, it's a good thing; I wonder if Opera users weren't ashamed all this while to be identified as IE users? :P
  • by Sr. Pato ( 900333 ) on Sunday July 31, 2005 @02:34PM (#13208823) Homepage
    Opera has the option to identify itself as Opera, Mozilla, and IE. IE is by default, for some reason which I don't know (anyone care to explain?) why. Anyone using Opera would probably already be savvy enough to change those settings if they wanted too. But some people are just too lazy, and since there's no real benefit to it, they just leave it as is.
    Expect IE's market share to drop a bit, and for Opera's to go up. :-) Not significantly though, but it's a step in the right direction.
    It's useful, but there's no reason why someone else's browser should be set by default. Don't know, I just never really understood why they did that to begin with.

    All-in-all, my point was, that although this is a good thing for the numbers, it's not something largely significant.
    • by kronocide ( 209440 ) on Sunday July 31, 2005 @02:38PM (#13208852) Homepage Journal
      Because some sites will simply give you an error page if the agent is not IE or possibly Mozilla. Since Opera is highly IE compatible, it's meaningful to circumvent that "feature" of some sites and just pretend to be an IE browser. I hope this is a sign that Opera is now common enough so that the Opera people feel confident that site owners will not filter them out.
      • That does not fix the problem. What will fix the problem is accurate reporting of the user agent and repeated complaints to the site authors until the site is fixed to not blindly cough up error messages, or at least allow users to browse with a warning that not all functions may work.

        I do that for my page -- it's written to W3C standards and is validated by their checker. It tells you that it's standardized and to use a standards-complaint browser -- but it won't stop you from deciding to browse anyway wit
        • I simply test my site in IE, Opera, and Firefox, and assume that people not using them know what they're doing. ;-) Well, if it works in those three without any browser-dependent code fixes, it usually works everywhere.
          • Yah, that's pretty much what I did. I coded it and then I validated it and ran it past the major browsers (and asked an opera-using friend to browse and send me a few screenshots). Works all right in everything.

            I really don't understand this current mentality that it's OK for a site to turn anyone away instead of letting them try browsing anyway. And I also don't understand why some of them use some strange scripting language that Firefox (I use a Mac) can't figure out at all. And then depend on it for almo
  • I tried to read the linked article with the new Opera release, but it said it only supports IE ...
  • With MS warning everyone to update the browser sniffing libs for the IE7 release, now is the time to make changes for those who are trying to do something where the client browser gets funky with specific HTML coding.
  • Comment removed (Score:5, Informative)

    by account_deleted ( 4530225 ) on Sunday July 31, 2005 @02:43PM (#13208888)
    Comment removed based on user account deletion

    • Exactly. Only really really braindead software actually misidentifies Opera, so its usage stats will likely not shoot up any significant amount. What will happen though is webpages from 1998 will have to be updated to stop checking for IE vs NS4 with silly useragent checks and start using object existance checks.
      • What will happen though is webpages from 1998 will have to be updated to stop checking for IE vs NS4 with silly useragent checks

        Fat Chance. Anyone who is still doing NS4 detection is probably using a library script that they don't understand. If they aren't checking for Opera now, this change alone won't cause them to start.

        and start using object existance checks

        I tested a script-heavy site with an old version of Opera. It turns out the DOM objects and methods existed, but didn't do anything. I guess Opera
    • Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1) Opera 7.54 [en]

      The interesting part is, that if read the HTTP [sunsite.dk] specification, and then interpret accordingly, this string contains three product tokens and one comment. The product tokens are "Mozilla/4.0", "Opera", and "7.54". So actually it does not identify itself as MSIE, neither does IE. The [en] part is syntactically wrong, and really should have been part of a comment. Another mistake is that "Opera" and "7.54" are given as two different product na
    • So in reality, it's Opera spoofing Mozilla 4.0 spoofing MSIE 6.0 spoofing Opera...
  • by DamienMcKenna ( 181101 ) <{moc.annek-cm} {ta} {neimad}> on Sunday July 31, 2005 @02:43PM (#13208889)
    In the past few days Eric Meyer [meyerweb.com], CSS guru and general cool guy, released a version 1.1 of his wondeful S5 presentation system [meyerweb.com]. Right afterwards a part-time employee of Opera Software posted a rant on his weblog [opera.com] bitching that Eric gives Opera the "cold shoulder" and questioning S5's status as being cross-browser compatible. As Eric says in a follow-up blog on the topic [meyerweb.com]
    Lying about S5's cross-browser nature? Giving Opera the cold shoulder? Utterly wrong on both counts. I've done everything I can to make sure Opera is still at this particular table.

    As a test Eric disabled the Opera-validation code, changed Opera to properly identify itself and ran the default S5 slideshow...
    Everything worked just fine except for two things. One, the browser window had a vertical scroll bar for no apparent reason. Two, the controls were nowhere to be found, either by hovering over where they're supposed to be or using the "C" key to toggle them.

    So is it possible that Opera took this as a slap in the face and maybe are starting to change their opinion of their place in the world, i.e. "if I can't easily detect your browser I can't begin to fix my code"? Are they trying to stand up against the PR machine that Firefox has behind it to say that they're still in the running, and maybe also make life easier for web developers who'll finally be able to easily identify their browser?

    No matter what the reasons, its a good decision IMHO.

    Damien
    • "So is it possible that Opera took this as a slap in the face and maybe are starting to change their opinion of their place in the world, i.e. "if I can't easily detect your browser I can't begin to fix my code"?"

      No.

      First of all, Meyer might be a big CSS guru and all, but the creator of CSS actually works for Opera, and Meyer's word on browser useragent strings doesn't really make much of a difference if you are going to use Opera on real web sites.

      Also, you can easily detect Opera even when you ide

  • I haven't touched Opera since I switched to Firefox a year ago, but one feature I miss is Opera's ability to advertise itself as IE 5!
  • The browser name in the browser should be configurable. You end up with browser nazi sites like this one run by an anti-IE-nazi [caffeine-junkies.com] that put up nasty messages based on your browser. The ability to change the browser name could help get around this type of bad web site design.
  • Prediction: (Score:2, Funny)

    by Anonymous Coward
    IE usage statistics will be down by two.
     
  • This was kind of silly. About time.
  • Necessary evil... (Score:3, Informative)

    by Lisandro ( 799651 ) on Sunday July 31, 2005 @03:27PM (#13209139)
    Thing is, while Opera can render pages "designed for IE" just fine, a lot of sites still refuse to load [opera.com] if the browser's not IE. Nevermind sites like Hotmail, which deliver purposedly broken CSS [theregister.co.uk] if the browser detected is Opera - making the page look funny or disabling functionality like purging of the spam mail folder.

        Opera makes it easy to change the browser identification (via "Quick preferences"), but still, it can be annoying. Specially for non-technical users.
  • A Question (Score:3, Interesting)

    by fossa ( 212602 ) <pat7.gmx@net> on Sunday July 31, 2005 @03:44PM (#13209238) Journal

    Ok, browsers have User Agent strings. Not all browsers are compatible with every web standard. Websites are becoming more complex (google maps etc.) and taking advantage of newer browsers. So, the question is, do we limit ourselves to the lowest common denominator (among browsers above a certain market share threshold at least), or do we make sites that can change depending on the browser?

    If yes, then should the site do browser detection and serve up different pages? If not (and I think if certainly should be "not"), then how do we go about supporting an ever widening gap in browser features? Simply wait for all browsers above our threshold market share to catch up? I suppose that's what we do now, but it's quite annoying to not be able to use some nice features because of that.

    Another thought: web apps (vs. installed apps) have the great advantage of being upgradeable with no user action. But eventually we get to the point where upgrades require the user to take action and upgrade her browser... So the web app just serves as a buffer to user action.

  • In Two Minds (Score:3, Interesting)

    by gidds ( 56397 ) <[ku.em.sddig] [ta] [todhsals]> on Sunday July 31, 2005 @04:00PM (#13209351) Homepage
    ...about this issue, at least.

    On the one mind, I agree it's ghastly that Opera (or Safari, or Firefox, or whatever) has to pretend to be MSIE just to get served certain web pages. Changing the string might inconvenience some users in the short term, but it'll encourage web authors to better support other browsers, which is a Good Thing(tm) in the longer term.

    But on the other, aside from stats, why should it have to identify itself at all? What's wrong with something like

    Mozilla/5.0 (compatible) Exact browser name none of your business
    or similar? Groucho Marx is quoted as saying that he wouldn't want to belong to any club that'd have him as a member; I feel the same about web sites; if a site has to customise its pages for my browser, whatever that browser is, then I'm suspicious of it.

    • Re:In Two Minds (Score:5, Insightful)

      by myov ( 177946 ) on Sunday July 31, 2005 @05:47PM (#13209843)
      The Mozilla/5.0 string should also go. After all, every browser is pretending to be Netscape, and it's become redundant now.

      What the string should should indicate is what version of the various standards it supports. Something along the lines of:
      HTML/4.0 CSS/2 PDF JPEG PNG etc.

      Don't support CSS? You get the table layout. Don't support HTML 3? You get an upgrade message. Etc.

      The string itself would need to be enforced by the W3C so we can't get something like MS's "we'll impliment what we want or make our own standard" attitude. Supporting CSS 2 means you support the spec entirely, and it's no indication that the browser is IE, Firefox or anything else, which means you can't code to one browser.
  • We need a new borg icon for the Firefox fanboys. Just like the Billy-G worshippers, any time anyone mentions the virtues of a non-Firefox browser they shit their pants, whip out their willies, and start jacking off at the altar of free software.

    I've said it before and I'll say it again: ANY monoculture is a bad monoculture. It doesn't matter what the dominate monoculture is, it's always a bad thing to have a market overwhelmed by a single product. If the fanboys had any brains at all they'd welcome ever
  • Stop spoofing as Netscape 4?

  • by arose ( 644256 ) on Sunday July 31, 2005 @05:55PM (#13209872)
    I'm looking at you Internet Explorer [msdn.com], you Safari [apple.com] and you Konqueror [konqueror.org] (they don't even tell you the default, but on Ubuntu it spoofs as "Mozilla/5.0 (compatible;" as well as "(like Gecko)" ).

On the eighth day, God created FORTRAN.

Working...