Recommended Posts

Hey, today i noticed that google loads in nightly the old-fashioned search in nightly. I tried with a new profile too. It doesn't happen in Aurora, nor in Chromium, nor in IE9.

Some screenshots:

err2.PNG

Also in the italian version the logo is loaded incorrectly:

err1.PNG

This things make me realize how bad looking the internet was in the past :)

Google often use cookies to set how the site should appear to you when they are testing stuff (for example if you get the new Google bar or not). In this case, it appears your Nightly has somehow got one to give you the old design, which must be some sort of error on Google's end. Nice to see though!

Google often use cookies to set how the site should appear to you when they are testing stuff (for example if you get the new Google bar or not). In this case, it appears your Nightly has somehow got one to give you the old design, which must be some sort of error on Google's end. Nice to see though!

I thought that too, but after deleting the cache and cookies nothing changed

I got the same problem in nightly but changing the user agent fixes it for me. In about:config add "general.useragent.override" and string value "Mozilla/5.0 (Windows NT 6.1; WOW64; rv:10.0.2) Gecko/20100101 Firefox/10.0.2" or whatever previous version of firefox.

Websites sometimes appear differently on other browsers and they detect that by sniffing the user agent. Try using iPad safari browser user agent and you can emulate googles iPad layout on firefox.

Mozilla/5.0 (iPad; U; CPU OS 4_3 like Mac OS X; en-us) AppleWebKit/533.17.9 (KHTML, like Gecko) Version/5.0.2 Mobile/8F190 Safari/6533.18.5

edit:

It appears that today's nightly changed user agent formatting. https://bugzilla.mozilla.org/show_bug.cgi?id=588909

old nightly: Mozilla/5.0 (Windows NT 6.1; rv:13.0a1) Gecko/20120217 Firefox/13.0a1

today's nightly: Mozilla/5.0 (Windows NT 6.1; rv:13.0a1) Gecko/13.0a1 Firefox/13.0a1

This broke the google. So temp workaround is to manually change user agent or wait for google to fix it on their end.

The probably used that part of the useragent to detect if the rendering engine is modern enough for the standard design otherwise it will revert back to the old for compatibility reasons. The changing of the date format broke this backwards compatibility (Better to use dates than version numbers it would be simpler to do when Nightlies are involved which won't have a predictable version number)

Hey, today i noticed that google loads in nightly the old-fashioned search in nightly. I tried with a new profile too. It doesn't happen in Aurora, nor in Chromium, nor in IE9.

Some screenshots:

err2.PNG

Also in the italian version the logo is loaded incorrectly:

err1.PNG

This things make me realize how bad looking the internet was in the past :)

How bad looking it was? What's bad about that? It's called simple and simplistic and it loaded very fast too, I bet, which is how I wish it still was instead of all the bloat/crap Google has on their page and in their browser!! I'm not complaining about the speed of Google loading, but if everyone was still using dial up, I'd bet there be more complaining about it then?

90% of the reason I WILL NOT use anything Google!! It IS NOT the greatest thing since the wheel, or internet!! Google sucks green donkey d**k!!

How bad looking it was? What's bad about that? It's called simple and simplistic and it loaded very fast too, I bet, which is how I wish it still was instead of all the bloat/crap Google has on their page and in their browser!! I'm not complaining about the speed of Google loading, but if everyone was still using dial up, I'd bet there be more complaining about it then?

Nowadays version is minimalistic and loads almost instantly.

Just when you couldn't hate mozilla any more than you currently do.

They go and change this, which adds zero benefit but ****s every standard up.

Oh way to go...

What utter tosh. Why not actually read the bug ticket? UA strings are hardly a strict standard, as it is (and have a loooad of historical complexities around them). Sites should be feature sniffing, not browser sniffing. Do you have much experience with how UA strings work?

As it is, it's not going to stay broken/weird for long. Google is likely thinking the browser version is much older than it is.

"Sites should be feature sniffing, not browser sniffing. Do you have much experience with how UA strings work?"

Guess you've never been a web developer in your entire at all then.

Javascript comes with chome, firefox, etc. yes. A lot of browsers do not have javascript support (lynx, links, etc.), or have it disabled (no$cript).

This is just plain badly written code on Google's behalf, instead of checking for the version of the browser they check for the date it was compiled on, which is pretty damn meaningless (You could have a lower version with a higher build date, etc.) There's a reason Mozilla froze that date to be 2010-01-01 with 4.0, to discourage this, Google is just apparently slow to notice.

Edit: Ahh, I see the JS angle then. Still, your site shouldn't rely on JS to function and this is a server side issue on Google's behalf, not a client side issue.

"Sites should be feature sniffing, not browser sniffing. Do you have much experience with how UA strings work?"

Guess you've never been a web developer in your entire at all then.

Javascript comes with chome, firefox, etc. yes. A lot of browsers do not have javascript support (lynx, links, etc.), or have it disabled (no$cript).

It's not an either/or situation, though - browsers inherently without JS support could still be UA detected. I don't have much sympathy for people who would noscript Google - of course you're going to lose features that way. Just like if I turn off CSS, sites will also look fairly weird - but I'm not going to complain about it. (Turn off the scripts you don't want to run, for sure, but turning off *all* scripts and expecting everything to be okay is over the top)

There's many articles online about browser and feature sniffing, and the benefits of each - but they pretty much always come down on feature sniffing. I know browser sniffing is commonly used, but that doesn't make it good.

Besides, that's not a defence of your original comment at all, which was that somehow it was all Mozilla's fault.

This topic is now closed to further replies.
  • Posts

    • If I could, I would commemorate it the best way possible: Replacing old machines that are still running Windows XP with something more modern, stable and better.     Noone and nothing should be running Windows XP in 2026.
    • Google's new hand-wave reCAPTCHA can be bypassed with a stock photo by Ivan Jenic Image: Screenshot Google is testing a new reCAPTCHA method that asks you to wave at your camera to prove you're human. So, besides solving puzzles and reading distorted text, you can now use your computer’s camera to pass the verification test. When the hand gesture verification is triggered, your browser asks for camera access and prompts you to perform a simple gesture, like a wave or an open palm. Google says it records a short video of the movement and uses AI to extract 21 hand-knuckle coordinates to complete the verification process. The video is then immediately deleted, and Google swears it doesn't keep it. The process alone can be uncomfortable for people who wouldn’t want their biometric data, which hand scans technically qualify as, recorded. But it gets even more nuanced, as early testers discovered that the new hand-waving reCAPTCHA can be passed with a simple stock image. A user on X tested the new challenge using a stock image of a hand fed through OBS Virtual Camera, and it passed. I wanted to verify it, so I tried the same thing. It took me a few tries and a few stock images, but in the end, I was also able to pass the test. I simply had to readjust the stock image of a generic person waving inside OBS, and Google’s mechanism registered it as a legitimate hand gesture. Once again, it didn’t even have to be a video or an AI-generated hand animation. Given the simplicity of the process, the entire action can be automated in minutes. All it takes is a simple Python script to render the new reCAPTCHA method obsolete. And it doesn’t even have to be an AI bot, which is usually used for solving puzzles and other verification methods. The new reCAPTCHA method is still in its early phase, and Google will, hopefully, update its AI to at least reject still images. However, this incident, combined with users’ initial skepticism about Google’s practices regarding user data, likely won’t make too many people wave at the camera anytime soon.
    • 🤣🤣🤣🤣🤣 "to fund healthcare and tuition" 🤣🤣🤣🤣 Who do you think you are talking about, some COMMUNIST? We are better than them, doG bless Murica!!! p.s. I'm from a country where government does exactly that, i.e. not form US.
    • Apparently not. I know it is on Edge for business at the moment, but how long will it be before it become on the home version of Edge?
  • Recent Achievements

    • First Post
      carols23 earned a badge
      First Post
    • One Month Later
      Tom Willson earned a badge
      One Month Later
    • Apprentice
      Asgardi went up a rank
      Apprentice
    • One Month Later
      sunrisea2milk earned a badge
      One Month Later
    • Week One Done
      sunrisea2milk earned a badge
      Week One Done
  • Popular Contributors

    1. 1
      +primortal
      496
    2. 2
      +Edouard
      256
    3. 3
      PsYcHoKiLLa
      155
    4. 4
      Steven P.
      89
    5. 5
      macoman
      66
  • Tell a friend

    Love Neowin? Tell a friend!