Speaking of Firefox Usability



  • On the left, Internet Explorer 9.  On the right, Firefox 21.

    With Firefox, I can post a message, but I cannot reply to or quote a specific message because the necessary widgets are missising.  Also, using Firefox, I cannot post a comment on any blog on Google's Blogspot.com.  After hitting submit, the comment just disappears, never to be seen again.  IE works just fine.

    The number of websites that do not work properly with Firefox seems to be growing.

    Community Server, WTF - had to play around with resizing the pictures.  The two pictures would be side by side in compose mode, but the Firefox picture would be under the IE picture in the preview.

     



  • You are zoomed in slightly in Firefox, which may affect layout, because page coders never take that feature into consideration.

    Or, you are slightly zoomed out in IE.

    Better have a look!

     




  • @dhromed said:

    You are zoomed in slightly in Firefox, which may affect layout, because page coders never take that feature into consideration.

    Or, you are slightly zoomed out in IE.

    Better have a look!

    IE is zoomed in to 115%.  Necessary due to a combination of bad eyes and a big monitor.  Firefox is zoomed in slightly as well, but it doesn't give me an exact number like IE does.

    No amount of zoomng in or out fixes Firefox.  But it was worth a try.

     



  • @dhromed said:

    You are zoomed in slightly in Firefox, which may affect layout, because page coders never take that feature into consideration.

    Or, you are slightly zoomed out in IE.

    Better have a look!

    I never understood why browsers let the page know when they are zoomed in. Faking a smaller window size and then transforming the coordinates would probably cause a lot fewer layout bugs.



  • As much as I've begun to hate Firefox as of late, this really just seems like a classic mundane example of a site testing only in IE. Is there a specific reason you think this is a Firefox fail and not a web developer fail?



  • @SamC said:

    As much as I've begun to hate Firefox as of late, this really just seems like a classic mundane example of a site testing only in IE. Is there a specific reason you think this is a Firefox fail and not a web developer fail?

    Well, it's Ars Technica, and they're not people I'd assume developed only for IE (or in IE at all..)



  • Did you try disabling your adblocker?

    Okay, as much as troll with that one, serious question. So many sites seem to think it's a good idea to put their scripts that load their navigation shit in the same call that loads their pixel trackers and shit.  Or they put their navigation wireup code right after the call to LoadFifteenMillionTrackersOrCrashTheEntirePage();



  • @Lorne Kates said:

    Did you try disabling your adblocker?

    Okay, as much as troll with that one, serious question. So many sites seem to think it's a good idea to put their scripts that load their navigation shit in the same call that loads their pixel trackers and shit.  Or they put their navigation wireup code right after the call to LoadFifteenMillionTrackersOrCrashTheEntirePage();

    I think that's a perfectly reasonable tactic to make people turn off ad-blockers. I've seen a trend for copyright-ignoring streaming sites to cover functional sections of the page with banner ads that are unclosable if you're running an adblocker, which makes perfect sense to me.



  • @TDWTF123 said:

    I think that's a perfectly reasonable tactic to make people turn off ad-blockers.

    All people are going to see is a broken page and they won't realise that their adblocker is the cause.
    What they will do is tell all their friends not to use your website because it sucks balls.



  • @Salamander said:

    All people are going to see is a broken page and they won't realise that their adblocker is the cause.

    This is why I prefer a nice, big "Turn off your adblocker, cocksucker" goatse.

    @Salamander said:

    What they will do is tell all their friends not to use your website because it sucks balls.

    Meh, their friends are probably all crooks, too, so good riddance.



  • @dhromed said:

    You are zoomed in slightly in Firefox, which may affect layout, because page coders never take that feature into consideration.

    Or, you are slightly zoomed out in IE.

    Better have a look.

     Arstechnica is one of the few sites where it actually works properly.

     



  • @Ben L. said:

    I never understood why browsers let the page know when they are zoomed in. Faking a smaller window size and then transforming the coordinates would probably cause a lot fewer layout bugs.

    Well, you have it perfectly backwards. Browsers do not notify the page when they are zoomed in, but only notify the page that the window has been resized, and try to redo all the layout themselves. It's actually worse than that : browsers do not even allow the page to know what is the current zoom level, apparently because this is a user setting that belongs outside of the sandbox in which the page is allowed to play.

    In most cases it works just fine. However, when zoomed in our out, they need to play with floating point pixels and coordinates internally. And of course they all have a few floating point derived rounding errors. And none of them makes the same rounding errors, which typically lead to off-by-one errors in one or another browser, which in turn leads to what you can see there.

    This is especially common when you start using float:left in your css and that due to one of these off-by-one errors your floating content is pushed to below is sibling and then cropped out by it parents overflow:hidden attribute.



  • @LordOfThePigs said:


    In most cases it works just fine. However, when zoomed in our out, they need to play with floating point pixels and coordinates internally. And of course they all have a few floating point derived rounding errors. And none of them makes the same rounding errors, which typically lead to off-by-one errors in one or another browser, which in turn leads to what you can see there.

    This is especially common when you start using float:left in your css and that due to one of these off-by-one errors your floating content is pushed to below is sibling and then cropped out by it parents overflow:hidden attribute.

    That doesn't make sense. All the layout should be done with integers, and then scaled up (or down) for the actual drawing. No need for floats. In a perfect world, zooming would be a feature of the window system, so that even the browser wouldn't know about it and couldn't get it wrong.


  • Discourse touched me in a no-no place

    @Planar said:

    In a perfect world, zooming would be a feature of the window system, so that even the browser wouldn't know about it and couldn't get it wrong.
    That's actually one of the features of OSX. I hold down Ctrl and use the scrolling mouse action (which depends on whether I'm using a touchpad or a physical mouse with a scrollwheel) and everything scales. I find it's only really useful when you're looking at off-by-one errors in pixel layouts, but YMMV.

    Of course, browsers decide that it is necessary to implement their own scaling scheme as well…



  • @Ben L. said:

    Faking a smaller window size and then transforming the coordinates would probably cause a lot fewer layout bugs.
     

    No, because I hate fullpage zoom, since it also enlarges sidebars, titlebars, images etc, which distorts the layout significantly, and I only want text to zoom because that's what I can't read comfortably.

    Also, with full page zoom, images are upsampled and will always look worse than at 100%.

    Full page zoom bad. Text zoom good.



  • @TDWTF123 said:

    @Lorne Kates said:

    Did you try disabling your adblocker?

    Okay, as much as troll with that one, serious question. So many sites seem to think it's a good idea to put their scripts that load their navigation shit in the same call that loads their pixel trackers and shit.  Or they put their navigation wireup code right after the call to LoadFifteenMillionTrackersOrCrashTheEntirePage();

    I think that's a perfectly reasonable tactic to make people turn off ad-blockers. I've seen a trend for copyright-ignoring streaming sites to cover functional sections of the page with banner ads that are unclosable if you're running an adblocker, which makes perfect sense to me.
     

    I'm surprised that that tactic works, given that the purpose of an adblocker is to get rid of unwanted elements from the page. If someone's running an adblocker and comes across one of those banners, their most obvious response is going to be to tell the adblocker to block it.

    (A similar hate of mine: I browse with JavaScript off by default because it makes sites so much less annoying, and come across the occasional site that covers the entire page with an element that says that the page doesn't work with JavaScript off. My usual approach is to remove the element in question with Firebug, upon which the page normally works fine. An adblocker would typically work too; they're not just for blocking ads, people! I likewise use them to block interface elements of sites I disilke, such as avatars on forums. Or various people's signatures.)

     



  • I agree with the hate for thoses sites that require javascript. I can understand that for a tiny subsection of the web who actually do something useful on the page (say, they are hames, or google doc, or jukebox), but 99% of the time it's for shitty visual effects since I actually go on site to read information and not to do a shittier version of applications already installed on my computer. I find especially jarring the fact that they often don't explain why they need javascript. Unknown program that want special right in my computer without explanation are suspiciously like malware.



  • If I ever designed a video streaming site, the ads would come from a URI like http://data2.example.com/0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33.flv.

    The videos would come from a URI like http://data2.example.com/bb928ca332f5f4fa2cdaef238672e0fbcf5e7a0f.flv.

    Seriously, all it takes it to have first party ads that don't look like ads.



  • @TDWTF123 said:

    I think that's a perfectly reasonable tactic to make people turn off ad-blockers.
     

    Sure, because hosting critical scripts that your site absolutely 100% requires on unrelated, unreliable 3rd party servers is a great idea.

    No, wait, even ignoring the whole ads thing, that's a goddamn fucking stupid idea.  Have fun spending hours troubleshooting why a large swath of customers can't access the site, only to find out there's the occasional network outage in some part of your ad network's CDN.

    ESPECIALLY given that it's not just <script src='adnetwork.com/nav.js' />.  It's usually src='cdn.com/loader.js', which loads three different 'cdn.partners.com/tracker.js', which each in turn load three different 'a.cdn.inedo.com/tools.js', which in turn load 'api.google.com/jquery-latest.js'.

    I would never even use a direct include of api.google.com/jquery if my site was critically dependent on jquery. I'd suck it up and stick the 10k of minified js on a nicely cached first party server.



  • @MiffTheFox said:

    If I ever designed a video streaming site, the ads would come from a URI like http://data2.example.com/0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33.flv.

    The videos would come from a URI like http://data2.example.com/bb928ca332f5f4fa2cdaef238672e0fbcf5e7a0f.flv.

    Seriously, all it takes it to have first party ads that don't look like ads.

     

    People have been saying they should do this for years. And given the saturation of the market with 3rd party advertisers, they never will. They needs their tracking and marketing data to keep their business afloat. Why are you trying to steal money from the mouths of the ad networks with your shitty suggestion, you fucking leech?

    Also, video streaming sites can't afford the bandwidth to host first party ads. What do you think they are, video streaming sites or something?

     


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.