3 ms·

> The situation isn't as bad for desktop browsers Don't worry, it will get worse. Firefox is now forced to accept `-webkit-` in css (http://www.theregister.co.

by skarap 11y ago

> The situation isn't as bad for desktop browsers Don't worry, it will get worse. Firefox is now forced to accept `-webkit-` in css (http://www.theregister.co.uk/2016/01/04/firefox_webkit_css_support/ http://www.theregister.co.uk/2016/01/04/firefox_webkit_css_s...) not to be "broken". This will obviously lead to even less Firefox testing. Hell, even I (my team) am contributing to this! We're working on Angular-based web SPA now. It hasn't got much functionality yet, but already has quite a few bugs when viewed in Firefox. And the usual answer from frontend guys is "It's a firefox bug!".

digi_owl 11y ago
I suspect it will go beyond CSS. There are a certain kind of app on Android that makes the device FS available via a web browser. Thing is that these apps invariably makes use of a Chrome-ism when dealing with drag and drop. In Firefox you can only drag a single file at a time. drag over multiple, or a folder, and there won't be much of a result. In Chrome however it will transfer over just fine. This has been sitting in the Mozilla bugtracker for some time, because the Chrome behavior goes beyond the HTML spec for file uploads.
- yyhhsj0521 11y ago
  Yeah sometimes I wonder if providing extra behaviors is good or bad. Certainly it could save developers a lot of works and save users a lot of time, but it may become next ActiveXObject("Microsoft.XMLHTTP"); Maybe Chrome should put a warning into the console, saying that this function is experimental.
emn13 11y ago
I remember fixing one of frontend bugs in an SPA and contrary to the claims it turned out to be a webkit bug that the code relied on (the app had been developed on a mac, so it's fairly easy to only test on webkit/blink). Lovely. Aside/Rant: I also get the impression that frontend devs have never read the CSS2.1 spec or bothered to understand it (CSS3 is longer, and consists largely of special-purpose extra modules, so a complete reading is less valuable and more difficult). There's somehow this feeling that since CSS "isn't a real programming language" you should be able to use it intuitively. And when things don't work out and you fiddle until it works without giving thought to why, that's OK. That's such a huge waste of time. The box model may not be 100% intuitive, but boy, if that's your standard, what you doing in IT?
- masklinn 11y ago
  > the app had been developed on a mac, so it's fairly easy to only test on webkit/blink It's no harder to only test on Chrome in Linux and IE in Windows, I fail to see how that's an excuse. The only excuses are IE testing when developing on OSX and Safari testing when developing on Windows, it's trivial to install both Firefox and Chrome on any OS.
  - emn13 11y ago
    Actually, it's easier to test for IE versions on linux/mac due to http://xdissent.github.io/ievms/ http://xdissent.github.io/ievms/ - You can install the VMs on windows too (of course), but I've never stumbled upon quite as simple a solution. On windows I tend to just rely on compat mode, which is surprisingly good, but far from perfect.

by skarap 11y ago

digi_owl 11y ago

I suspect it will go beyond CSS. There are a certain kind of app on Android that makes the device FS available via a web browser. Thing is that these apps invariably makes use of a Chrome-ism when dealing with drag and drop. In Firefox you can only drag a single file at a time. drag over multiple, or a folder, and there won't be much of a result. In Chrome however it will transfer over just fine. This has been sitting in the Mozilla bugtracker for some time, because the Chrome behavior goes beyond the HTML spec for file uploads.

yyhhsj0521 11y ago
Yeah sometimes I wonder if providing extra behaviors is good or bad. Certainly it could save developers a lot of works and save users a lot of time, but it may become next ActiveXObject("Microsoft.XMLHTTP"); Maybe Chrome should put a warning into the console, saying that this function is experimental.

emn13 11y ago

I remember fixing one of frontend bugs in an SPA and contrary to the claims it turned out to be a webkit bug that the code relied on (the app had been developed on a mac, so it's fairly easy to only test on webkit/blink). Lovely. Aside/Rant: I also get the impression that frontend devs have never read the CSS2.1 spec or bothered to understand it (CSS3 is longer, and consists largely of special-purpose extra modules, so a complete reading is less valuable and more difficult). There's somehow this feeling that since CSS "isn't a real programming language" you should be able to use it intuitively. And when things don't work out and you fiddle until it works without giving thought to why, that's OK. That's such a huge waste of time. The box model may not be 100% intuitive, but boy, if that's your standard, what you doing in IT?

masklinn 11y ago
> the app had been developed on a mac, so it's fairly easy to only test on webkit/blink It's no harder to only test on Chrome in Linux and IE in Windows, I fail to see how that's an excuse. The only excuses are IE testing when developing on OSX and Safari testing when developing on Windows, it's trivial to install both Firefox and Chrome on any OS.
- emn13 11y ago
  Actually, it's easier to test for IE versions on linux/mac due to http://xdissent.github.io/ievms/ http://xdissent.github.io/ievms/ - You can install the VMs on windows too (of course), but I've never stumbled upon quite as simple a solution. On windows I tend to just rely on compat mode, which is surprisingly good, but far from perfect.