Which browsers should I support?
-
Dear God no. That means standardising on an incredibly awful implementation set. IE6 is at least a whole generation old and should be treated as a special case, not a direct target.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
I decided to post here instead of the web forum to get more opinions. My boss has asked me to come up with a 'browser support strategy' that covers both internal projects and external projects. The reason for this is mostly: in a just released a public web site we are getting feedback like "this doesn't look right on my Kindle Fire" there is too much white space at the top (or whatever). The problem is that we could chew up the rest of our careers responding to this if we aren't careful - currently the person in charge says 'we should support all mobile devices out there' and we have spent hours on stuff that (in my opinion) we shouldn't have. We are getting some loaner devices so we can at least see for ourselves what the problem is (rather than rely on an email description of what's wrong - that's an improvement. We have to support IE7 because of the number of users on it within our organization - I'd love to ditch it but people are only moving off about 0.5% per month and it has a 30% share of current page hits. All help is appreciated. I searched and haven't find anything relevant except a graded browser support idea that I found at yuilibrary.com. My current thought is to have categories like: A = we will fix it if something looks wrong, QA will test these browsers B = it only has to look roughly the same as tier A, we will only fix it if something is missing or doesn't work, QA does NOT test these browsers C = These are considered rare browser platforms and users should not expect them to work. If money is provided to support a specific broser in the category we can do it, but it will not be supported ongoing (if it works now and breaks in a month we will not fix the break). F = These are antiquated browsers which are not supported for any reason Do any of you have a browser list that you target support for? Any recommendations? (Aside from drop IE7 - we're working on that battle) Comments on how to word my ABCF scale and what should be in it are welcome also. Thanks, -Chris C. EDIT: Example of Tier "F" would be IE6, IE5.5, etc.)
Your approach is the right one and Jörgen is wise. Aim to support the most common browsers (FF - Gecko, Chrome/Safari - webkit, and IE7+ - Trident). Focus on making the site usable (but not pixel perfect) on all browsers above your cutoff line, and provide different levels of features: 1. Mainstream browsers: The site should have full functionality and look great. Do as much as you can without resorting to HTML5/CSS3. Maybe you can't use a canvas, or maybe you can't use the file upload progress capabilities or local storage, but it all works. IE7+ and all other browsers fit in here 2. New browsers: Using progressive enhancement by sniffing for feature availability, add the features you feel are the icing on the cake and/or are feature you want as mainstream in a year or two. The experience on newer browsers should be the same, but little things just work nicer and look nicer. IE9, Chrome, Firefox and Safari latest fit into this. 3. Old dying browsers: information should be accessible and readable. Maybe something doesn't work like drag and drop, maybe you have no shadows or rounded corners. Maybe your DIVS are a little out of alignment. It doesn't matter. It all still works. 4. Mobile browsers: Don't even bother with non-HTML mobile browsers. Consider all mobile browsers HTML5 enabled and focus on a fast, clean, readable experience geared towards touch. It may mean a completely different site 5. Odd, dead or beta browsers. Chrome eleventy-one, Firefox 2.0. IE5 on a Mac. Lynx. Take a moment to see if you have customers that actually use, and need to use, any of these. If they do, support it on an as-needed basis similar to #3. In all of this, ask yourself whether you expect all your users will have Javascript and images enabled, and whether you can reasonably require this to be the case in order to view the site. And then, when that's all done, fix the issues in the correct order.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
Sorry, but that doesn't work. Browser support isn't limited to Javascript - it includes the gamut of CSS quirks that drive sane men mad.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
eh?
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert Einstein -
eh?
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert EinsteinYou said "Do feature detection rather than browser detection" and I was assuming you meant javascript based feature detection. This means you're only handling script issues, not layout issues, and the majority of issues with browser support for the most painful (read: older IE) browsers is layout. Of course, maybe I'm reading too much into the question and simply having 'nam flashbacks of my own adventures in browser support. I find the javascript support the easy stuff. Layout and styling? That's a killer.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
You said "Do feature detection rather than browser detection" and I was assuming you meant javascript based feature detection. This means you're only handling script issues, not layout issues, and the majority of issues with browser support for the most painful (read: older IE) browsers is layout. Of course, maybe I'm reading too much into the question and simply having 'nam flashbacks of my own adventures in browser support. I find the javascript support the easy stuff. Layout and styling? That's a killer.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
What would you want for mailing these words to my boss? :) We (the R&D staff) has managed to hold him off HTML (vs native mobile UIs), for some time now, but he keeps picking up the subject during meetings. It's getting bothersome... :)
-- Kein Mitleid Für Die Mehrheit
-
You said "Do feature detection rather than browser detection" and I was assuming you meant javascript based feature detection. This means you're only handling script issues, not layout issues, and the majority of issues with browser support for the most painful (read: older IE) browsers is layout. Of course, maybe I'm reading too much into the question and simply having 'nam flashbacks of my own adventures in browser support. I find the javascript support the easy stuff. Layout and styling? That's a killer.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
Well, I'm no expert so perhaps I shouldn't speak. But I was primarily talking about layout & CSS feature detection. Javascript libraries such as Modernizr[^] and it's brethren allow you to write both conditional Javascript and CSS.
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert Einstein -
Dear God no. That means standardising on an incredibly awful implementation set. IE6 is at least a whole generation old and should be treated as a special case, not a direct target.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
Well, I'm no expert so perhaps I shouldn't speak. But I was primarily talking about layout & CSS feature detection. Javascript libraries such as Modernizr[^] and it's brethren allow you to write both conditional Javascript and CSS.
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert EinsteinThat's the point, though: detecting differences in browsers is easy. Making the different browsers work properly is where the work is.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
What? Just because when I say "IE6 should be treated as a special case" and I mean "IE6 should be taken out back and bludgeoned" doesn't mean I have anything against IE6. Much.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
-
That's the point, though: detecting differences in browsers is easy. Making the different browsers work properly is where the work is.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
True
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert Einstein -
What? Just because when I say "IE6 should be treated as a special case" and I mean "IE6 should be taken out back and bludgeoned" doesn't mean I have anything against IE6. Much.
cheers, Chris Maunder The Code Project | Co-founder Microsoft C++ MVP
The only thing that should be supporting IE6 is TNT. With a lit fuse. s s s s s s s s s s s s s s s s s s s s KABOOM!!!!!
I think computer viruses should count as life. I think it says something about human nature that the only form of life we have created so far is purely destructive. We've created life in our own image. Stephen Hawking
-
I decided to post here instead of the web forum to get more opinions. My boss has asked me to come up with a 'browser support strategy' that covers both internal projects and external projects. The reason for this is mostly: in a just released a public web site we are getting feedback like "this doesn't look right on my Kindle Fire" there is too much white space at the top (or whatever). The problem is that we could chew up the rest of our careers responding to this if we aren't careful - currently the person in charge says 'we should support all mobile devices out there' and we have spent hours on stuff that (in my opinion) we shouldn't have. We are getting some loaner devices so we can at least see for ourselves what the problem is (rather than rely on an email description of what's wrong - that's an improvement. We have to support IE7 because of the number of users on it within our organization - I'd love to ditch it but people are only moving off about 0.5% per month and it has a 30% share of current page hits. All help is appreciated. I searched and haven't find anything relevant except a graded browser support idea that I found at yuilibrary.com. My current thought is to have categories like: A = we will fix it if something looks wrong, QA will test these browsers B = it only has to look roughly the same as tier A, we will only fix it if something is missing or doesn't work, QA does NOT test these browsers C = These are considered rare browser platforms and users should not expect them to work. If money is provided to support a specific broser in the category we can do it, but it will not be supported ongoing (if it works now and breaks in a month we will not fix the break). F = These are antiquated browsers which are not supported for any reason Do any of you have a browser list that you target support for? Any recommendations? (Aside from drop IE7 - we're working on that battle) Comments on how to word my ABCF scale and what should be in it are welcome also. Thanks, -Chris C. EDIT: Example of Tier "F" would be IE6, IE5.5, etc.)
You may want to look at http://remysharp.com/2010/10/08/what-is-a-polyfill/ and https://github.com/Modernizr/Modernizr/wiki/HTML5-Cross-browser-Polyfills. Personally I don't work on websites, but on web apps, and use http://qooxdoo.org/. But the framework also has website development support, possibly integrating much stuff listed in the second page linked to above already - I have no idea, since I don't use their website stuff, but judging from what their desktop stuff looks like, they build powerful, solid and comprehensive solutions - nothing half-baked.
-
I decided to post here instead of the web forum to get more opinions. My boss has asked me to come up with a 'browser support strategy' that covers both internal projects and external projects. The reason for this is mostly: in a just released a public web site we are getting feedback like "this doesn't look right on my Kindle Fire" there is too much white space at the top (or whatever). The problem is that we could chew up the rest of our careers responding to this if we aren't careful - currently the person in charge says 'we should support all mobile devices out there' and we have spent hours on stuff that (in my opinion) we shouldn't have. We are getting some loaner devices so we can at least see for ourselves what the problem is (rather than rely on an email description of what's wrong - that's an improvement. We have to support IE7 because of the number of users on it within our organization - I'd love to ditch it but people are only moving off about 0.5% per month and it has a 30% share of current page hits. All help is appreciated. I searched and haven't find anything relevant except a graded browser support idea that I found at yuilibrary.com. My current thought is to have categories like: A = we will fix it if something looks wrong, QA will test these browsers B = it only has to look roughly the same as tier A, we will only fix it if something is missing or doesn't work, QA does NOT test these browsers C = These are considered rare browser platforms and users should not expect them to work. If money is provided to support a specific broser in the category we can do it, but it will not be supported ongoing (if it works now and breaks in a month we will not fix the break). F = These are antiquated browsers which are not supported for any reason Do any of you have a browser list that you target support for? Any recommendations? (Aside from drop IE7 - we're working on that battle) Comments on how to word my ABCF scale and what should be in it are welcome also. Thanks, -Chris C. EDIT: Example of Tier "F" would be IE6, IE5.5, etc.)
It's not just the browser qua browser. It's the OS, too. I'm redesigning my site right now and it looks like crap in Chrome and Firefox. Layout? No. Features? No. Freaking fonts. I can't even get Google fonts to look good in Chrome. Blocky, pixelated, ugly. Looks great in IE. Makes no sense. I hear tell it's because of a setting in Windows itself (font smoothing? Something. Doesn't matter. I have no control over the users' systems, so I have to find some other way of making pretty.)
No dogs or cats are in the classroom. My Mu[sic] My Films My Windows Programs, etc.
-
I decided to post here instead of the web forum to get more opinions. My boss has asked me to come up with a 'browser support strategy' that covers both internal projects and external projects. The reason for this is mostly: in a just released a public web site we are getting feedback like "this doesn't look right on my Kindle Fire" there is too much white space at the top (or whatever). The problem is that we could chew up the rest of our careers responding to this if we aren't careful - currently the person in charge says 'we should support all mobile devices out there' and we have spent hours on stuff that (in my opinion) we shouldn't have. We are getting some loaner devices so we can at least see for ourselves what the problem is (rather than rely on an email description of what's wrong - that's an improvement. We have to support IE7 because of the number of users on it within our organization - I'd love to ditch it but people are only moving off about 0.5% per month and it has a 30% share of current page hits. All help is appreciated. I searched and haven't find anything relevant except a graded browser support idea that I found at yuilibrary.com. My current thought is to have categories like: A = we will fix it if something looks wrong, QA will test these browsers B = it only has to look roughly the same as tier A, we will only fix it if something is missing or doesn't work, QA does NOT test these browsers C = These are considered rare browser platforms and users should not expect them to work. If money is provided to support a specific broser in the category we can do it, but it will not be supported ongoing (if it works now and breaks in a month we will not fix the break). F = These are antiquated browsers which are not supported for any reason Do any of you have a browser list that you target support for? Any recommendations? (Aside from drop IE7 - we're working on that battle) Comments on how to word my ABCF scale and what should be in it are welcome also. Thanks, -Chris C. EDIT: Example of Tier "F" would be IE6, IE5.5, etc.)
There is a difference between having to support IE7, and having to make it look good in IE7. If I were you I would create an adaptive layout that looks good in modern browsers and works on different sized windows to solve the mobile device vs. desktop problem. For the layout details, though, don't limit yourself to CSS that works in IE7. Just accept that things may look not quite right. It's not too hard to make something that works in IE7, though you'll go crazy trying to make it look exactly the same (unless you want it to look like it's 2005 in ALL browsers). If your organization can accept that using an old browser means their web sites aren't going to look like a modern browser, and may have layout issues, then I'd go with that approach. If they expect everything to look snappy in IE7 too, then you're screwed.
-
You should also prioritize in which order you fix problems.
- Broken functionality should be highest priority and include all categories, maybe except F.
- Having all elements looking roughly right and being in the right positions should be next priority and include categories A and B.
- Pixelpushing is time demanding and only important to the marketing department and should therefore have the lowest priority and be limited to category A.
Light moves faster than sound. That is why some people appear bright, until you hear them speak. List of common misconceptions
> You should also prioritize in which order you fix problems. Very good point, that's a fight waiting to happen, I'll add that
-
I suspect you were kidding (and P.S. it did make me grin) IE6 isn't a problem as it is no longer patched by msft. Encouraging users to stay on IE6 would not be a public service. The current 'boat anchor around our neck' is IE7 as it needs its own full section of CSS support and I'm told is difficult to match up to modern browsers (with modern defined as IE9+, FF, Chrome) FWIW
-
I suspect you were kidding (and P.S. it did make me grin) IE6 isn't a problem as it is no longer patched by msft. Encouraging users to stay on IE6 would not be a public service. The current 'boat anchor around our neck' is IE7 as it needs its own full section of CSS support and I'm told is difficult to match up to modern browsers (with modern defined as IE9+, FF, Chrome) FWIW
I like Microsoft. However, even though I'm a huge fanboy I have to admit they've been a total screw up when it comes to browsers. I cannot understand why they've a history of dragging their feet when it comes to one of the most influential pieces of software on the market these days (the browser). Despite the fact I detest Google, even I can appreciate the fact they made a plug in for IE that makes IE work better. I cannot imagine a better definition for "owned" then that right there. Also, printing from Chrome is a revelation - it's such a nice interface for printing. Print preview in a browser and so on - it's very well done. So Microsoft - just offer Google a couple of billion for Chrome, slap your logo on it and be done. Either get serious about IE or cancel it already. Sheesh.
-
There is a difference between having to support IE7, and having to make it look good in IE7. If I were you I would create an adaptive layout that looks good in modern browsers and works on different sized windows to solve the mobile device vs. desktop problem. For the layout details, though, don't limit yourself to CSS that works in IE7. Just accept that things may look not quite right. It's not too hard to make something that works in IE7, though you'll go crazy trying to make it look exactly the same (unless you want it to look like it's 2005 in ALL browsers). If your organization can accept that using an old browser means their web sites aren't going to look like a modern browser, and may have layout issues, then I'd go with that approach. If they expect everything to look snappy in IE7 too, then you're screwed.
Thanks, well said. > If they expect everything to look snappy in IE7 too, then you're screwed. That's where we were for this release, and what I'm trying to avoid in the future. For anyone that still needs to estimate what it would take a competant designer with a lot of CSS experience to make a site that looks good in Chrome/FF look pixel perfect in IE7 a good estimate is +100-150%. And, yes I mean multiply your IE7 estimate by 2 or 2.5 And, no I'm not the one that did it (I just did the web application).
-
Do feature detection rather than browser detection. Then gracefully downgrade the respective page if the desired feature is not detected.
If your actions inspire others to dream more, learn more, do more and become more, you are a leader.-John Q. Adams
You must accept one of two basic premises: Either we are alone in the universe, or we are not alone in the universe. And either way, the implications are staggering.-Wernher von Braun
Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.-Albert EinsteinSorry I wasn't able to post this earlier - I lost connectivity then had to go out of town (where there's no internet except a borrowed computer I can read email on) We aren't even at feature detection, because we aren't doing some of the cool stuff yet (in public apps). All of this frustration is over 'almost pixel perfect' HTML5 CSS3 sites
-
I like Microsoft. However, even though I'm a huge fanboy I have to admit they've been a total screw up when it comes to browsers. I cannot understand why they've a history of dragging their feet when it comes to one of the most influential pieces of software on the market these days (the browser). Despite the fact I detest Google, even I can appreciate the fact they made a plug in for IE that makes IE work better. I cannot imagine a better definition for "owned" then that right there. Also, printing from Chrome is a revelation - it's such a nice interface for printing. Print preview in a browser and so on - it's very well done. So Microsoft - just offer Google a couple of billion for Chrome, slap your logo on it and be done. Either get serious about IE or cancel it already. Sheesh.
Agree with most of that. I've actually had more trouble printing in Chrome than IE8 or FF. The interface is great - but the results are iffy. Maybe it is because I'm still on XP or something my desktop people have done. I tried to print pages 3-5 of an article Thursday and ended up with 7 pages and only a few words on each page. I'll be on Win7 in a few months and will revisit.