22 ms·

Modern Browsers Ship

127 points by jdub 16y ago

simonw 16y ago
Wow, 2007 was a bad year for browser releases.
btipling 16y ago
Is the height supposed to mean anything? Why did they start with Safari? It's the first modern web browser? It points to a page that says IE9 is not a modern browser yet shows IE7 and IE8.
- jdub 16y ago
  No, the icons just indicate releases. They're at different vertical levels so they don't obscure each other. :-) I decided to start in 2006 to catch Firefox 2 and Internet Explorer 7, otherwise you don't fully see the gaping hole which represents the Dark Ages of the modern web. (Safari was added later, and 2.0 happened to slot in at the start of 2006.)
  - phoboslab 16y ago
    I think it would be better to have every browser on it's own line. Especially with IE and Safari as their icons look quite the same from a distance. Edit: My attempt at it. Not as pretty, but a bit more readable: http://www.phoboslab.org/crap/modern-browsers-ship.png http://www.phoboslab.org/crap/modern-browsers-ship.png
    - jdub 16y ago
      Each set of browser icons are at the same vertical height. If it gets more complicated, I might move things around a bit.
    - Timmy_C 16y ago
      In this pic you can really see that IE has HUGE gaps between releases.
  - simonw 16y ago
    It would be pretty cool if you extended it back to 2001, when IE6 was released - then you could also show the rise and fall of Mozilla and the birth of Firefox. Even better, roll it back to 1991! Edit: and another pony request - how about putting the release date in the tooltips (yyyy-mm-dd) to clarify any confusion about when exactly it was?
- jinushaun 16y ago
  IE7/8 are there to show scale. The premise is that modern browsers release frequently. You can see that in the time it took IE to release 2 versions, the other browsers released at least double that same amount. Thus IE is not a modern browser.
  - jdub 16y ago
    The contrast I most hoped to illustrate was Chrome vs. everything else. :-)
- Hovertruck 16y ago
  I don't think the height means anything. As far as starting with Safari, they probably just picked 2006 as the start point and it was the first one with a release that year. The reason it shows IE7 and IE8 is to illustrate that all of the modern browsers are shipping pretty regularly, while IE only has two releases on there.
chaosfox 16y ago
where is opera ?
- jarin 16y ago
  Hanging out on 2.2% of computers
  - rimantas 16y ago
    Opera's popularity is highly geographically dependent, unlike any other browser, I think. It is very popular in Russia and Ukraine for example (almost on par with Firefox). See http://my.opera.com/dstorey/blog/2009/03/16/a-look-at-desktop-market-share-cis-edition http://my.opera.com/dstorey/blog/2009/03/16/a-look-at-deskto... and http://www.opera.com/press/releases/2010/04/14/ http://www.opera.com/press/releases/2010/04/14/
- jdub 16y ago
  Adding it now. :-)
- thomasz 16y ago
  Last time I checked it was crawling around 2.5 % market share.
davej 16y ago
This chart doesn't show the 5 year sabbatical that Microsoft decided to take between August 2001 and October 2006 (The amount of time between releases of IE6 and IE7). I reckon they shipped IE7 at exactly the most damaging time for the web because If they waited another year or so then it might have been long enough to kill off IE altogether.
- rapind 16y ago
  That sabbatical is precisely why I will never ever promote Internet Explorer. It was ridiculous. I don't even care if suddenly MS produces the best technical browser available (highly doubtful, but even so). I don't care if it's a new team because it's the same company, and yes much of the same leadership. They blew it and they don't get another chance. They had the best browser back when IE4 launched. Given the amount of pain IE has caused over the years, it surprises me that any self respecting web developer could ever forgive them. "IE you are dead to me."
  - masklinn 16y ago
    > That sabbatical is precisely why I will never ever promote Internet Explorer. It was ridiculous. Some (including standardistas and other web luminaries) argue it was a good thing overall (though it got long in the tooth around 2005-2006) because it allowed the web to settle and developers to better understand what they were working with (because the target got static, it became possible to explore in depth, not just in breadth). > They had the best browser back when IE4 launched. They also had the best browser back when IE5 launched and back when IE6 launched.
    - EiZei 16y ago
      IE6 the best browser available. Now there's a frightening thought.
      - davej 16y ago
        This is 2001, pre-firefox. If you want a frightening thought then try it's main competitor: Netscape Navigator 4! I remember Opera was my browser of choice around that time, it was a seriously nippy browser back then, way ahead of it's time. Unfortunately the fact that there was no free version prevented them from getting a decent share of the browser market. Later on, they did eventually release a free version but with an ad banner in the navigation bar (of course Opera is now completely free and doesn't have any ad banner).
yread 16y ago
Hmm Opera is missing. 2006 9.1 2008 9.5 2009 10 10.10 2010 10.5 11
- zootm 16y ago
  It's in there now, although it only shows major version numbers (generally not unfair, although a major version of Chrome doesn't necessarily mean the same as a major version of, for example, Firefox). Edit: Actually, I just noticed that Opera 9.1 and 9.5 are there, along with Firefox 3.5, so scratch that!
  - Tyrannosaurs 16y ago
    This is my main issue with it. It shows releases which increment major/minor version number but not all releases (by this definition) are equal. Ultimately all it says is Chrome do lots of smaller releases, IE few bigger ones, the others in the middle.
    - jdub 16y ago
      These really do reflect the most significant releases of each browser, regardless of arbitrary major/minor/micro version number changes... Firefox 3.6 was as major a release in that universe as Chrome 8 was in theirs.
      - bzbarsky 16y ago
        The point, which you seem determined to not get, is that the "major release" designation is an arbitrary one, and some major releases have more scope than others. Which is what Tyrannosaurs said. Then you objected to something he didn't actually say....
        Tyrannosaurs 16y ago
        Absolutely - Chrome 7 wasn't far off a bug fix release (also contained some improvements to HTML5 support but little in the way of new end user features) but it's clearly not the same order as the IE releases which aren't far off ground up rewrites with new UIs and so on.
Qz 16y ago
This 'Modern Browser' stuff is getting kind of old. The question is relatively meaningless, but if it were meaningful, the answer would be more complex than something you can diagram in ~5 minutes.
- jokermatt999 16y ago
  "Modern browser", "open/not open", "killer", "dead", and recently, "Quora" articles can probably be ignored if they're the focus of the title. They're pretty much always petty arguments. I think I'd respect these sort of articles a lot better if they skipped the rhetoric and just focused on their point. "IE is not a modern browser" is a great soundbite, but it also means that's what rebuttal articles will focus on. "IE is holding back the web by not implementing features the majority of other browsers are" is less catchy, but it conveys the actual focus of the article. Sadly, I think most tech journalism would distill an article like that into another meaningless soundbite.
- jdub 16y ago
  Think of this as an answer, not the answer. :-)
podperson 16y ago
One of the things that killed Macromedia Director Schockwave (remember that?) which was in many respects superior to Flash (SWF stands for "shockwave flash" -- Flash originally rode on Director's coat-tails) was too-frequent releases. As a Director coder you never knew what platform you needed to target and the end-user experience usually comprised being pestered to update their plugin before seeing your content. Flash forward to 2010-11 and Google is doing the same thing with Android and Chrome (and Firefox is following suit). This is actually a BAD thing, even though we all have broadband, etc. now. There's some kind of happy medium between significant, reasonably well-paced releases and shoving stuff out the door constantly. "Real artists ship," but they don't spam you with incremental releases.
- jdub 16y ago
  I think that might be true in cases where you're not building on such a widely distributed and incredibly backwards-compatible / gracefully-degrading platform such as the web.
  - nlogn 16y ago
    And Android for that matter. I really don't see podperson's point; what Android builds have broken backwards compatibility?
    - qeorge 16y ago
      IIRC, carriers not updating their customers to the most recent Android build has been a big problem. Not really Google's fault, but perhaps Android isn't a great example.
      - jokermatt999 16y ago
        That causes issues for being able to use the latest features, which is somewhat of a problem, but should not affect backwards compatibility. It means you have to possibly be aware of the OS features you use in your apps, but you don't have to worry about updating them due to some API being deprecated.
defroost 16y ago
The "Grade A" browser ratings are only slightly more meaningful than the "Modern Browser" label. And IE fans (apologists?) can rationalize from now until Sunday, but web developers who project after project are forced to write CSS Hacks, conditional comments, etc. know that improvements to IE have been slow and minimal. The "modern", standards compliant browsers are the ones that continue to improve and push the envelope re: HTML5/CSS3 features. All this talk about IE doing well by not adding HTML5 features because the W3 spec may change is just an excuse for continued suckiness. You'd think the fact that so many people (me not included) use IE give MS a greater pressure to keep pace with the Chromes and Safaris of the world, but I guess they don't agree.
- Joeri 16y ago
  "web developers who project after project are forced to write CSS Hacks, conditional comments, etc. know that improvements to IE have been slow and minimal" Not my experience. IE8 seems to have its act together, at least on CSS 2.1 compliance. Granted, conditional comments are still necessary for IE6 and IE7, but MS has made big improvements.
  - bkhl 16y ago
    Same here. I agree that IE6&7 were bad, but I didn't have to "hack around" for IE8 and IE9. Of course, if you do need to hack around, then you are probably not doing what is considered as "standard". A lot of things that we, early adapters, talk about (e.g. HTML5) aren't standard, yet. This modern browser crap is getting old. I have no problem browsing internet with both IE9, Chrome, and Firefox.
ozten 16y ago
Using this logic, lynx is fairly modern having released within the year: June 2010 lynx 2.8.7 Oct 2006 lynx 2.8.6 Feb 2004 lynx 2.8.5 ... lynx, now with WebGL, WebSockets (disabled) and IndexedDb!
Detrus 16y ago
It's a bit strange comparing Chrome's releases with the others. Firefox had many more pestering minor updates in that time, which are comparable to Chrome's, only Firefox labeled them differently. Safari also had major changes between version numbers. To get an accurate picture the diffs between updates should be compared, Chrome made version numbers meaningless.
- blocke 16y ago
  Firefox's minor updates were not really feature updates. On the chart Chrome whole number releases are new feature releases. Firefox 3.5 and 3.6 are seperate because they are similar feature releases. 3.6.0 to 3.6.1 are not.
geekfactor 16y ago
There's a certain arbitrariness to this analysis, since all that is being considered is an increment in the major version number. If the author were looking at operating systems this way, Apple hasn't released a new operating system in nearly 10 years! (OS 10.0 Cheetah was released on 3/24/01) Not to offend anyone, but Chrome major version releases tend to feel relatively incremental compared to what I had come to expect from Firefox.
- code_duck 16y ago
  I use Chrome every day, and would be unable to list one major difference between Chrome 6 and 7. Or 5 and 6. Or 7 and 8.
  - robryan 16y ago
    The developer tools have been steadily improving, apart from that I can't recall to much exactly. Will be interesting to see if Firefox can keep to their new release schedule after 4. Otherwise it seems that Chrome just iterates to fast compared to Firefox. Chrome 9 is a long way ahead of Firefox 3.6, sure Firefox 4 might even it up but if every release has this kind of beta time Chrome will just keep skipping ahead.
    - code_duck 16y ago
      Hey, there you go. I noticed some great additions to the Inspector. I'm pretty sure that the reason I can't come up with the differences is that the change is so gradual. I know Chrome has been improving, it's just not a dramatic change like FF 1 to 2 or IE 6 to 7. But Chrome 4 to 9 doesn't even seem like a dramatic change. if they were using normal versioning, I don't even know if it would be time for Chrome 2.0 yet.
- InclinedPlane 16y ago
  Why would that be offensive? Chrome major version number changes are a: typically invisible to the user, and b: chosen for merely engineering reasons. If tech blogs didn't exist to tell people about major version changes in chrome most user's wouldn't have the slightest clue what version they were running. tl;dr Chrome's major version releases are supposed to be incremental.
blocke 16y ago
All the nitpicking people are doing on this and ignoring the point behind that graphic... The Point: "Modern Browsers" are the browsers that actually get around to actually @#$@ing SHIPPING new features and NOT TAKING FOREVER TO DO IT. Iterate faster and get new crap to users NOW.
- user24 16y ago
  > get new crap to users NOW thing is, 'new crap' is also unfamiliar crap. Unfamiliar to developers. 5 browsers, 5 teams of developers, results is 5 different ways of doing things. -moz-border-radius and the like are BAD for the web, because everyone shipped new crap without any kind of agreement on how it was supposed to work. And now we've got vendor specific code in our websites. That's not how the web is supposed to work. Impatient browser vendors shipping unfinished specs like css3 are as bad as proprietary spec extensions like 'filter' etc.

127 points by jdub 16y ago

Wow, 2007 was a bad year for browser releases.

Is the height supposed to mean anything? Why did they start with Safari? It's the first modern web browser? It points to a page that says IE9 is not a modern browser yet shows IE7 and IE8.

jdub 16y ago
No, the icons just indicate releases. They're at different vertical levels so they don't obscure each other. :-) I decided to start in 2006 to catch Firefox 2 and Internet Explorer 7, otherwise you don't fully see the gaping hole which represents the Dark Ages of the modern web. (Safari was added later, and 2.0 happened to slot in at the start of 2006.)
- phoboslab 16y ago
  I think it would be better to have every browser on it's own line. Especially with IE and Safari as their icons look quite the same from a distance. Edit: My attempt at it. Not as pretty, but a bit more readable: http://www.phoboslab.org/crap/modern-browsers-ship.png http://www.phoboslab.org/crap/modern-browsers-ship.png
  - jdub 16y ago
    Each set of browser icons are at the same vertical height. If it gets more complicated, I might move things around a bit.
  - Timmy_C 16y ago
    In this pic you can really see that IE has HUGE gaps between releases.
- simonw 16y ago
  It would be pretty cool if you extended it back to 2001, when IE6 was released - then you could also show the rise and fall of Mozilla and the birth of Firefox. Even better, roll it back to 1991! Edit: and another pony request - how about putting the release date in the tooltips (yyyy-mm-dd) to clarify any confusion about when exactly it was?
jinushaun 16y ago
IE7/8 are there to show scale. The premise is that modern browsers release frequently. You can see that in the time it took IE to release 2 versions, the other browsers released at least double that same amount. Thus IE is not a modern browser.
- jdub 16y ago
  The contrast I most hoped to illustrate was Chrome vs. everything else. :-)
Hovertruck 16y ago
I don't think the height means anything. As far as starting with Safari, they probably just picked 2006 as the start point and it was the first one with a release that year. The reason it shows IE7 and IE8 is to illustrate that all of the modern browsers are shipping pretty regularly, while IE only has two releases on there.

chaosfox 16y ago

where is opera ?

jarin 16y ago
Hanging out on 2.2% of computers
- rimantas 16y ago
  Opera's popularity is highly geographically dependent, unlike any other browser, I think. It is very popular in Russia and Ukraine for example (almost on par with Firefox). See http://my.opera.com/dstorey/blog/2009/03/16/a-look-at-desktop-market-share-cis-edition http://my.opera.com/dstorey/blog/2009/03/16/a-look-at-deskto... and http://www.opera.com/press/releases/2010/04/14/ http://www.opera.com/press/releases/2010/04/14/
jdub 16y ago
Adding it now. :-)
thomasz 16y ago
Last time I checked it was crawling around 2.5 % market share.

davej 16y ago

This chart doesn't show the 5 year sabbatical that Microsoft decided to take between August 2001 and October 2006 (The amount of time between releases of IE6 and IE7). I reckon they shipped IE7 at exactly the most damaging time for the web because If they waited another year or so then it might have been long enough to kill off IE altogether.

rapind 16y ago
That sabbatical is precisely why I will never ever promote Internet Explorer. It was ridiculous. I don't even care if suddenly MS produces the best technical browser available (highly doubtful, but even so). I don't care if it's a new team because it's the same company, and yes much of the same leadership. They blew it and they don't get another chance. They had the best browser back when IE4 launched. Given the amount of pain IE has caused over the years, it surprises me that any self respecting web developer could ever forgive them. "IE you are dead to me."
- masklinn 16y ago
  > That sabbatical is precisely why I will never ever promote Internet Explorer. It was ridiculous. Some (including standardistas and other web luminaries) argue it was a good thing overall (though it got long in the tooth around 2005-2006) because it allowed the web to settle and developers to better understand what they were working with (because the target got static, it became possible to explore in depth, not just in breadth). > They had the best browser back when IE4 launched. They also had the best browser back when IE5 launched and back when IE6 launched.
  - EiZei 16y ago
    IE6 the best browser available. Now there's a frightening thought.
    - davej 16y ago
      This is 2001, pre-firefox. If you want a frightening thought then try it's main competitor: Netscape Navigator 4! I remember Opera was my browser of choice around that time, it was a seriously nippy browser back then, way ahead of it's time. Unfortunately the fact that there was no free version prevented them from getting a decent share of the browser market. Later on, they did eventually release a free version but with an ad banner in the navigation bar (of course Opera is now completely free and doesn't have any ad banner).

yread 16y ago

Hmm Opera is missing. 2006 9.1 2008 9.5 2009 10 10.10 2010 10.5 11

zootm 16y ago
It's in there now, although it only shows major version numbers (generally not unfair, although a major version of Chrome doesn't necessarily mean the same as a major version of, for example, Firefox). Edit: Actually, I just noticed that Opera 9.1 and 9.5 are there, along with Firefox 3.5, so scratch that!
- Tyrannosaurs 16y ago
  This is my main issue with it. It shows releases which increment major/minor version number but not all releases (by this definition) are equal. Ultimately all it says is Chrome do lots of smaller releases, IE few bigger ones, the others in the middle.
  - jdub 16y ago
    These really do reflect the most significant releases of each browser, regardless of arbitrary major/minor/micro version number changes... Firefox 3.6 was as major a release in that universe as Chrome 8 was in theirs.
    - bzbarsky 16y ago
      The point, which you seem determined to not get, is that the "major release" designation is an arbitrary one, and some major releases have more scope than others. Which is what Tyrannosaurs said. Then you objected to something he didn't actually say....
      - Tyrannosaurs 16y ago
        Absolutely - Chrome 7 wasn't far off a bug fix release (also contained some improvements to HTML5 support but little in the way of new end user features) but it's clearly not the same order as the IE releases which aren't far off ground up rewrites with new UIs and so on.

Qz 16y ago

This 'Modern Browser' stuff is getting kind of old. The question is relatively meaningless, but if it were meaningful, the answer would be more complex than something you can diagram in ~5 minutes.

jokermatt999 16y ago
"Modern browser", "open/not open", "killer", "dead", and recently, "Quora" articles can probably be ignored if they're the focus of the title. They're pretty much always petty arguments. I think I'd respect these sort of articles a lot better if they skipped the rhetoric and just focused on their point. "IE is not a modern browser" is a great soundbite, but it also means that's what rebuttal articles will focus on. "IE is holding back the web by not implementing features the majority of other browsers are" is less catchy, but it conveys the actual focus of the article. Sadly, I think most tech journalism would distill an article like that into another meaningless soundbite.
jdub 16y ago
Think of this as an answer, not the answer. :-)

podperson 16y ago

One of the things that killed Macromedia Director Schockwave (remember that?) which was in many respects superior to Flash (SWF stands for "shockwave flash" -- Flash originally rode on Director's coat-tails) was too-frequent releases. As a Director coder you never knew what platform you needed to target and the end-user experience usually comprised being pestered to update their plugin before seeing your content. Flash forward to 2010-11 and Google is doing the same thing with Android and Chrome (and Firefox is following suit). This is actually a BAD thing, even though we all have broadband, etc. now. There's some kind of happy medium between significant, reasonably well-paced releases and shoving stuff out the door constantly. "Real artists ship," but they don't spam you with incremental releases.

jdub 16y ago
I think that might be true in cases where you're not building on such a widely distributed and incredibly backwards-compatible / gracefully-degrading platform such as the web.
- nlogn 16y ago
  And Android for that matter. I really don't see podperson's point; what Android builds have broken backwards compatibility?
  - qeorge 16y ago
    IIRC, carriers not updating their customers to the most recent Android build has been a big problem. Not really Google's fault, but perhaps Android isn't a great example.
    - jokermatt999 16y ago
      That causes issues for being able to use the latest features, which is somewhat of a problem, but should not affect backwards compatibility. It means you have to possibly be aware of the OS features you use in your apps, but you don't have to worry about updating them due to some API being deprecated.

defroost 16y ago

The "Grade A" browser ratings are only slightly more meaningful than the "Modern Browser" label. And IE fans (apologists?) can rationalize from now until Sunday, but web developers who project after project are forced to write CSS Hacks, conditional comments, etc. know that improvements to IE have been slow and minimal. The "modern", standards compliant browsers are the ones that continue to improve and push the envelope re: HTML5/CSS3 features. All this talk about IE doing well by not adding HTML5 features because the W3 spec may change is just an excuse for continued suckiness. You'd think the fact that so many people (me not included) use IE give MS a greater pressure to keep pace with the Chromes and Safaris of the world, but I guess they don't agree.

Joeri 16y ago
"web developers who project after project are forced to write CSS Hacks, conditional comments, etc. know that improvements to IE have been slow and minimal" Not my experience. IE8 seems to have its act together, at least on CSS 2.1 compliance. Granted, conditional comments are still necessary for IE6 and IE7, but MS has made big improvements.
- bkhl 16y ago
  Same here. I agree that IE6&7 were bad, but I didn't have to "hack around" for IE8 and IE9. Of course, if you do need to hack around, then you are probably not doing what is considered as "standard". A lot of things that we, early adapters, talk about (e.g. HTML5) aren't standard, yet. This modern browser crap is getting old. I have no problem browsing internet with both IE9, Chrome, and Firefox.

ozten 16y ago

Using this logic, lynx is fairly modern having released within the year: June 2010 lynx 2.8.7 Oct 2006 lynx 2.8.6 Feb 2004 lynx 2.8.5 ... lynx, now with WebGL, WebSockets (disabled) and IndexedDb!

Detrus 16y ago

It's a bit strange comparing Chrome's releases with the others. Firefox had many more pestering minor updates in that time, which are comparable to Chrome's, only Firefox labeled them differently. Safari also had major changes between version numbers. To get an accurate picture the diffs between updates should be compared, Chrome made version numbers meaningless.

blocke 16y ago
Firefox's minor updates were not really feature updates. On the chart Chrome whole number releases are new feature releases. Firefox 3.5 and 3.6 are seperate because they are similar feature releases. 3.6.0 to 3.6.1 are not.

geekfactor 16y ago

There's a certain arbitrariness to this analysis, since all that is being considered is an increment in the major version number. If the author were looking at operating systems this way, Apple hasn't released a new operating system in nearly 10 years! (OS 10.0 Cheetah was released on 3/24/01) Not to offend anyone, but Chrome major version releases tend to feel relatively incremental compared to what I had come to expect from Firefox.

code_duck 16y ago
I use Chrome every day, and would be unable to list one major difference between Chrome 6 and 7. Or 5 and 6. Or 7 and 8.
- robryan 16y ago
  The developer tools have been steadily improving, apart from that I can't recall to much exactly. Will be interesting to see if Firefox can keep to their new release schedule after 4. Otherwise it seems that Chrome just iterates to fast compared to Firefox. Chrome 9 is a long way ahead of Firefox 3.6, sure Firefox 4 might even it up but if every release has this kind of beta time Chrome will just keep skipping ahead.
  - code_duck 16y ago
    Hey, there you go. I noticed some great additions to the Inspector. I'm pretty sure that the reason I can't come up with the differences is that the change is so gradual. I know Chrome has been improving, it's just not a dramatic change like FF 1 to 2 or IE 6 to 7. But Chrome 4 to 9 doesn't even seem like a dramatic change. if they were using normal versioning, I don't even know if it would be time for Chrome 2.0 yet.
InclinedPlane 16y ago
Why would that be offensive? Chrome major version number changes are a: typically invisible to the user, and b: chosen for merely engineering reasons. If tech blogs didn't exist to tell people about major version changes in chrome most user's wouldn't have the slightest clue what version they were running. tl;dr Chrome's major version releases are supposed to be incremental.

blocke 16y ago

All the nitpicking people are doing on this and ignoring the point behind that graphic... The Point: "Modern Browsers" are the browsers that actually get around to actually @#$@ing SHIPPING new features and NOT TAKING FOREVER TO DO IT. Iterate faster and get new crap to users NOW.

user24 16y ago
> get new crap to users NOW thing is, 'new crap' is also unfamiliar crap. Unfamiliar to developers. 5 browsers, 5 teams of developers, results is 5 different ways of doing things. -moz-border-radius and the like are BAD for the web, because everyone shipped new crap without any kind of agreement on how it was supposed to work. And now we've got vendor specific code in our websites. That's not how the web is supposed to work. Impatient browser vendors shipping unfinished specs like css3 are as bad as proprietary spec extensions like 'filter' etc.