I am called Skapare. I've been called Skapare since I played text MUD games online. I do my best to annoy Slashdotters. My phone runs Android. So now I guess everyone knows everything there is to know about me.
Except there is a problem with your line of thinking, and that is how quickly the internet evolves. Firefox 3.6 has webpage rendering bugs which have been fixed in more modern versions. It doesn't support new features that all other recent browsers do. Get with the times, or the internet is going to be a very strange and buggy looking place.
Except there is a problem with your line of thinking, and that is how non-quickly the world moves. Every version of Firefox has webpage rendering bugs which eventually get fixed in latter version. Every version of Firefox didn't support some features that a more recent one did. That has never yet justified rapid update cycles.
Updating software costs time. In business it also costs money. Being behind in software versions is already understood as a cost, too. Theses costs are weighed and balanced against each other. Business has also found and interesting cost bias that influences when software is upgraded. That is the life cycle of hardware that runs the software. There are two such cycles in play. The slower one is when hardware just dies. The somewhat faster one is when hardware is obsoleted by newer faster bigger hardware. And that has resulted in a typical upgrade cycle of 3 to 4 years. It costs less to upgrade software when the upgrade comes in the form of new hardware. And that also often allows running the old and new side by side for a while to work out all the kinks, new bugs, and regressions.
Have you ever worked in an IT department when everyone was upgraded? When I did work in such a department (in a different role... I and 5 others ran the network and servers, leaving 16 to support end users), I found they used a rotation strategy for upgrades to avoid completely overwhelming the staff, which would happen if the entire company of 1000 were upgraded all at once. We divided the company up into segments that upgrades were rotated around. The end result is everyone got a new PC about every 2.5 years, and a new version of Windows with it (along with a larger hard drive, etc). Sometimes some people had exceptional needs and were moved around between groups. I used 2 versions of WIndows, 1 version of Solaris, and half a dozen versions of Linux in the 4 years I was there.
It is lying only to the extent that it gives some people false expectations. The statement that they are lying is only relevant to be expressed to such persons that misunderstand the change. If someone claims that FF has now picked up the pace and is doing 3 to 4 years worth of major innovative development in 2 months, over and over again, of course it is that person whose perspectives need to be changed. So I say TO THAT PERSON that the FF devs "are lying" as part of changing that person's perspective of understanding. Within the incorrect understanding that person has about it, that is the effect. As seen by that person, FF going from version 5 to version 6 to version 7, and so on, within that person's belief that the first digit is always a major change, this must be making FF doing major changes rapidly. It's all about getting that person to change their understanding to realize "oh sh.. these are not major changes".
I stand by my use of it "it is lying"... as a tool to knock some sense into some people.
And we appear to have a number of people who have lost their way as a result of Firefox's altered scheme of numbering. While many software projects do have widely varying version schemes, few projects change the scheme in mid-course. In part I do blame Firefox devs for doing that. And in part I blame the people who think what Firefox has done was a spurt of innovation (when really it is a spurt of marketing waywardness).
Normally, a major change of software makes it time to take a look and see when an upgrade can fit into your plans. But instead we get a bunch of people saying everyone must upgrade NOW because they are 6 or 7 major versions behind.
IMHO, it's time for an independent group to evaluate software and report when releases do constitute a reasonable time to start exploring and scheduling an upgrade. In the case of Firefox, some people have upgraded TO the 3.6 class of browsers a mere year ago. Some even more recently due to the release cycles of distribution adoption, because distributions intent on being more stable will spend some time to evaluate a new project release (both in in-house testing and in watching the information flow about it) before committing it to their next distribution release.
Someone on Firefox 2.0? Sure, I'll say it in time to upgrade real soon now (as in expedite that upgrade scheduling). But for 3.6 I'll say they should just be forwarding looking and begin to examine where an upgrade fits into their plans for the future. An actual upgrade from 3.6 might well be a year or so from now in that case.
Profiles are poorly managed. For example, Firefox won't start the correct profile when visiting a specific site. Stuff gets mixed up when they do because the provide identities are unpredictable. Instead, I did develop (way back in version 0.8) a system to properly manage the browser in multiple instances. And mine does automatically run the appropriate "profile" based on what site is being visited. But it does take longer and longer with each new version to adapt this to the new version. Maybe they are trying to thwart it? Or it could be the fact that they just change everything around. But in any case, it is much work to figure out how they changed it and adapt to it. And that work seems to be increasing. So my upgrading is in larger steps... 0.8... 1.5... 3.6... ???
Profiles do not make for new versions. Versions collide, at least for the common packages, because file names and directory names are the same. Just try installing two different versions of Firefox under Ubuntu. You end up with a mess.
Seriously, that browser is ridiculously outdated and web devs shouldn't (and often don't) have to cater to people who willingly choose to be so far behind standards. Fine, you don't like the new Mozilla, go find another modern web browser you do like. Either way, get over it and move on.
And what version do you think was current just ONE year ago (relative to your post and my reply on 2012-03-04)? Hint: 4.0 had not even come out yet.
Don't be confused by the strange new numbering system the Firefox devs started to use at 4.0 and beyond just because its competitors were using accelerated release numbering. Under the traditional numbering scheme, the current version today (2012-03-04) would be around 4.6.1, making 3.6.27 merely one major version behind.
Or maybe we should be using release numbers based on year-month (like Ubuntu). Then we'd be seeing major release numbers 04.xx, 06.xx, 08.xx, and now 11.xx. Sure, it is time to be moving off of 08.xx. But it is NOT yet time to have expected everyone to complete that move, especially if they just got ON to 08.xx right before 11.xx came out (less than a year ago today).
At the very minimum, upgrades should not be required more often than every 2 years. 3 years is more reasonable. Ubuntu LTS releases are supported for 3 years (5 years for server versions). Slackware has been doing security updates to releases as old as 6 years or more.
It's been less than a year since 4.0 came out. Yeah, all that flurry of new major versions ever few weeks is probably confusing your memory. The latest version today would be roughly equivalent to 4.6.1 had Firefox used the traditional numbering scheme. 3.6.X is NOT old. It just didn't have HTML5 and lots of web developers wanted to use HTML5 in a non-"degrade gracefully" way. So you get lies on major websites that say you need Flash or HTML5 with H.264 to view videos. It's not true because I watched a video just a while ago on my Flash-Free 3.6.17.
Now, if the Firefox developers had set up a file naming scheme for Firefox files that allowed multiple versions to be installed on the system (keep themselves separate by default), and used by users (keep per-user state separate by version, too... only looking to older versions on a read/only basis to initialize a new version state), then it would at least be easier to upgrade (technical term: "graceful upgrade"), while retaining the functionality of old sites that don't upgrade very fast.
My FF 3.6 plays videos all by itself from web sites run by webmasters smart enough to realize this works, and adhered to the altruistic principle of gracefully degrading (and, what an older version can do, let it still do it). It can't render the non-Flash video sites only because those sites put in some gotcha-code to break it and spit out some HTML5 BS. It's the website doing it, not the browser.
You don't need ASM code to end up with a platform specific program. You can do it in C and C++ very easily. And idiots have figured out how to do it in Java and Perl.
Firefox has only had HTML5 for a year. And HTML5 (nor Flash) is NOT required for video (despite the lies from Youtube, Vimeo, and others). People should not be required to upgrade more than every 2 to 3 years. Businesses don't do that until typical funding cycles, which are 3 to 5 years. I know Firefox developers have been on a version frenzy lately. But don't be confused by the numbers. The current Firefox is essentially just the equivalent of 4.6.1 under the old version numbering scheme.
If the developers were true to the principles of real GUI apps, you could just drag the tabs to the bottom and it would just work, even if no developer ever imagined anyone would put the tabs at the bottom. And they would be there when you restart unless you revert to an older version of your preference state.
There are bunches of website out there that don't understand the concept of "degrade gracefully" and bitch and moan about the lack of HTML5 support (came out in FF 4.0 about a year ago), even though I have seen some web sites that do video just fine in HTML4.
If they were smart, they would make ALL plugins run in a separate process, with the option to jail it. There's no reason for Flash to have access to even the files in your home directory. Jail it in its own process with an empty chroot directory. Then even I wouldn't have issues with it. I don't WANT to hate Flash. I just hate the way it gets used. And I don't install it because of that. I would install it if it were run the safe way.
It would have been helpful if the Firefox developers had provided for a means to install, AND run, EVEN concurrently, more than one version of a browser at the same time (without the need for virtual machines or even different user logins). But it seems from their actions that their intention is to have you either upgrade and break current web based apps, or stay behind.
So why is it that every time we end up with twice the memory capacity on new machines, the application (especially GUI application) developers feel the need to double the size of their programs?
Windows 7 now frequently suggests I close FF to free memory, whereas it never did before.
I close FF frequently to free memory. That's because I set it up in a way that each time I start it for a new site I visit, it starts a whole new separate instance that actually stays separate. I end up with many browser instances. But they are NOT collecting tons of orphaned memory from all the sites I visit, because when I leave a site, I quit the browser instead of close the window. Sure, they do free allocated memory when a window closes, but those allocated pieces are all over the place, interleaved with other still allocated pieces, preventing page size units from being discarded back to the kernel.
This is in Linux, not Windows 7.
I do this in 3.6. I don't know if Firefox 10 will work that way.
More likely a video driver issue. That has been stated over the years as one of the many issues with Flash. While I'd agree the driver should not fail, Nvidia just isn't listening.
How often should people be required to upgrade? Keep in mind that upgrading does break things very often, and Firefox is among the leaders of breakage. Upgrading takes time. And with something like Firefox, that's a very critical point to be broken, because you may not be able to access anything until the glitches and other bugs are worked around (which is often slow when answers are not forthcoming on the forums where asked). My last Firefox upgrade took 2 weeks to get it working right.
And of course you will see a lot MORE straggling this year because of Ubuntu's switch to Unity, and people holding back in versions 10.04 or 10.10 to "wait and see" what happens. At least I'm going to try Xubuntu 12.04 on a separate machine and see how well it works. If it works OK and I can get Compiz running under Xfce this time, and the Firefox version on it can be made to work in multi-instances OK, then in about 2 months after that I'll switch my main desktop over to that, and then be on whatever that version of Firefox is.
I now wonder if the question I started this post with will even be answered.
If the developers of Firefox properly understood just how many things BREAK when upgrading a browser, maybe then they would design things to make it easy for two or more versions of Firefox to co-exist (even if there is a requirement that any one user only be using one version at a time and thus require switching user to use a different version). Then, it would at least be easier to migrate gracefully to their new versions.
As it is now, it's a major pain in the arse to upgrade Firefox, usually much worse than upgrading Linux. Because they already do some things wrong, I have to work around that. And that breaks when I upgrade. For example, when I do upgrade, I'll need to build new multi-instance templates and update all the static instances. And that's all because Firefox developers are trying to make it all run under one instance. Very annoying.
And the Firefox people went on a version frenzy, so I just STOPPED upgrading altogether until things settle down. I am most certainly NOT going to upgrade more often than every 2 years.
Slashdot Mirror
"Intellectual property" is a contradiction.
I am called Skapare. I've been called Skapare since I played text MUD games online. I do my best to annoy Slashdotters. My phone runs Android. So now I guess everyone knows everything there is to know about me.
Except there is a problem with your line of thinking, and that is how quickly the internet evolves. Firefox 3.6 has webpage rendering bugs which have been fixed in more modern versions. It doesn't support new features that all other recent browsers do. Get with the times, or the internet is going to be a very strange and buggy looking place.
Except there is a problem with your line of thinking, and that is how non-quickly the world moves. Every version of Firefox has webpage rendering bugs which eventually get fixed in latter version. Every version of Firefox didn't support some features that a more recent one did. That has never yet justified rapid update cycles.
Updating software costs time. In business it also costs money. Being behind in software versions is already understood as a cost, too. Theses costs are weighed and balanced against each other. Business has also found and interesting cost bias that influences when software is upgraded. That is the life cycle of hardware that runs the software. There are two such cycles in play. The slower one is when hardware just dies. The somewhat faster one is when hardware is obsoleted by newer faster bigger hardware. And that has resulted in a typical upgrade cycle of 3 to 4 years. It costs less to upgrade software when the upgrade comes in the form of new hardware. And that also often allows running the old and new side by side for a while to work out all the kinks, new bugs, and regressions.
Have you ever worked in an IT department when everyone was upgraded? When I did work in such a department (in a different role ... I and 5 others ran the network and servers, leaving 16 to support end users), I found they used a rotation strategy for upgrades to avoid completely overwhelming the staff, which would happen if the entire company of 1000 were upgraded all at once. We divided the company up into segments that upgrades were rotated around. The end result is everyone got a new PC about every 2.5 years, and a new version of Windows with it (along with a larger hard drive, etc). Sometimes some people had exceptional needs and were moved around between groups. I used 2 versions of WIndows, 1 version of Solaris, and half a dozen versions of Linux in the 4 years I was there.
It is lying only to the extent that it gives some people false expectations. The statement that they are lying is only relevant to be expressed to such persons that misunderstand the change. If someone claims that FF has now picked up the pace and is doing 3 to 4 years worth of major innovative development in 2 months, over and over again, of course it is that person whose perspectives need to be changed. So I say TO THAT PERSON that the FF devs "are lying" as part of changing that person's perspective of understanding. Within the incorrect understanding that person has about it, that is the effect. As seen by that person, FF going from version 5 to version 6 to version 7, and so on, within that person's belief that the first digit is always a major change, this must be making FF doing major changes rapidly. It's all about getting that person to change their understanding to realize "oh sh.. these are not major changes".
I stand by my use of it "it is lying" ... as a tool to knock some sense into some people.
And we appear to have a number of people who have lost their way as a result of Firefox's altered scheme of numbering. While many software projects do have widely varying version schemes, few projects change the scheme in mid-course. In part I do blame Firefox devs for doing that. And in part I blame the people who think what Firefox has done was a spurt of innovation (when really it is a spurt of marketing waywardness).
Normally, a major change of software makes it time to take a look and see when an upgrade can fit into your plans. But instead we get a bunch of people saying everyone must upgrade NOW because they are 6 or 7 major versions behind.
IMHO, it's time for an independent group to evaluate software and report when releases do constitute a reasonable time to start exploring and scheduling an upgrade. In the case of Firefox, some people have upgraded TO the 3.6 class of browsers a mere year ago. Some even more recently due to the release cycles of distribution adoption, because distributions intent on being more stable will spend some time to evaluate a new project release (both in in-house testing and in watching the information flow about it) before committing it to their next distribution release.
Someone on Firefox 2.0? Sure, I'll say it in time to upgrade real soon now (as in expedite that upgrade scheduling). But for 3.6 I'll say they should just be forwarding looking and begin to examine where an upgrade fits into their plans for the future. An actual upgrade from 3.6 might well be a year or so from now in that case.
Yay! Classic Coke is back!
Profiles are poorly managed. For example, Firefox won't start the correct profile when visiting a specific site. Stuff gets mixed up when they do because the provide identities are unpredictable. Instead, I did develop (way back in version 0.8) a system to properly manage the browser in multiple instances. And mine does automatically run the appropriate "profile" based on what site is being visited. But it does take longer and longer with each new version to adapt this to the new version. Maybe they are trying to thwart it? Or it could be the fact that they just change everything around. But in any case, it is much work to figure out how they changed it and adapt to it. And that work seems to be increasing. So my upgrading is in larger steps ... 0.8 ... 1.5 ... 3.6 ... ???
Profiles do not make for new versions. Versions collide, at least for the common packages, because file names and directory names are the same. Just try installing two different versions of Firefox under Ubuntu. You end up with a mess.
Seriously, that browser is ridiculously outdated and web devs shouldn't (and often don't) have to cater to people who willingly choose to be so far behind standards. Fine, you don't like the new Mozilla, go find another modern web browser you do like. Either way, get over it and move on.
And what version do you think was current just ONE year ago (relative to your post and my reply on 2012-03-04)? Hint: 4.0 had not even come out yet.
Don't be confused by the strange new numbering system the Firefox devs started to use at 4.0 and beyond just because its competitors were using accelerated release numbering. Under the traditional numbering scheme, the current version today (2012-03-04) would be around 4.6.1, making 3.6.27 merely one major version behind.
Or maybe we should be using release numbers based on year-month (like Ubuntu). Then we'd be seeing major release numbers 04.xx, 06.xx, 08.xx, and now 11.xx. Sure, it is time to be moving off of 08.xx. But it is NOT yet time to have expected everyone to complete that move, especially if they just got ON to 08.xx right before 11.xx came out (less than a year ago today).
At the very minimum, upgrades should not be required more often than every 2 years. 3 years is more reasonable. Ubuntu LTS releases are supported for 3 years (5 years for server versions). Slackware has been doing security updates to releases as old as 6 years or more.
Håkon? Is that you?
It's been less than a year since 4.0 came out. Yeah, all that flurry of new major versions ever few weeks is probably confusing your memory. The latest version today would be roughly equivalent to 4.6.1 had Firefox used the traditional numbering scheme. 3.6.X is NOT old. It just didn't have HTML5 and lots of web developers wanted to use HTML5 in a non-"degrade gracefully" way. So you get lies on major websites that say you need Flash or HTML5 with H.264 to view videos. It's not true because I watched a video just a while ago on my Flash-Free 3.6.17.
Now, if the Firefox developers had set up a file naming scheme for Firefox files that allowed multiple versions to be installed on the system (keep themselves separate by default), and used by users (keep per-user state separate by version, too ... only looking to older versions on a read/only basis to initialize a new version state), then it would at least be easier to upgrade (technical term: "graceful upgrade"), while retaining the functionality of old sites that don't upgrade very fast.
My FF 3.6 plays videos all by itself from web sites run by webmasters smart enough to realize this works, and adhered to the altruistic principle of gracefully degrading (and, what an older version can do, let it still do it). It can't render the non-Flash video sites only because those sites put in some gotcha-code to break it and spit out some HTML5 BS. It's the website doing it, not the browser.
So, basically, everyone else was lying about how advanced they were, so Firefox should, too?
You don't need ASM code to end up with a platform specific program. You can do it in C and C++ very easily. And idiots have figured out how to do it in Java and Perl.
There is no technical reason videos and games need Flash. And I don't like making system configuration decisions based on PHB reasoning.
Firefox has only had HTML5 for a year. And HTML5 (nor Flash) is NOT required for video (despite the lies from Youtube, Vimeo, and others). People should not be required to upgrade more than every 2 to 3 years. Businesses don't do that until typical funding cycles, which are 3 to 5 years. I know Firefox developers have been on a version frenzy lately. But don't be confused by the numbers. The current Firefox is essentially just the equivalent of 4.6.1 under the old version numbering scheme.
If the developers were true to the principles of real GUI apps, you could just drag the tabs to the bottom and it would just work, even if no developer ever imagined anyone would put the tabs at the bottom. And they would be there when you restart unless you revert to an older version of your preference state.
There are bunches of website out there that don't understand the concept of "degrade gracefully" and bitch and moan about the lack of HTML5 support (came out in FF 4.0 about a year ago), even though I have seen some web sites that do video just fine in HTML4.
If they were smart, they would make ALL plugins run in a separate process, with the option to jail it. There's no reason for Flash to have access to even the files in your home directory. Jail it in its own process with an empty chroot directory. Then even I wouldn't have issues with it. I don't WANT to hate Flash. I just hate the way it gets used. And I don't install it because of that. I would install it if it were run the safe way.
It would have been helpful if the Firefox developers had provided for a means to install, AND run, EVEN concurrently, more than one version of a browser at the same time (without the need for virtual machines or even different user logins). But it seems from their actions that their intention is to have you either upgrade and break current web based apps, or stay behind.
So why is it that every time we end up with twice the memory capacity on new machines, the application (especially GUI application) developers feel the need to double the size of their programs?
Windows 7 now frequently suggests I close FF to free memory, whereas it never did before.
I close FF frequently to free memory. That's because I set it up in a way that each time I start it for a new site I visit, it starts a whole new separate instance that actually stays separate. I end up with many browser instances. But they are NOT collecting tons of orphaned memory from all the sites I visit, because when I leave a site, I quit the browser instead of close the window. Sure, they do free allocated memory when a window closes, but those allocated pieces are all over the place, interleaved with other still allocated pieces, preventing page size units from being discarded back to the kernel.
This is in Linux, not Windows 7.
I do this in 3.6. I don't know if Firefox 10 will work that way.
More likely a video driver issue. That has been stated over the years as one of the many issues with Flash. While I'd agree the driver should not fail, Nvidia just isn't listening.
I couldn't get Chrome to compile.
I was using FireBird 0.8 until recently. FireFox 3.6 is a lot slower. I'd go back if it wasn't for the case that I can't find a 64-bit version.
How often should people be required to upgrade? Keep in mind that upgrading does break things very often, and Firefox is among the leaders of breakage. Upgrading takes time. And with something like Firefox, that's a very critical point to be broken, because you may not be able to access anything until the glitches and other bugs are worked around (which is often slow when answers are not forthcoming on the forums where asked). My last Firefox upgrade took 2 weeks to get it working right.
And of course you will see a lot MORE straggling this year because of Ubuntu's switch to Unity, and people holding back in versions 10.04 or 10.10 to "wait and see" what happens. At least I'm going to try Xubuntu 12.04 on a separate machine and see how well it works. If it works OK and I can get Compiz running under Xfce this time, and the Firefox version on it can be made to work in multi-instances OK, then in about 2 months after that I'll switch my main desktop over to that, and then be on whatever that version of Firefox is.
I now wonder if the question I started this post with will even be answered.
If the developers of Firefox properly understood just how many things BREAK when upgrading a browser, maybe then they would design things to make it easy for two or more versions of Firefox to co-exist (even if there is a requirement that any one user only be using one version at a time and thus require switching user to use a different version). Then, it would at least be easier to migrate gracefully to their new versions.
As it is now, it's a major pain in the arse to upgrade Firefox, usually much worse than upgrading Linux. Because they already do some things wrong, I have to work around that. And that breaks when I upgrade. For example, when I do upgrade, I'll need to build new multi-instance templates and update all the static instances. And that's all because Firefox developers are trying to make it all run under one instance. Very annoying.
And the Firefox people went on a version frenzy, so I just STOPPED upgrading altogether until things settle down. I am most certainly NOT going to upgrade more often than every 2 years.