The Enterprise Is Wrong, Not Mozilla

There's been a lot of noise about Mozilla's new rapid release leading to conflict with Enterprise users. Kethinov found an Ars article that points out that "Now that Mozilla has released Firefox 5, version 4, just three months old, is no longer supported. Enterprise customers aren't very pleased with this decision, and are claiming it makes their testing burden impossible. We're not convinced: we think Mozilla's decision is the right one for the Web itself.'"

Comments

Think of it as 4.0.2

[jonescb]

If the version number were 4.0.2 instead of 5.0 Enterprises wouldn't be getting their panties in a bunch over this.

Re:Think of it as 4.0.2

[Lunix Nutcase]

Except it's not just enterprises. Tons of average users are getting headaches over this as well when suddenly an unjustified version jump is making it so their plugins get disabled.

Re:Think of it as 4.0.2

[phantomfive]

Right, as the article points out, the changelist for Firefox 5 is not much more expansive than the changelist for Firefox 3.6.

Some enterprise users have internal apps that they need to test, and some of them are upset about such a 'big' change. In reality they shouldn't be looking at version numbers, they should be looking at a list of potential impacts, to make their testing easier.

If Mozilla wants to handle this PR challenge well, it might help announce that from now on they are going to support they enterprise better (everyone likes to know they are being thought of), then from now on point people to the changelist, or add a 'potential impacts' section to the list. Simple enough, and lets people know they are considered.

Re:Think of it as 4.0.2

[Dunega]

It's not about the version number, it's the "not-supported" part that's the issue.

Re:Think of it as 4.0.2

[Antimatter3009]

But the idea is to speed up the release cycle into what is almost a "rolling release" style. And, in fact, this is exactly what the Ars article is arguing is a good thing (which I agree with). I think if they're going to follow Chrome's release style, though, they need to get the rest of what makes it work for Chrome. By that I mean automatic, almost silent updates and an almost total disregard for the version number. Chrome still has versions, but they don't really mean anything significant. Firefox needs to stop calling this Firefox 5 and start calling it just Firefox. The version is no longer important. Similarly, extension support needs to stop being based on the version number and go to some other system. My initial thought would be to assume all extensions will work and allow the community of users to report broken extensions which can be automatically tallied and turned into a warning of some sort when you install. Think something like: "This extension has been reported to be incompatible with Firefox since dd/mm/yyyy."

Re:Think of it as 4.0.2

[Fantom42]

Right, as the article points out, the changelist for Firefox 5 is not much more expansive than the changelist for Firefox 3.6.

This may be true for this particular instance, but Firefox certainly isn't guaranteeing that going forward. What happens with Firefox 9 is released with a feature that breaks their enterprise, and Firefox 8 is suddenly no longer supported?

This whole attitude I hear parroted that "release numbers are irrelevant because they are just numbers" ignores a whole bunch of realities regarding how new features are introduced and developed to different classes of users. And in the case of Firefox, this new strategy sends a disturbing message to enterprise customers that new and potentially disruptive features will be introduced "when they are ready" and support for previous versions will be immediately dropped.

Re:Think of it as 4.0.2

[Bacon Bits]

Some enterprise users have internal apps that they need to test, and some of them are upset about such a 'big' change. In reality they shouldn't be looking at version numbers, they should be looking at a list of potential impacts, to make their testing easier.

The point is that most every enterprise IT department treats incrementing the major version number more seriously than minor version numbers, and much more seriously than revision numbers. The de facto standard for version numbers is that major version increments mean major changes which require major testing. That's how everybody else -- short of Google Chrome -- operates. Corporate policies are built around these de facto standards. Abandoning them with no justifiable reason is obnoxious and frustrating.

As far as Google Chrome, they've always operated like this. So it's nothing new. They've always had rolling releases with the major version number representing the stable/beta/dev branches more than anything. Additionally, the software is already corporate-unfriendly due to the fact that it allows non-admins to install so nobody in enterprise IT supports it. It's essentially already carrying a sign that says "NOT FOR BUSINESS". It's getting much better (and appears to have better support than Firefox now) but there hasn't been much press around Chrome for the Enterprise. It's just not on anybody's radar like Firefox is (yet).

Re:Think of it as 4.0.2

[Goaway]

Or they could not expose the internals of the application to plugins, and therefore not force them to be upgraded for every browser upgrade.

Re:Think of it as 4.0.2

[bloodhawk]

I can't understand how everyone here is completely missing the point, for the enterprise the version number change is minor. It really doesn't matter, it is the fact of the previous version becoming officially UNSUPPORTED. Firefox is effectively removing themselves from the list of enterprise products as with long testing and release cycles for many enteprises the concept of something being unsupported so fast is completely unacceptable for an enterprise piece of software.

No, they aren't.

[Anonymous Coward]

No, they aren't. EOLing something after 4 months and breaking tons of user plugins for no reason is not good for users or the Web itself. It's needlessly churn to rapidly inflate version numbers for no gain for anyone.

Mozillacide

[denis-The-menace]

We are witnessing "Mozillacide"

Damn "ordinary users", they don't need plugins that work.
Damn the enterprise, they are not the target market.

The version number is now Mozilla's priority.

Re:Mozillacide

[IgnoramusMaximus]

Translation: "Stupid users want to keep their stupid plugins! Ha! What do these idiots know?! We the developers, not some moron users, just know better what users really want! Didn't we tell them that all of their stuff is optional (to us)?! Well, we will just shove our choices down their throats and they will like it! Or else! ... Wait, what happened?! They all left! What's going on?!"

Dear Mozilla

[JamesP]

This is the reason IE continues to stay strong in enterprise.

Yes, corporate users are small-minded, and you're incurring in the same error.

Fix, stabilize, make a 'corporate version'. You don't need many resources for that.

Basically, sell a way for them to use Mozilla.

You're making IT people that root for you look bad. And making the dolts that only know IE look good.

Re:Dear Mozilla

[ElVee]

This move just crushed any chance of Firefox being approved as an 'alternate' browser at the large faceless corporation I toil for, and I'm one of those Firefox 'fanbois' that was pushing for this, and I'm going to look like an idiot now. I'm not real happy with Mozilla right now.

I'm guessing somebody at Mozilla just doesn't understand the size of the testing effort that was underway to get Firefox 4 approved for enterprise distribution. Let me scale it for you: 40,000 workstations in several dozen countries running several hundred 'critical' webapps. We have to certify exactly what works and what doesn't, in each and every language we support for each specific VERSION that is published. Testing is EXPENSIVE, and every dime we put into this effort is now wasted. That leaves Chrome as the only other "corporate-friendly" choice, and now management is going to ask "Well, what if Google decides to pull the same stunt?".

Microsoft gives us written guarantees as to how long they will support previous versions of IE. That means a lot to my corporate overlords. Mozilla might want to consider doing the same thing. It's called "being a reliable business partner".

Soon it may not even matter.

[blind biker]

Firefox's usage share has been slowly declining since quite some time. They introduced the rather universally hated moron-bar, and paid no attention to the feedback. Then they introduced the unwelcome changes in the UI with Firefox 4, and paid no attention to the feedback. Now they decided to piss off the plugin authors and enterprise customers. In the end, they may become a niche browser, and even Google could decide that their money is better spent elsewhere, than on a bunch of idiots.

Re:A release every 6 weeks is really stupid

[the_raptor]

This is exactly my thinking. I don't care about the version numbers, as version systems are entirely arbitrary, but just the drive by Mozilla to subject us to new "features" (like removing established UI elements) constantly.

Browsers are old tech. Browsers are utilitarian. Non-technical people don't want a constantly evolving piece of basic software.

Mainstream browsers are not the place for "cool and cutting edge" development. I want a browser that focuses on security and standards compliance. New features outside that should be addons/plugins until they are so widely adopted, or self-evidently useful, that they get moved into the core of the browser. I call this the Blizzard model because that is the method they follow for World of Warcraft.

Mozilla seem to have adopted We-are-graphic-designers-and-so-know-better-than-you-plebs model that turned "Web 2.0" into a steaming pile of shit.

Enterprise is wrong.

[faedle]

If a web-browser change causes a "mission-critical web app" to break, one of the words in "mission-critical web app" is a lie.

Re:AAT is golden

[lostmongoose]

How about they use better development practices such as not breaking plugins for people by bumping a version number for no reason?

Or how about plugin authors using the Beta or, better yet, the Aurora release to get their shit updated for the final release? God forbid the extension/plugin authors actually do anything to alleviate a problem with a simple solution. No, they'd rather bitch about having to update it instead.

No, Mozilla's delusional

[Tridus]

This is not good for the Enterprise. It's not good for Firefox or Mozilla, which is already losing marketshare and isn't going to benefit from pissing off very large users. It's not even good for "the web" despite their nebulous and poorly supported claim that it is.

In reality this is some blowhards like Asa making poor decisions and then trying to defend them when people point out that it's a poor decision. Normal users don't particularly benefit from more big downloads that break things more often and will sometimes get a new gee-whiz HTML 5 feature out the door a bit sooner (which then won't be adopted by any websites until a couple of versions of FF later because of the lag time required to, you know, develop stuff). Enterprise users clearly suffer because keeping up with this requires throwing testing out the window and will effectively just reinforce the idea that you should stick with IE (where Microsoft actually wants your business and doesn't give you a middle finger).

If driving people away from Firefox is "good for the web", then I guess this is good for the web. But here in reality it's good for IE and Chrome.

Re:A release every 6 weeks is really stupid

[Kjella]

Yes and no. Releases are fine, as long as they add features in a backwards-compatible matter. This is 4.0, all plugin interfaces are stable for the 4.x series. The thing is that with major version numbers, you can't tell because there's nothing bigger. What's the interface for version 5-6-7 going to be like? They could break *everything*, so no plugin is guaranteed compatible. You have to either force them on and pray, or hope the maintainer is on top of the game every few months. Chrome doesn't care because they don't need to care, It also tends to bring a little responsibility to developers if they have to support their bloopers for a while, then you start making sure what you have is really what you want not just a WIP.

Re:What is the purpose of Mozilla?

[mbrinkm]

At Mozilla, all I see is mismanagement. They can't control their code. They can't control their staff. And they are continually lagging behind all competition, which is especially sad given their rock star performance not too long ago, with social buzz propelling a large install base.

I agree with your observations whole heartily and it feels like a giant fuck you to me and I would assume to a lot of people that have been praising and endorsing Firefox for years.

Oh well; on to something else.