Jono recently wrote a blog post about Firefox updates, and Atul wrote a follow up. They are two of the brightest usability thinkers I know. When they talk about users, I listen. I listen, even though some of the things they say sound confused to me, and some are plain wrong. And I listen because if people as bright and in tune with Mozilla as them think these things, I bet others do, too.
When I read (and re-read) the posts, I see 3 main points:
- The constant interruption of updates is toxic to the usability of any piece of software, especially one as important as your web browser.
- Our reasons for frequent updates were arbitrary, and based on the wrong priorities.
- We take our users for granted.
To be honest, if it weren’t for the third point, I wouldn’t be writing this. Anytime you do something that impacts lots of people, especially lots of impassioned and vocal people, you’re gonna get criticism. Listening to that is essential, but fanning the flames can consume all your energy and even still some people won’t be convinced. The third point, though, made by two people who know and love Mozilla even if they haven’t been close to the release process, isn’t something I want to leave sitting. I understand how it can fit a narrative, but it’s just not true.
Since I’m writing anyhow, though, let’s take them in order.
Interruptions Suck
Yes. They do. One criticism that I think we should openly accept is that the move to regular releases was irritating. The first releases on the new schedule had noisy prompts (both ours and the operating systems’). They broke extensions. Motives aside, our early execution was lacking and we heard about it. Plenty.
Today our updates are quiet. Addons have been compatible by default since Firefox 10 back in January. But that was a mountain of work that would have been much nicer to have in hand up front. As Jono says, hindsight is 20/20, but we should have done better with foresight there.
Motivations
It was hard for me to read the misapprehension of motives in these posts. Hard because I think Mozilla’s earned more credit than that, and hard because it means I haven’t done a good job articulating them.
Let me be clear here because I’m one of the guys who actually sits in these conversations: when we get together to talk about a change like this, concepts like “gotta chase the other guys” are nowhere in the conversation. When we get together and draw on whiteboards, and pound on the table, and push each other to be better, it is for one unifying purpose: to do right by our users and the web.
I wrote about this a while back, but it bears repeating. We can’t afford to wait a year between releases ever again; we can’t afford to wait 6 months. Think how much the web changes in a year, how different your experience is. Firefox 4 was 14 months in the making. A Firefox that updates once every 14 months is not moving at the speed of the web; we can’t go back there. Every Firefox release contains security, compatibility, technology and usability improvements; they should not sit on the shelf.
There’s nothing inviolate about a 6 week cycle, but it’s not arbitrary either. It is motivated directly from our earnest belief that it is the best way for us to serve our users, and the web at large.
And so the hardest thing for me to read was the suggestion that…
We Take Our Users For Granted
Nonsense. I don’t know how else to say it. In a very literal way, it just doesn’t make sense for a non-profit organization devoted to user choice and empowerment on the web to take users for granted. The impact of these changes on our users was a topic of daily conversation (and indeed, clearly, remains one).
To watch a Mozilla conversation unfold, in newsgroups or in blogs, in bugzilla or in a pub, is an inspiring thing because of how passionately everyone, on every side of an issue, is speaking in terms of the people of the web and how we can do right by them. We are at our most excellent then.
There’s beauty in the fact that this is another of those conversations. It is not lost on me, nor on Jono and Atul, I’d wager. They are Mozillians. And I believe they care deeply about Firefox users. I hope they realize how much the rest of us do, too.
meandering wildly 
For a bit I tried to use Chrome, around late 2009, and the thing that bugged me the most was the automatic update. The worst is when I realized it had done it behind my back without me running using it: I grew tired of it and returned to Firefox. But month later, Chrome was still up to date: it had installed something I would call “malware”: a piece of software that change other software on my machine without telling me.
I think we really have to be prudent about the silent update, as power users, those that people turn to for recommendations, notice that kind of things.
(Disclaimer: I work for MoCo. And I still prefer Firefox to the other browsers)
From the perspective of most of the users I know, projects like memshrink, the (mostly stalled) papercuts and silent update should be top priorities. Now it is said with hindsight, but it’s not like there weren’t concerns raised before Firefox 4 was released. And those concerns were sometimes ignored and sometimes acknowledged, but if there was a response, it was generally to repeatedly articulate why Mozilla believed what it was doing was right. The question is whether it would have been better to change the plan.
I can accept that it is your earnest belief that the 6 week cycle is the best way of serving users. The problem is that many users earnestly believe that it is not the best way of serving them.
I know 2 or 3 people who have computers that they only use once or twice each month – sometimes not for a couple of months at a time. Every time they start the computer up, they have to fight with a bunch of updates from various bits of software. Then they use the computer for an hour or two, and then shut it down for another long period. Is that update you’re bothering them with really improving their experience in that hour or two?
I think it’s a fact that every release has come with important security fixes. To say that every release comes with “compatibility, technology and usability improvements” is a stretch. Take the release notes for (desktop) Firefox 12 – I don’t see a whole lot of significance in there. I’m sure if you look through the whole pushlog it would be possible to find improvements in each category that would have at least a small effect on a small proportion of users. But many people would probably want to trade the hassle of an update for not having those improvements for another 6 weeks (and security updates don’t give a noticable improvement to the experience, they just avoid a potentially extremely negative one…).
I’m not sure the assertion was supposed to be that Mozilla was intentionally taking users for granted (I wouldn’t agree with that). But having an earnest belief that you are doing the right thing for someone does not necessarily mean that you are right. Spending more time articulating why you believe what you believe may or may not help. Maybe what is (or, at least, was) needed is to pay more attention to the current experiences of current users, and less to delivering things that Mozilla believes will be best for the web in the future.
Thinking about how much the web will change in the next year does not mean that problems that happened last year (and the year before, and the year before that) do not also need attention. However, I can see that changing priorities to be short term can mean falling behind in the longer term – it’s a balance.
Mozilla has lost users and potential users never give it a chance. Mozilla was never going to be able to compete with Chrome is terms of advertising as Chrome is force-fed down users throats who opt to use the browser from a non-Google browser. On the various tech sites every time Firefox is mentioned, you have users screaming about it “looking and feeling like a Chrome clone” and yet, when you mention this to Mozilla staff, they act as though it’d be ludicrous for someone to think such a thing.
Word-of-mouth is Firefox’s friend and the only way to compete with Chrome’s advertising is the more personal word-of-mouth recommendation. However if we’re unable to convince the people who shout the loudest about browsers, we’re walking down a dark path and that’s sad.
Firefox isn’t as bad as was made out in Jono’s post. But it is in fact the perception of the layman and yet, nothing is being done to challenge that. Instead resources were more thinly spread rather than concentrated on rectifying the pains of the update procedure and the misconceptions regarding add-ons not working.
There’s a followup here http://evilbrainjono.net/blog?permalink=1097 which I think clarifies a lot.
Personally, I have to say the rapid release has been nothing but a massive improvement for me. I’m getting the effects of projects like memshrink within reaasonable timescale.
I understand where others are coming from but as an unusual ‘user’ I believe in updates, because software will always need improvement, so setting aside some time each week to do all my updates is completely worth it.
And Mozilla has definitely fixed the problems mentioned above with updating, so definitely full speed ahead!
There’s one urban myth that bugs me as long as we’re talking about this.
That Firefox 2, 3, 3.5 or 3.6 wouldn’t have updated every 6 weeks, and that there weren’t prompts about those.
As far as I can think back, we’ve always sent out security updates, on a rather frequent cadence. They weren’t silent either.
I guess we shot ourselves in the foot by having frequent addon-busting updates, ’cause that’s pretending that we’re doing “arbitrarily large” changes every six weeks. Which is totally the opposite of what we want, of course. We want small incremental change on cadence, nothing scary.
I’d expect that we need about the same amount of “non-breaking” updates as we had “halloween” ones to make people feel safer about our progress again. We also may need to be looking into the future a tad more. My mom got scared by the new-tab page, which showed her bank screens. If we’d learn where change is scary for good reasons, we’ll win.
And I think that’s the thing. Firefox isn’t necessarily bad now, but they badly stuffed up the transition to the rapid-release cycle – and did so in ways that people were predicting in advance, not just in hindsight.
Now, you might see that as a thing of the past, but it drove away a lot of users, and those users are still talking about the bad experiences they had with Firefox and discouraging others from using it. And that’s a big problem.
The assertion that Mozilla can’t possibly be taking its users for granted concerns me a bit.
Anyone familiar with Mozilla knows how unlikely it is that they are consciously taking users for granted, but that doesn’t stop people from making well-intentioned decisions that nevertheless have a similar effect.
Mozilla cares, but caring alone isn’t enough. We need to be constantly aware of whether our good intentions are bearing fruit or rot.
Disclaimer: I like the rapid releases.
Jono does have a point and the responses on Hacker News and Reddit proves it. Many people dislike the change (and some are just Chrome fanboys fighting the holy browser war).
Axel Hecht is dead on. Firefox 3.6 got 16 point releases in 14 months – not quite once every 6 weeks, but not far from it. The major “culprit” here, I’d say is the adoption of Windows Vista/7. Before UAC, updates were a lot less obtrusive. And when Firefox 4 was released only about a third of the world’s internet users had Vista/7.
Anyhow, what’s done is done. Let’s just move on. But with the knowledge of 1+ years of rapid releases.
Jono’s followup post (see Awad’s post above) is food for thought. “Nobody” would complain about security, speed, stability, memory and web enhancements. UI and API changes should perhaps be released all at once when enough of them have been developed? This would a) lessen the problem with relearning, 2) lessen the problem with breaking add-ons (binary included?). And have the added benefit of making certain releases newsworthy, just like in the old days!
@Nuss: A big UI change instead of many smaller ones would “lessen the problem with relearning”? Are you sure about that?
Whiners gonna whine. I don’t mind the updates, I love how Firefox handles them compared with say Adobe or Oracle, I appreciate the improvements. Onward!
I feel that the rapid updates were more a knee-jerk reaction to just how horribly long Firefox 4 took to release; at the time, at least, Mozilla seemed simply incapable of shipping anything on a schedule. In response, bits now get pushed every six weeks.
I suspect though, that at least unconsciously, Chrome was also a big motivation; after all, the primary justification for thinking it was going to work was that Chrome was already doing it. If that hadn’t been in the discussion, perhaps there would have been more thought given to the user impact – but then, perhaps not.
Axel mentioned the urban myth about the frequency of releases prior to FF4.
Another urban myth is the one about the UI. “The UI changes every 6 weeks!” Um, no it doesn’t.
“Another urban myth is the one about the UI. “The UI changes every 6 weeks!†Um, no it doesn’t.”
You’re wrong. Just ask in the Theme developer forums. Many third party themers have given up updating their themes because they just can’t keep up with the continuous stream of changes.
Phil
“To watch a Mozilla conversation unfold, in newsgroups or in blogs, in bugzilla or in a pub, is an inspiring thing because of how passionately everyone, on every side of an issue, is speaking in terms of the people of the web and how we can do right by them. We are at our most excellent then.”
Unfortunately to most outsiders this tends to come across as Mozilla saying “We know what the user wants better then the users themselves”. Perhaps you should spend more time listening to actual users in the user newsgroups and forums instead of basing your decisions on theoretical end user personas/profiles.
Phil
“@Nuss: A big UI change instead of many smaller ones would “lessen the problem with relearningâ€? Are you sure about that?”
Yes, It takes time to adjust to a UI change, while adapting your muscle memory, productivity suffers. Doing this once a year in one go is less cumulatively traumatic than having to do that every six weeks.
Phil
As the comments here show the people here are thinking very different about the mentioned points.
“The constant interruption of updates is toxic to the usability of any piece of software, especially one as important as your web browser.”
Personally I prefer being notified about new updates of a software because I like seeing software being constantly improved. And I am also one of the persons wanting to know about what has changed.
“Our reasons for frequent updates were arbitrary, and based on the wrong priorities.”
As stated by the other commenters a rapid release cycle has positive and negative effects.
There is a clear benefit of having regular updates by getting fixed bugs to the people as fast as possible. Though as Axel mentioned earlier this was already working before Mozilla switched to the new release cycle.
Personally e.g. I still just don’t like the new version numbering principal. It leads somebody to believe that each release brings bigger changes than it actually does. One of the reasons for this surely was to catch up with Chrome’s version numbers.
Anyway, the fast release cycle works and disregarding the reasons behind it with each release users and developers get fixes and enhancements.
“We take our users for granted.”
From my experience this is depending on what you’re asking for. Some(times) Mozilla guys are caring about their users and sometimes not. That’s how people are.
E.g. I have several issues nobody at Mozilla cares about even when they are affecting a big bunch of people, which really bugs me a lot. On the other side some issues are handled within days or even hours.
Also I read many threads of people asking for something useful and it’s rejected by Mozilla or done in an unsatisfying way.
So you could indeed get the impression that sometimes Mozilla doesn’t care about their users, but you can’t generalize this statement.
Sebastian
It’s NOT about HOW OFTEN the updates happen. It’s how they’re done. The little details matter.
We could do better. More listening to users. More ears to the ground. More research and experiments.
Keep up the good work!
Reading through Jono’s and Atul’s posts, it strikes me that the issues they raised are essentially the issues that Asa Dotzler rather curtly dismissed a year ago with the infamous comment that he didn’t “care about making Firefox enterprise-friendly.” My hunch at the time was that many consumers would have exactly the same concerns as the enterprise users (even if they might not articulate them in the same terms that enterprise IT staff would use), and that Asa’s apparent contempt for those concerns translated to contempt for a large number of Firefox users. While Mozilla has worked hard to mitigate the damage since then, there is a lot of truth to the old saying: “It takes 20 years to build a reputation and five minutes to ruin it.” It will take some time for Firefox to recover its reputation.
Hi Jonath,
I am deeply, deeply sorry that my post caused PR damage for Mozilla. That was never my intent. I never imagined my post would go viral, nor did I imagine how much the Internet would twist the point of what I wrote.
The main point I wanted to make was about the distance between the developer perspective and the user perspective, the costs for users of updates (even good updates), and the reasons why developers (everywhere, not just Mozilla) may have trouble seeing updates from the user perspective. I stand by that point. I just wish I had chosen my words more carefully.
I was naiive to think that my post would stay confined to the small group of people who read my personal blog. It seems like a lot of the news sites reblogging my post thought that “Firefox dev Jono DiCarlo says Firefox suxxx” would make an irresistable headline — never mind that I’m not a Firefox dev, my name isn’t DiCarlo, and I don’t think Firefox suxxx!
I was heartbroken when I saw that you had to write a press release to respond to all the flak that my post was generating. Thank you for writing this very thoughtful response, and thank you for giving me the benefit of the doubt in my motives. I’m just sorry that you had to write it at all.
Please accept my personal apology.
TL;DR – I didn’t mean for this to happen, I’m just bad at the Internet.
P.S. I do think Mozilla has been doing a lot to address the problems with the update system, and the updates since version 13 or so have actually been quite good.
(I mentioned that in my post and in the follow-up, but it was at the end and I think everybody missed it.)
I’d like to help get the word out to as many people as possible that if they have given up on Firefox due to update problems, it’s come a long way and they should give it another try.
It is sad to see that you do not understand the problem.
You don’t take your users for granted.
Sorry, but yes you do. Over the past couple years the Firefox developers have made a number of very, very bad design decisions and every complaint from users was met with the same response. “This is the way it is. it’s not going back to the way it was so just shut up and get used to it”. Jono said it very well in his now controversial blog post:
“Users who pointed out flaws in this plan were dismissed as being just a loud minority, or just irrationally afraid of change, or just focused on some marginal edge-cases which would be easy to fix once we got the hang of the new release cycle.”
“Firefox 4 was 14 months in the making. A Firefox that updates once every 14 months is not moving at the speed of the web; we can’t go back there.”
Firefox wasn’t updated once every 14 months. it was updated regularly. There was 3.6.1, then 3.6.2, then 3.6.3, etc., etc., etc. What you have done, starting with Firefox 4 is NOT “Rapid Release”. What you have done is “Rapid Version Number Inflation”. What you call Firefox 13 should really be called Firefox 4.1.8 or something. Maybe Firefox 5 at the most. But that wouldn’t look good when all your competitors have much higher version numbers and thereby make it appear that Firefox is lagging behind.
Are we really supposed to believe that in the past it took well over a year, sometimes two, to go from one major release to another, but now you can do it in 6 weeks? Utter rubbish and nonsense. What you are doing now is making the same small incremental improvements that you’ve always made in the past but now call them major releases, with an accompanying change in the major version number.
“The constant interruption of updates is toxic to the usability of any piece of software.”
The problem is not the update mechanism. Jono was wrong when he claimed that Chrome got it right and designed a good update mechanism right from the start and then built the browser around that. Firefox’s update mechanism is fine. It’s the updates themself that are the problem. Time after time features are removed or changed in a way that makes them less useful. People complain but are ignored (see point 1). This results in people being forced to use extensions to get back features they have lost. And then those extensions break when Firefox is updated.
“Our reasons for frequent updates were arbitrary, and based on the wrong priorities.”
This is absolutely true., as I’ve already explained in my previous points.
In conclusion, it’s very sad. Firefox started out as a fantastic product. And in many ways it’s still better than all the other browsers. But you keep making it harder and harder to like Firefox.
@Gern Blaanston: I actually think that it is people like you, who spread uninformed information, who are making it harder and harder to like Firefox.
Whether Firefox 13 should be called 4.1.8 or 5 or 13 or 2012.3 is completely irrelevant. It’s practically invisible to users. There is no canonical version numbering scheme. They don’t mean anything except that a larger number is released later.
You really don’t prove any point (except perhaps in 3) in your point. No examples, just anecdotal evidence. If you want to convince anybody that rapid releases are bad, bring up some hard facts.
I much liked to read the criticism and I dislike the tone of defense that still is between the lines in your article. I can understand it, but it’s not the point this is about: One of Firefox strengths were the users promoting it, the hype, the movement. Last year was like a hit, like that somebody has ignited a nuke in the middle of us. That were this insane update brouhaha. I think this hit has hurt over the day.
I still use Firefox because it ships with my OS. And it is still a great browser.
But with everything bad there is something good if lessons are learned. If there is something Firefox will be able to compete with Chrome are those users. It’s more than only listen to users, the development should be user-driven. Add-Ons were an important corner-stone in that process. These updates last year were basically an Add-On slaughtering, a killing spree with your users. The software has taken over it’s users: Update or I’ll kill you.
Silent updates alone might not solve this. You can still break stuff – now even silently.
Now you tell something about lessons learned and I see first-hand technical feature descriptions as strategy of defense. But you can not solve communication problems with a technical processing. The problem is buried much deeper. Mozilla need to analyse and explain by the best sense of the word why such massive wrong decisions have been made last year. I bet the reasons are non-technical.
I am tired to listen to technical explanations. That’s how you can talk with developers if you try to convince them for something. But I’m a user, too.
One of the best benefits we have is that the only thing we need to care about are our users. There is no other agenda. No company agenda or whatever. Let’s become free again, I’d like to see more free thinking in dealing with all these problems.
Mozilla is not Google. There is no company agenda that needs to create crippled smartphone or minimal desktop devices that only need to run a browser that does everything. That’s even a very short-term strategy taking into account how fast those devices change.
Why do you aim at those devices? Who needs that crap? Let the commercial development sector feed those devices and can’t we just get a working browser that is safe, flexible, takes care of privacy, takes care about web-standards and has long term support for websites of day 0 and beyond?
Thank you. Yes you’re getting better again. But part of it would be if you find out why you did such a wrong decision first-hand.
How about you start to pick the oldest open bugs and think about why you were not able to fix them since ages? As a developer I much love these bug databases. Those issues all tell a lesson, and the older bugs tell the best lessons. There are things we don’t want to do as developers, there are things we are afraid to touch and all that stuff. There are even just things that went out of sight.
Can you stand to fix the ten oldest open tickets within each of your six week release cycles. Just the ten oldest. Just fixed. Per each release. Can you stand it?
I am a well seasoned IT professional. I too have expressed my displeasure of not so much the constant stream of updates but the seeming CONSTANT changes to the UI. Changes that appear to happen unilaterally without any room for user feedback. These changes seem to take perfectly functioning features and substitute newer and less well behaved AND feature rich replacements OR just remove features that I and others have come to rely on. Things like the URL of a link showing the “real” location at the bottom of the FF widget.
All this seems to happen without ANY regard to the users. I’m a Nightly fan and love getting the newest things and reporting bugs too. Things like the replacement of the “downloads” popup vs the one Nightly has now that’s imbedded are a good example of my complaints. The new popup seems to steal the input focus so that I can not switch to an xterm or other app without having this new popup disappear. If I’m trying to copy data from the popup, I’m just out of luck.
All these problems that I see and hear about really do point to a seeming lack of consideration of the users by FF developers AND architects and what the users want. In my experience, if you tell customers no too frequently, they’ll go elsewhere.
I want FF to succeed! This is why I use Nightly and suffer (sic) with the daily updates.
I just wanted to add that Yahoo and it’s “new” UI is another case in point. Seemingly changed RADICALLY, removing features, substituting features that don’t work as well… ALL done without customer feedback. Sigh…
George…
The perception is the reality you have to address. Whatever you feel is going on, users feel that Mozilla disregards them and even treats them with contempt. (I’ve seen it many times, but the most memorable was the response of a high-ranking Mozilla employee to corporate IT complaints about rapid release.)
You can say it’s not happening in that Mozilla isn’t doing it, but it’s certainly happening in the sense that users feel that way.
I suspect that Mozilla’s non-profit nature may insulate it from users. For-profit businesses are highly dependent on their customers and go out of their way to maintain a good relationship with them; for example, nobody would make public statements like the Mozilla employee mentioned above, and if they did the company would apologize, retract the statement, and likely fire the employee to assuage customers — compare that to Mozilla’s response. In a for-profit business, if sales drop then the people responsible are often replaced– does anyone lose their job at Mozilla if Firefox usage drops? It seems like the same pressure and relationship with customers is not there.
I’m not advocating that Mozilla needs to be for-profit, but that one it should be aware that those market signals are what make businesses responsive to customer demand, and one weakness of being non-profit is that Mozilla doesn’t receive them as strongly as for-profit companies.
I think the complaints about the UI changes miss the fact that landing a huge swath of UI changes all at once is at least as disruptive (and in my opinion more so) than landing those same changes bit by bit as they become available. It’s much easier to adapt to two or three UI changes every 6 weeks than 30 UI changes in one hit every 14 months since the latter invariably means a completely revised UI.
I like the rapid release and I like not having to wait a year or more for improvements.
I hated having my extensions break every six weeks – that was the disruptive part and it’s been resolved. And before rapid-release, x.y updates were just as disruptive to extensions. I detested pulling down a security fix and finding that 1/3 of my extensions didn’t work – the more so because the extensions I run are critical to my workflow. All RR did was expose how broken the relationship between extensions and updates truly was.
I am neutral about silent vs. doing the update explicitly, as long as it works with a restricted user – i.e. the updater must know that I am restricted and either elevate or not attempt (and then fail) the update. Personally, I would prefer a notification popup with “Do It” and “Remind me later” options, as long as “Do It” actually worked as a restricted user.
Fine, argue semantics. If anything, Jono’s only guilty of a little hyperbole, and that’s covered under common rhetorical license.
What we all know he means to level against Mozilla is…
“We [Took] Our Users[‘ Aversion to Intrusion] For Granted[, As Clearly Evidenced By Our Now Admitted Lack of Foresight Over The Update Experience]”
Saying that “we didn’t take our users for granted” when the crappy update process did, for a time, exist, doesn’t excuse, but prove Jono’s point. You were so far in the forest, all you saw were trees. “Sure, it sucked, and sure, Chrome’s % is going up, but it was all for YOU!” These are not the boyfriends girls keep (SOs we keep?).
What you do now is thank Jono for identifying a weak period in Firefox’s development, try to have better and more representative user testing in the future, apologize to Jono and users, turn the bad pub into good by capitalizing on the opportunity to pimp the new update process, wear the egg, and move on.
Don’t fight great criticism. Embrace and redirect it.
@Nuss,
You stated, “Whether Firefox 13 should be called 4.1.8 or 5 or 13 or 2012.3 is completely irrelevant. It’s practically invisible to users. There is no canonical version numbering scheme. They don’t mean anything except that a larger number is released later.”
It is relevant to anyone using addons. Major version number changes break addons. And that is the biggest ongoing complaint by any user, even since “Compatible by default” was featured in v10. And I’ve responded to over 8,000 email over the last year my Mozilla users.