Read the complete 8,000 word interview with Mozilla CEO Mitchell Baker, who talked to APC about the future of Firefox. If there's one tech article you read this year, read this one.
How 12 people made Firefox 1.0
|Hair, there and everywhere: according to Baker, her hairstyle's similarity to a Firefox tail is entirely coincidental
Dan Warne (APC): Firefox seems to have been a stunning success, but it can be hard to get a true idea of just how small it was to begin with, and how big it has become.
Mitchell Baker: Yes -- the Mozilla project started in 1998 and was originally housed inside Netscape - a virtual organisation. We had thought about having an independent organisation for a long time and finally in 2003 it was decided it was time to give us some seed money, some critical assets and the Mozilla name.
But in the short‑term even the development machines we had were pretty critical, as there was no known financial sustainability path at the time.
So the foundation started with ten or eleven employees and we grew by maybe two or three employees in the next 15 months. So when we shipped Firefox we had maybe 15 employees, it was very slim.
Dan Warne (APC): That's an amazingly small team for such a complex product.
Mitchell Baker: It might have even been 12. It was very slim at the time. It was really a leap of faith on the part of all of us that it had to happen and something would work out.
We had a pretty good inkling before we shipped Firefox 10 that we were on to something and that our product and the timing was right. Probably as early as summer or maybe May or June of 2004 we could tell by the level of technology interest and the number of people using it that we were in a good spot.
But we couldn't predict what happened with the actual release of Firefox 1.0 and how it took off. So that was unexpected or unpredictable. I think a launch like that is pretty hard to predict. It suddenly seemed to storm out of nowhere, although we had been doing it for a long time.
One of the things that we spent a long time thinking about as we got Firefox 1.0 ready for release was how to make it as much a consumer-focused product as possible. This was a big change for us; the design of the product has always been focused towards your grandmother or your grandfather, some smart respected person who was technologically not focused. How would that person enjoy using the Internet?
So that had been built into the product, but as we came down to the end stretch there were certain other things like what did it look like the first time you started it up? And today I think Firefox is pretty well known for the search box and the simple start page and so on, but that was a long, long, long discussion for us.
Dan Warne (APC): Really? It seems like such an obvious thing.
Mitchell Baker: Yes, because this was one of the very early moves from being really developer-centric, not in the product design, but in the presentation of it.
We had very long discussions about whether it should be a portal page, what do people really want, should we leave it with the Mozilla "please help us write code, here's our bug tracking system" page, which is what it was before, things like that.
We finally came to the conclusion that the old developer start page was not appropriate and that the one thing we knew people did was search. People do many things on the web, but if you're trying to find something that almost everybody does search was the only thing that we were really comfortable with.
So that made our choice for the start page search and it turned out that Google was interested in doing that with us and creating a joint start page. So that's how we ended up with the start page. I think that was a good decision because a lot of people really do like it and even now, when you look at people who customise their start page it varies widely and trying to pick something other than search I still today wouldn't know what to pick.
We also came to agreement with search engines about revenue related to search. So as Firefox market share took off - and I think it grew at a percent a month for the first X months - it turned out of course many more people were using it than we expected and the combination of search and the start page and the search box turned out to make search really useful.
So more people used Firefox and more people did more searches than we expected out of that. And so our hope for some sustainability model turned out to be positive, it turned out well and it also generated more money than we would have expected in the first place.
How Firefox makes $US55million a year
Dan Warne (APC): I was going to ask you about the money you earn from your search box in Firefox because it is pretty widely known that Google pays for inclusion of Google search functionality, but people have very little idea of how much money Google pays.
Mitchell Baker: Well a couple of things to note here; it's not just Google -- they're not the only ones. One of the things that we pioneered - I mean people do think of Google because they are the default and the search box and the start page, but one of the things that we pioneered was multiple choices for the user.
If you look there's Google and there's Yahoo right next to each other and maybe people take that for granted now, but it was not taken for granted when we first did our early deals. Google is the default, but we were adamant that there had to be a choice and that other search engines that we thought users were interested in that made sense we might ship by default and that users can add anything.
It is very easy to add in a search engine. So that was a fundamental element of the user choice aspect. So we have arrangements with more than Google . We're non-profit so our numbers are audited and on the public record. We haven't finished 2006 yet so they're not public, but the 2005 numbers are around $US55million.
Dan Warne (APC): Wow! That's a spectacular amount of money. You could do a lot of work with $55million a year I would have thought!
Mitchell Baker: Yep it's I think unprecedented for an open source project like us. Of course currently that money comes into the corporation as taxable so you don't get that much of it, but nevertheless it was unexpected and very welcome and it allows us to do many things which is important because the browser is so fundamental to the Internet.
We were trying to do it on a shoestring and we were pretty remarkably successful through 1.0. The opportunities have grown and so trying to do it on that shoestring now would be very hard to be successful so that's been a boon particularly because users like it, search is useful as opposed to other things, like selling off buttons on the browser interface which might make money, but not be useful to people.
Maybe some day we'll find another one that people like and that would be great, but in the meantime there's a lot of opportunity to generate revenue out of the browser as it is now.
Conventional business processes say you absolutely diversify your revenue sources, which ideally would be nice to do but not at the cost of product.
Dan Warne (APC): Yes, many internet applications in the past have crossed the line and really annoyed users with intrusive add ons.
Mitchell Baker: Right ... so we ended up or started down this path of people really loving Firefox and sometimes I think it sounds funny to describe how much people actually love a piece of software and how excited they get and how much people are willing to do to help other people adopt that and to build it, and create it. But it actually happens [laughs].
So we saw our user base grow, we saw the level of excitement -and another overused word -- ‘passion' -- related to Firefox - extend beyond our core development community to unbelievable numbers of people and that is an asset that's priceless. And there is that sense of trust that yes it's a great browser and it's better than I had before, but I also trust it.
But that is the most fundamental aspect I think of Firefox and partly it's because the product is great, partly for those that know it's because we are a public benefit organisation and we are not trying to maximise our revenue and we're not trying to generate massive private wealth for a few people. The asset is owned by the public.
I think that many people don't know about the open source nature or how it is built, but actually somehow feel that in the product, that the end result reflects how it's made and it is made in a very community‑based, very user focused way and that comes through somehow.
Dan Warne (APC): Well it seems to me that one of the most attractive things about Firefox is the plug-in eco system and I think it's amusing to watch Microsoft trying very, very hard to replicate that, but their plug‑in eco system is full of "pay $30 to register this", "pay $50 to register that" - it's all commercialware and I think it is testament to the fact that in your open source model it's not easy to replicate that unless you are actually open source.
Mitchell Baker: Yes, yes well first of all it's hard to replicate interest in public benefit as opposed to shareholder personal wealth because fundamentally you're two different organisations and each has legal constraints that drive you in a different direction, so that, I think, by definition can't be replicated.
The framework that we operate under with decentralised voluntary action moderated by intense discipline and quality control at the centre also is pretty hard to replicate in a setting. Certainly Microsoft has had various shared source and sort of collaborative or sharing initiatives and good for them. I mean that's a step forward right but it is not the same as fundamentally sharing what you are doing through a leadership structure, but not an employment and financial structure.
So that's hard to replicate and then the ability to actually work with Firefox and the quality of the technology and the ability to work with it and make something interesting that solves your problem using Firefox, it's certainly nowhere. It is pretty extraordinarily hard to replicate.
Putting Firefox on mobile phones
Dan Warne (APC): Something I wanted to ask about is one area that Firefox hasn't seemed to have delved into much is putting Firefox on mobile devices and Opera obviously has a pretty good spot in that space and even Microsoft to an extent. Is there any move in that direction?
Mitchell Baker: Yes it is a long‑term move though -- it is not in the next weeks or months. The Mozilla Foundation's mission in life is to improve Internet experience and that is increasingly on devices other than PCs. If we're not there then we won't be able to live the kind of vision that we helped grow.
So that needs to happen. We had a small project looking at it but we decided that the right thing to do is to look first at the technology and look at our core technology and really tune it so that it's best suited for that. We are at work on that now, however it will take a while.
We are also looking at how to reflect the richness of the entire web on a small device with the current constraints and that one we don't know. There is no easy answer for that because the web is growing and the functionalities of the web are growing. We're looking closely at that one.
We have an experiment underway which is clearly a PC-centric experiment related to mobiles (so this is not a pure mobile strategy), but we are experimenting with the relationship between Firefox users and their mobile devices. We know people like Firefox because of the add-ons and the customisation and the ability to get particular information that you want through Firefox and extensions.
So the experiment that's underway is called Joey - we are looking at how you can take information that people like to access and deliver it to a mobile device. Clearly you can go to the web and you can SMS yourself various things ... but what else could be done? If you have Firefox, you already have the ability to customise it and gather certain kinds of information, so what could you do with that so that being a Firefox user makes your mobile experience better.
Dan Warne (APC): So are you talking about a kind of a back end service solution that helps pre-format content for mobiles etc?
Mitchell Baker: It's got a couple of pieces, it's got a server piece, it's got a little client piece that sort of thing and that one we'll launch in labs pretty soon. I think there's information up there about it now so that will launch as an experiment, not necessarily as a product plan, but as a way to start to get information because right now in many countries the user of a mobile device interacts with a carrier and not directly with the software vendor.
So even if we had a great product to do what we do best which is to touch human beings there's no current way to get that on most devices. Another thing that we're trying to sort through and maybe if that doesn't change then maybe this ability to get different kinds of information would make sense, we're not sure.
Opera is probably better suited to us as a supplier to carriers because our DNA is really very consumer and individual-focused. So all of those things have led us to try this experiment whilst we tweak our technology and see what things look like.
Dan Warne (APC): Firefox has a bit of a reputation - I'm not sure whether it's right or wrong - of having a hefty code‑base as a renderer and I think that largely might have come about when Apple chose the KHTML rendering core used in the Linux Konqueror browser. They said at the time reason they chose KHTML over Gecko [Firefox's rendering core] was because it was very lightweight. So is it true that Firefox internally is quite hefty and might be a bit difficult to shoehorn onto a mobile device?
Mitchell Baker: Oh well all of them are difficult to shoehorn onto a mobile device, so we should be clear about that. Opera has done a pretty good job of getting something useful on to a mobile device, but it's not a full fledged and doesn't have the capabilities of Firefox. That's hard to get on any mobile device so that's a separate question.
But yes I think it's fair to say that the actual and better piece - it's called WebKit by Apple - that particular piece is easier to work with currently than our analogous technology.
Now some of that is that when you get our analogous technology you get a whole bunch of other things that allow the creation of the communities that we do. So for extensions and the language of XUL and a whole range of other things come with our technology.
So some of that is just more capability but some of it is that it is a smaller piece and probably more approachable and that's part of what we're looking at when we say before we really launch into a better space we know we have these advantages that I think are probably unmatchable - building that kind of community that we talked about - it's probably unmatchable.
But even given that we should really work harder and smarter to make the core as approachable and easy so that it's easy to develop and you get all the other benefits that come with Firefox and Mozilla technology.
Business giving the cold shoulder to IE7; recontemplating move to Firefox?
Dan Warne (APC): Something's up with IE7 adoption. What we see on the APC site - which would be almost exclusively tech early adopters and enthusiast readers -- is that a lot more people are still using IE6 than IE7. Do you think Microsoft made too much of a dramatic change with IE7, which might have prompted users to stick with IE7, and has that somewhat unsuccessful strategy informed your point of view on how you should move ahead with changes to Firefox? Like do you see a need to be more subtle in each evolution?
Mitchell Baker: Well a couple of things I am certainly not an expert on Microsoft. I look at IE once in a while, but it's not a browser I would ... use very much.
Dan Warne (APC): [laughs]
Mitchell Baker: Firefox was actually designed to make migration from IE very easy, so we know many people almost don't realise the change.
I have been told by some large business uses that when they look at IE7 they are re‑contemplating a switch to Firefox or support of Firefox because the move from IE6 to Firefox might, in their case, be easier than the move from IE6 to IE7.
So that's an interesting discussion point. I am not running sophisticated studies, but when you start to hear these things, which we didn't hear so much before (we just heard "oh I don't want to make a change.")
But Microsoft certainly does not inform our product plan. Not at all. We've been the leaders for a number of years now in browsers and how to help users understand the Internet.
The underlying question we look at very carefully because we have a great interest in a set of users, early adopters and power users for whom more new things constantly is a requirement and drives people and they get bored really easily. So we have a very vibrant and active and important part of the Firefox community for whom new and moving forward and new possibilities is fundamental.
Dan Warne (APC): One thing us easily bored tech early adopters love to argue about is the penetration of Firefox. How many people are using it now?
Mitchell Baker: As best we can estimate our user base it's somewhere between 75 and 100 million users ...it's a lot! These numbers aren't exact because they are all correlations of different data points, but if you take 15 per cent of the worldwide web and you look at the number of people who come to get security updates and try to correlate that's about what it looks like. Every rule of thumb that we can find suggests something like that.
Dan Warne (APC): So is it tricky to keep everyone happy?
Mitchell Baker: Most of those people are not power users and don't value constant change above complexity. Many of those people really don't know - even what's Firefox and the what's the chrome of Firefox versus the dynamic content that comes from a web page.
So that's very complex and very frightening to people so the issues that inform our product planning are how to continue to expose the richness of the web which is changing all the time; how to bring the new ideas that are early adopter advocate community wants to see and is producing, and at the same time how to continue to ship a product that the general consumer can use?
That is what informs our thinking constantly and it's quite separate from what decision Microsoft may or may not make.
Of course we use the add-on system to meet many of those needs and so I think you will see more and more new ideas show up as add-ons that we may even develop or promote to test for experiments like The Coop to see for people who are early adopters or interested in something, what really works, what works well and how do we take the best of that and put it in a core Firefox so the general consumer can actually approach it without being terrified.
Why there's no inbuilt adblocker in Firefox yet
Dan Warne (APC): So speaking of add-ons I'm sure a lot of people are curious to know why you guys haven't built an ad blocker into Firefox?
Mitchell Baker: I don't know I would have to ask actually!
Dan Warne (APC): Really okay? I hypothesised while thinking of the question list that the reason is that the web more or less survives on ad revenue and you wouldn't want to piss off every publisher out there. But I have no idea really which is why I was curious.
Mitchell Baker: I don't know actually. It may be that they are too technically complex, but I would have to ask.
Dan Warne (APC): No worries. I was just curious because there are popular add-ons that do it.
Mitchell Baker: Right and I guess some of those add-ons are pretty complex. When I look at them, they ask you what element you want to block and so on. How many ordinary users would be able to do that, I wonder.
Dan Warne (APC): I think most people just download the automatic list that has everything that has been pre‑populated by someone else.
Mitchell Baker: Right maybe that's it maybe that would be a better extension. Still, I actually don't know what the answer to that would be without looking into it further.
Firefox 3.0 – what’s all this ‘lock-in branding’ stuff about?
Dan Warne (APC): I was going to ask you something else about plug-ins. We had a look at the Gran Paradiso alpha release notes and one of the things we noticed in there support for persistent branding of Firefo I assumed it was so, for example, Dell could pre‑install Firefox on their PC and have a little Dell logo in Firefox or something? I don't know what it meant exactly, but I'm hoping it's not going to be like the IE4 days where Microsoft made it possible to brand e-v-e-r-y element of IE4 and so you had these disgusting browser windows with giant writing and logos all over them.
Mitchell Baker: [Laughs] No, no I can't imagine we'd be doing anything like that...
Mitchell Baker: Speaking of branding, the code names, as you may know, are park names and so Firefox 3.0 is codenamed Gran Paradiso, and there's a Gran Paradiso Park. Recently when we were looking at the code names and for some reason we went to the site of the Gran Paradiso Park, there at the bottom of the site was a ‘get Firefox button', I went "yes! that's the right code name!"
Dan Warne (APC): Ha - that's a funny coincidence isn't it. Let's look it up the Gran Paradiso alpha notes on the web and see what they say. [Turns to Edelman public relations person] Can I use your laptop for a sec?
PR guy: Sure. [To Mitchell Baker] You have to excuse this laptop ... it doesn't actually use Firefox.
Mitchell Baker: Now is when I storm out of here [laughs].
PR guy: I was going to install it myself but I don't have access rights.
Mitchell Baker: To that machine?
PR guy: Yes unfortunately.
Mitchell Baker: What kind of machine is it? Where's your managing director? He came in earlier today ... now I'll complain [laughs].
Dan Warne (APC): Ok, so it says here about Gran Paradiso: "ability to lock in branding."
Mitchell Baker: Right... well I think the key thing is to figure out what "branding" means in this circumstance. Sometimes branding is shorthand for a set of things.
For example, a telco, T‑Online, has distributed a version of Firefox in Germany in which the start page is T‑Online and it's co‑branded. The idea was that it would be simple for T-Online to distribute the browser with their start page.
But it turns out that when you have to do updates it is very, very complex and making sure that when you do the update that the start page remains T-Online rather than the start page for our German default, which is I think Google, is really tricky.
I think "branding" encapsulates all those sorts of things.
Dan Warne (APC): Like how mobile phone "branding" encapsulates everything on the phone for a carrier?
Mitchell Baker: Right, for things like that. So it's not what you are worried about.
Firefox to go head-to-head with Flash and Silverlight
|Flash vs Silverlight vs Cairo: can Mozilla take animation and video open-source?
Dan Warne (APC): Another thing that I saw in the Gran Paradiso notes was some mentions of Cairo rendering engine and as far as I have read about it it's a sped up, good quality rendering engine for scaleable vector graphics.
Mitchell Baker: Yes, 2D graphics at this stage.
Dan Warne (APC): So this might be a long shot, but I wondered if that at all will come into competition with Flash and Silverlight?
Mitchell Baker: Well clearly graphics is an area where the web could be better and so we're all focused on it. We are focused on it with really open technologies and of course Silverlight is completely proprietary, and clearly Flash also proprietary, though Adobe is making some roads towards openness.
So we actually - and I think Brendan and his roadmap - has been suggesting to Adobe even more moves towards openness that we might do together and we would love to see an open version of Flash, but that's Adobe's decision not ours.
So I guess in the sense we're all looking at graphics yes and ultimately I guess if Flash were to remain as proprietary as Microsoft technology then we'd have to continue to try and develop graphics in an operable way.
But Flash is clearly dominant in the graphics' space.
Dan Warne (APC): One does wonder why Microsoft would bother with Silverlight when it is so late into the game.
Mitchell Baker: But it is so critical, I mean we're doing the same thing and we're doing the same thing because Flash - yes it's proprietary so to us that's kind of a problem, but why does it really matter? Well it doesn't live - it is on the web, but not of the web, it's not searchable, it doesn't share all the features of the browser, you can't operate it, it lives in a little box.
I imagine there are probably efforts to move it out of the box, but to really integrate with the rest of the web, some of those capabilities ought to be in the web client, which is the browser. And so we continue to hope on that front, but also to develop graphics capabilities ourselves.
Dan Warne (APC): That's really cool. Would that encapsulate any video capability as well?
Mitchell Baker: We're looking at that so there are a couple of things. One is clearly the text‑based web that browsers grew up looking up is changing and video is going to be increasingly important.
Dan Warne (APC): It's really boomed in the last sort of six months to a year hasn't it?
Mitchell Baker: Yes once capabilities are there, there's an easy mechanism. So there are a couple of them there is the question what should a browser do to support that and both at the application level that human beings see and on an underlying technology level.
So we are looking at both of those to try and figure them out. It turns out that once you get into some audio but video things there are a lot of complexities.
Dan Warne (APC): Is there ever. Endless combinations of CODECs ...
Mitchell Baker: You have to work your way through; some of these things are clearly patented so you have to work your way through some of those things. So we are looking at it, but we haven't made our way through all of those things yet.
Dan Warne (APC): Do you run into that very often; run into like we have an idea we would like to do something, but then realise, "oh crap someone already has a patent on that?"
Mitchell Baker: Well it's constantly in the background. We haven't struggled with it too much.
Hot or not: the relationship between Microsoft and Mozilla
|Anti-crust: not strictly part of the legal settlement, the IE cake Microsoft sent to Mozilla upon shipping Firefox 2.0 put frosting on an icy relationship...
Dan Warne (APC): Also an interesting thing that happened a little while ago was Microsoft's very public invitation for Firefox people to come into the Vista compatibility labs. I remember asking Microsoft at the time if they could give me more background to that and they said to me something along the lines of "well we've extended an invitation to our competitor and if they choose to accept our invitation well that's up to them, but if they don't it's not our problem." Do you have any comment on how that went?
Mitchell Baker: I think it is fair to say that Microsoft did make an effort and it was welcome. There are many areas in which we would like to work well with Microsoft and certainly I think Vista's compatibility there might be some legal requirements there on their part as well, but whether there are or not they did extend that offer and we did take them up on it and they did host Mozilla time up there, which I think was useful for us; and I think that's worth some credit.
Microsoft is back in the browser space now and has to pay attention to the web again so we'll probably see them in the standards bodies and that's probably where we will see - not those particular people in the Vista lab and I think they did make an effort and they did work so that we could get up there and do some testing on Vista - but I think we will see in the standards bodies whether Microsoft continues to push its own proprietary technology at the expense of open and operable technologies. That happens a lot in standards bodies meetings. That is where their true colours will show and we'll see what that is probably pretty soon.
Dan Warne (APC): That's interesting. Were there any issues with Firefox and Vista because I mean I was running all Vista betas and I installed Firefox on all of them and they basically worked fine.
Mitchell Baker: Yay! We did do some work - I am trying to remember it was; I think there are some installation differences that are required.
Dan Warne (APC): That would be probably be likely given all the changes to permissions and stuff in Vista.
Mitchell Baker: Yes and so - yes actually the basic product worked extraordinarily well we were proud of that, but there has been a set of tuning and installation and I think there was one other category and those things came out and - two or three I think. So not at the 2.0 release time, but we've made a lot of Vista-related changes, although again not in the basic functionality, but just pieces around it.
The strategy for getting Firefox onto more desktops
Dan Warne (APC): Last time I spoke to you I asked you about the distribution strategy for Firefox and I think it was probably around the time of Firefox 1.0 and you at the time were saying something along the lines of, "well, we're not commercial and we'd rather work with users, not corporations; we hope users take it up by themselves," which evidently they did in great numbers.
But nonetheless can you talk a bit about what is your strategy for distribution and are you are looking at deals with PC manufactures pre-installed or anything like that?
Especially now that Dell is doing Ubuntu pre-installed onto PCs... I suppose it means that Firefox goes out with every Dell PC that has Linux on it.
Mitchell Baker: One would think so!
Dan Warne (APC): So, looking at one major distributor you have - Google. What is in it for them, how they distribute Firefox and actually pay bloggers a $US1 commission every time someone downloads Firefox from a link on their blog?
Mitchell Baker: Well for the official Google rationale you need to talk to Google. I think it is some variant of the theme that Firefox is good for the world.
|Google dollars: stick a Firefox badge on your blog and Google pays $US1 for each new user who installs it based on your referral
Dan Warne (APC): I guess a company with as much money and autonomy as Google can do that without worrying about what shareholders might say about squandering funds.
Mitchell Baker: Yeah, but I am not sure it's squandering funds.
Dan Warne (APC): Sorry, I don't mean that in an offensive way. I just mean that often, shareholders are judgemental about the way money is spent if they can't see a direct financial return to a company.
Mitchell Baker: Well it's also pretty clear that Microsoft is aiming at Google and they have said that over and over and over pretty explicitly. So it's probably hard to argue they're squandering in any sense when you think of what are the battles that Google is fighting. Microsoft seems to be on the table there.
Dan Warne (APC): What about other distribution methods... preinstalls onto PCs for example.
Mitchell Baker: We've tried other distribution alternatives and we continue to explore them. Firefox has been included on a couple of PC production runs; one in Europe I think, but those are new enough we're not sure how they're going to work out.
Actual PC distribution would be interesting, but mostly we have been very successful in distribution ourselves, driven by community word of mouth, or the marketing or whatever it is that we do that brings people to us is far and away the most effective. We keep looking at all the angles though.
Mozilla Japan's cartoon character, "Foxkeh" and more Firefox development in the Asia/Pacific region
Dan Warne (APC): Since you're here in Australia - in the Asia/Pacific region -- are you doing any APAC initiatives or do you have any links with organisations here that are interesting?
Mitchell Baker: We have for many years had an active community in Japan and it is a community‑based group with a very small organisation, Mozilla Japan, which is non‑profit. Once we had revenue about a year ago we started to fund an investment in Mozilla Japan so that's probably six or seven or eight people with some developers, some outreach and doing very interesting things.
So in fact I will just show you this. This little creature here is Foxkeh which is a creature that the folks in developed to make Firefox accessible and for viewing in the market there.
|Kawaii: Oh so cute... Mozilla Japan's firefoxy cartoon character Foxkeh
They are doing some pretty excellent work on extensions and also developing documentation in Japanese and so on.
We have also had for a couple of years now a very, very low key small effort in China which we are a small good project hosted by the Chinese Academy of Sciences and in the last few months we've decided that we should be a little more engaged in China.
We've hired someone to help us. I was there a few months ago and there is a set of people in China, maybe the best we can tell like a million users so not a lot of users for China. But an interested and excited user base and the bloggers and so on and some of them are very interested.
So our thinking is: well let's go and find out who is there and who is already interested. Let's go and see if we did more, whether more people would be interested and if this vision of the Internet that we are working so hard on resonates there. If it does and there's a community of people then let's see if we can be helpful.
We expect that to be a pretty significant exploration in the next year.
Dan Warne (APC): Any other contributors here in Australia or nearby?
Mitchell Baker: There are a few yes. We have for many years had a wonderful contributor, Robert O'Callaghan , who is a New Zealand native and he's been living in the United States - or he was living in the United States - for a number of years and desperately wanted to go back.
So he is back in New Zealand now and he works for us and we've started a small office in New Zealand.
Dan Warne (APC): Really?
Mitchell Baker: Yeah because it turns out when you have one really great contributor who is also a good teacher and willing to mentor and can attract a few other smart people around him that's worth a lot to us, that's everything.
In fact the offline apps demo that was done this morning that showed a demo of Zimbra running offline, without an internet connection, that work was done out of the New Zealand office.
|Robert O'Callaghan: back where he belongs, developing for Mozilla from New Zealand
Dan Warne (APC): Wow! Are you doing any work in Australia as well?
Mitchell Baker: Not currently. However, one of our longest contributors to the Mozilla project is a lecturer here in Australia at University of Queensland and he's been involved in the project since 2000 I think. He's a mathematician or at least a scholar in an area in which mathematic equations are important.
So way, way back when he decided - there's a language called MathML - and he decided that Firefox Mozilla in those days ought to have it. And he has been contributing ever since and is single‑handedly responsible for the ability to really persuade confidence in this area.
Dan Warne (APC): Right... I have always been curious about why on earth it is that Mozilla has such comprehensive support for mathematical equations?
Mitchell Baker: It's all due to Roger Sidje from UQ. It was a long hard battle for him in the early days.
Dan Warne (APC): Really?
Mitchell Baker: Well we're talking 2000 so actually getting that complex and different rendering elements into the core rendering engine is hard.
Dan Warne (APC): It's almost like writing a whole different HTML renderer, I can imagine, because mathematical equations can be formatted in very complicated ways.
Mitchell Baker: Right and he's been active so long now. Since we started getting revenue, we've been hosting a Firefox summit of key contributors and non‑employees from around the world, so Roger has come to visit us a couple of times because he has been such a great contributor for us.
The stoush over Linux distributions using the Firefox trademark
|The making of the Firefox logo: (hicksdesign.co.uk)
Dan Warne (APC): I was actually just in China and I met the guys from Red Flag Linux where they are working with Intel on a new Linux operating system called MIDINUX for mobile Internet devices.
But they were also telling me that Red Flag Linux is actually the most used distribution of Linux - more popular than Ubuntu for example - because of China's huge population. Yet, I discovered it's nearly impossible to download from anywhere due to the incredibly slow and unreliable Red Flag download servers - plus it's only available in Chinese and Spanish at the moment which I guess limits its appeal to the rest of the world somewhat.
But anyway, the guys from Red Flag gave me a set of install CDs and we're publishing them this month on APC's cover disc. We installed it to try it out, and Firefox is on there and it's an icon that's remarkably similar in style to the Internet Explorer icon. [laughs].
Mitchell Baker: Well that's interesting.
Dan Warne (APC): And the open source media player has the Windows Media Player icon too!
|Red Flag Desktop 5: first we'd heard of Windows Media Player on Linux too! (Screenshot - distrowatch.com)
Mitchell Baker: Really well we should look at that that's really ... interesting.
Dan Warne (APC): It's bizarre because what they've done is the entire operating system looks exactly like Windows XP - it's almost as if they've taken all the Microsoft icons and stuff applied to open source software.
Mitchell Baker: To the open source software that's not remotely related to Windows XP though.
Dan Warne (APC): Exactly! But evidently the felt people were most comfortable with Windows XP, so they'd just make it look like that. It was really quite bizarre.
Mitchell Baker: Yeah that is.
Dan Warne (APC): But on this general subject, there has been quite some controversy around the use of the Firefox trademark in various Linux distributions. What's all that about?
Mitchell Baker: Some of the Linux distributions ship Firefox code but not Firefox product -- not the Firefox brand and so on. Some of the Linux distributions are very sensitive to trademark law and so they brand the browser as something else which allows their groups to work within in the framework they're comfortable with.
Dan Warne (APC): Do you, as CEO of Mozilla, feel that's necessary? Would Mozilla ever take legal action against a Linux OS using the Firefox trademark?
Mitchell Baker: Well it's not that it has Firefox in it that would be a problem for us. It's if an OSS group had made significant changes to it and then shipped it as Firefox. That would be a problem and this is a philosophical kind of difference with the Linux distributors sometimes. It is a pretty complex issue, but even something as precise as - when somebody ships us Firefox we want to know exactly what version is in it and that every security patch that we think is in Firefox is in that version.
So that's the issue for us. If they are just shipping unchanged Firefox we would let the world do that, but the Linux distributions of course ship the product, so that's where we run into the setting where we say you are welcome to ship whatever you want, but that's not what we shipped, then call someone else.
Dan Warne (APC): Which is more than fair.
Mitchell Baker: Well, yeah, but it has been pretty contentious.
Dan Warne (APC): I must admit I'd never heard a clear explanation on the issue. I have to confess I'd assumed it was just some sort of turf war that's typical in the technology industry - small open source product grows into big success and starts acting more commercially.
Mitchell Baker: Oh no, not at all. We absolutely want them shipping our trademark for our product, and we even have a process which says these are some changes you can make and still call it Firefox because there are some installation changes to get it installed on different Linux distributions, for example, so of course that makes sense.
So we're even OK with shipping "Firefox" when it's not 100 per cent exactly the same, but when we've agreed and know what those changes are.
But when you get into difference in functionality, you strike trouble down the track with upgrade cycles and those sorts of things. Long ago I think we found some distros were shipping different builds of Firefox - something really fundamental that was different and so then we couldn't assure upgrades would work well. So that's what that was about.
Mozilla working on Web 3.0: web apps that run offline too
|Web apps without a net connection: would you still need Microsoft Office then? (pic - zimbra.com)
Dan Warne (APC): Your CeBIT keynote on May 1st talked about running web apps offline without an internet connection. That's incredibly interesting because everyone's waiting for the ‘holy grail' Microsoft Office killer to come along, yet anyone who's ever used one of the upcoming online office suites is aware of the problem that you can't use them when you're on a plane, or otherwise disconnected from the net, for example.
Mitchell Baker: Right. The big picture is we are investing enormous amounts in the web itself as a platform, which is what capabilities does the web have and are they robust enough to be viable alternatives to proprietary technology like Silverlight, for example?
And are they also robust enough to make the web attractive in comparison to desktop software, because that's the platform that we care about.
And so offline apps are the ability to work offline for some period of time with the web - it's one of those things that would move the web platform forward. We have actually done a fair amount of work already, so what's happened so far is that most or all of the basic code - the databases and storage work necessary to support offline web apps -- is in the shipping versions of Firefox now.
Those capabilities have been in there since Firefox 2, but people haven't known and picked up on it. We thought, "okay, we should do a little demonstration so people can see what it is and we are now growing enough that we have enough resources to be able to do some of this." The guys in New Zealand working on this are phenomenal.
So we thought Mal would be a good one to demo it and so he worked with Zimbra, which of course he used to work with as an open source developer. So he has done a pretty nice demo of what can be done.
Then there has been some work in the Mozilla Development Centre to explain what that is and provide some assistance. So you could find info on Mozilla offline apps in the Mozilla Development Center - we are very proud of that. We actually have a pretty professional set of documentation now. There is sample code there I am not sure if that is there yet.
So the plan is that all the functionality required for shipping Firefox Two and we are hoping that we'll have some other, not just applications that people can sit and touch and feel and so on so that will be exciting.
Dan Warne (APC): Very exciting stuff. And so the last question... how long until Firefox Three?
Mitchell Baker: [Laughs]
What we are aiming for is the end of this year but we won't know until we do the betas. That's where we get the degree of testing that we can actually make an accurate estimate.
It takes somewhere between 50 and 100,000 people - so maybe 70,000 people -- using something to really have an idea of where you are. That is what we find that open source testing is a phenomenal benefit and when we get to the betas and we get those kinds of numbers because the web and the use cases are so broad that testing against the web - even as you automate you can't automate everything.
So when you get that number of people you get a lot of information that tells you a lot more than whatever you thought before. So that's when we start assessing how many betas [inaudible].
Dan Warne (APC): I will be one of them. Thank you very much for sharing your thoughts so generously.