An increasing number of computer users are turning to online applications instead of ones on their desktop. It started with webmail and has moved to productivity/office tools. With the emergence of online applications that have no desktop equivalent, and mobile devices that are browsers in your pocket, things are looking up. But what about free software? If the software we are using is not run on the computer on our desk/lap/hand what does the licence matter? For some time now I've been reading predictions where the browser will be the computer. Does this future have space for free software?
For an online application, the licence terms shift from what you can do with the software more towards what the provider can do with your data. Google recently got in trouble with the licence for its Chrome browser. It had inadvertently included standard terms from its online application licence. Specifically those that said that Google could do pretty much what they liked with any your data used through the service.
So where does freedom and copyleft fit into this model? The four freedoms seem generally to be disregarded for the user of an online application. You must use the software within the limitations set by the application provider; you do not get the source code, so you can't study, modify, improve or redistribute it. But does it really matter? Is the licence important if it doesn't apply to the source code -- after all a number of free software advocates have Googlemail addresses? Even if the software had a free licence, would it make any difference? You can't -- after all -- force the provider to apply any patches you submit and you've no way of telling that that code you downloaded was exactly the code used on the main site.
So I ask again: if we're all heading towards a world where the browser is the computer will the licencing matter for much longer?
Of course this is Free Software Magazine so you'd be expecting me to say "yes" -- there , I've done it. But how can it matter given the circumstances above? I happen to believe it matters even more, in a couple of ways.
Firstly let's talk about trust. How much do you trust Google or any other online service provider? One of the reasons given for GPL freedom 1 (the freedom to study how the program works, and adapt it to your needs) is to look for security holes, back doors and "phoning-home" features. True it's mostly so we can study the code, learn from it and see how it works but by nature that means the code author/publisher will be found out if there are any less-than-forthright practices within it. Are online applications suddenly trustworthy? I say: no more than ones I install with apt-get.
In addition software innovation thrives in environment where coders can learn from each other. When the code is not kept secret the advancement and improvement of the software increases exponentially. Look at the free desktop environments around -- I think it's hard for anyone to argue they are of a standard lower than proprietary alternatives. If anything, they are at the leading edge in terms of features and quality. Have the leaps and advancements of the past few years been solely down to competition or has access to each other's code helped? I can't find any evidence either way, but something tells me that the KDE people study the GNOME code and vice versa. They may choose different ways to do things but in terms of getting it done -- sharing things works.
The same can and should apply to online applications - particularly because with an online application the software is no longer a commodity to be sold. Services are the commodities, as Tony Mobily says in his piece on the history of computing:
"The business model (and therefore the revenue source) will depend on the application, but one thing is sure: there is money to be made."
When you have to make money on what you are doing not what you have done the game changes. It's no longer who has the biggest feature list that counts: it's who can keep the site running and secure, attend to user queries and ensure the services scales well that will matter. Marketing will play it's part as always, so will establishing an early lead. People will often go with names they've heard of or that their friends recommend. But, the focus is now on service and not software provision so the old names everybody knew are pushed out by new ones. The new household names don't sell software they sell services (Facebook, Myspace, Google etc.). Even where the users don't pay for the service, the advertisers do and if the service is poor -- the users will leave and so will the advertisers. If the users lose trust in the provider, they'll leave -- and so will the revenue. Freeing the code we're all using would go a long way to keeping that trust. Part of that trust will be down to compatibility. As others have said, standards will become more important. If the users can't get their data out they will be more reluctant to put it in and they will tell others.
One last thing on this subject: does a model where it's service and not software that brings in the revenue sound familiar? It should do -- most of the successful free software providers operate on a similar model. Canonical (the people behind Ubuntu) mostly give their software away but they charge for professional support services. Sun offer support services not only for the paid-for StarOffice but also for it's free cousin OpenOffice.org. When the software is free, the service is what counts and then all those marketing claims will have to live up to something -- or else!
So this "the browser is the computer" thing: what's that about then? If the browser is to become the interface through which we access everything -- and it already has in some quarters -- what will this browser run on? So far I've not seen anybody suggest that the browser would run on anything other than an operating system of some kind. Even mobile browsers run on cut-down OSs. Android is much more than a browser, and the fact that it is free makes a big difference. Remember when the PC won because IBM allowed people to clone it? If a mobile provider needs software and one set has a licence fee but both are of equal standard: which one will they go for? Ask T-Mobile if you're not sure -- they've chosen Android and I suspect it's for more reasons than price alone.
So whatever shape and size they are, the devices of the future will need an OS and if the online applications require a browser and that's all -- why would a device manufacturer not want to use software which they could adapt, tweak and tailor for their users? A question was asked to a popular UK computing magazine recently. The magazine had run a feature on niche computing and had omitted GNU/Linux from the whole thing. "Why?" was the question that was asked. The answer? "Linux needs to become a reason for buying and not just a way of saving money.". They're right as well -- free software needs to be more than a cheap option. The problem is that it already is much more than that: you know it and so do I. The ones who don't know it yet are a majority of computer users and those providing their computing experience. With online applications, the licence does matter more than the cost. Imagine if an online provider had to get a software company to make changes for them: nothing would get done. In the online application world: everybody runs their own code and most of it will be built using and run on free software.
So here's the thing: I predict a greater shift to online applications by more people. No surprise there; but I also predict we'll be accessing these applications via a wider range of devices. Again nothing new there, but I wanted to say it. Finally I predict that more of these will be using free software. The days of proprietary software houses were numbered when GNU started to gain traction. GNU/Linux, Apache, Firefox, OpenOffice.org have all showed what can be done with free software. Google has added weight to this with Android and Chrome. On-line application providers operate under much tighter margins than traditional software providers. They need more than cheap software to run on, they need software that gives them the freedom to provide us with the services we demand.
So far from just having room for free software, the future could very well depend upon it.