Yesterday I was catching up on some magazine reading. Why else should you have a subscription, right?
PC-active, a Dutch PC magazine, in their July issue published an article about the importance openness for products and infrastructure in the computer industry. Trigger for this article was a presentation by Tristan Nitot of Mozilla Europe at the FOSDEM conference… in february of 2010. FOSDEM is an open source developers conference held yearly in Brussels, Belgium and quite an interesting one to put on your schedule. You can get the presentations at http://www.fosdem.org.
The presentation, according to pc-active, defended the openness of information technology and depicted closed source solutions as being dangerous. … I beg your pardon ?! DANGEROUS???
Openness in any circumstance, be it product development, human relations or information infrastructure, can have major boons. There is no denying that. Making the leap to defining closed solutions -in contrast- as downright dangerous however, for me seems like an Arnhem bridge debacle – i.e. a bridge too far.
Before we continue, let’s take a look at what we are talking about.
Hackability: the ability or possibility to customize a device or technology in a way not foreseen by the designer/inventor.
Further down in the article the example of telephones are given as being designed to remotely listen to opera, instead its omnipresent use now. A second example is that of the SMS as primarily intended by telecom operators to communicate the status of their network towards customers, and the third example is that of the WWW itselfs as being in origin a network for sharing and publishing scientific documents, making them easily available for consultation or even editing on other sites.
My question to you is, have these inventions been hacked, really? A telephone system is invented to transport an analog (audio) signal via a medium to a remote receiver that can interpret this communication. I do not see the hackability here, since basically the daily telephone conversations we are having now or the transmission of an opera house’s production is -in essence- the same use of the technology: transporting a signal over long distances. Secondly, SMS, is a technology to send short status messages to a remote party over a (GSM) network. That we are now using it daily for more than only receiving twitter-like status updates from our telecom provider, in my view, is not hackability. It is more the evolution of a product. Making creative use of it. Likewise with the final example of the world wide web. In essence it is still a technology to publish content. True, the nature of the content has changed enormously -from text only, to multimedia- but the basic idea is still the same: publish content so that others can read it or even edit it. Again product evolution, not hacking the product.
I’ll give a more accurate example of hackability : the good old commodore 64 had a video chip and an audio chip. They could be programmed to do either work: i.e. in the 80’s of last century there were a slew of computer hobbyist magazines that explained how you could program the c-64 to have no sound but let the audio chip help in rendering video or vice-versa.
Hackability is a change in functionality, not a customization. (Putting pink wallpaper in your room instead of white paint is not hacking, that’s customizing.)
Openness, when it suits, closed when needed.
Do not misunderstand me: I am all for openness in any designer or inventor related matter. Openness leads to public scrutiny and allows for more and maybe faster problem detection… and hopefully solutions to those problems. But it is not a quality assurance label, far from it.
My main issue with the article (and by consequence mr. Nitot’s presentation) is that none of the drawbacks or dangers of openness are addressed, while only the drawbacks or dangers of closed solutions are put into the spotlight. One of the main issues I have with the current state of so-called ‘open’ systems is consistency and stability. One of the main benefits of a non-apple personal computer is that you can customize the machine itself to your hearts desire. I love it, do it myself with vigor. However, one of the major drawbacks of this customizability is -according to accounts by microsoft itself- the interoperability of the multiverse of device drivers and the operating system. The most common cause of pc OS crashes are still device drivers. So I wonder, what is the actual value of being able to customize your system if the customization causes your system to become unreliable?
Besides, those that think a mac is not hackable: please go check out http://www.ifixit.com. You’ll find a rich compilation of step-by-step guides to customize your mac (or apple mobile device) to your heart’s content.
I do realize that closed solutions have the risk of making you producer-dependent. But let’s face it: in our daily life, what is really hackable? Your car? Hardly, except when you count tuning as hacking. In fact, in my humble opinion, it would be nice if the computer manufacturer industry would be more like the car industry. Make any PC closed. Make it customizable to a certain extent, with producer approved parts. But build in a stricter selection of hardware components so you’ll get a more robust and consistent platform. Will it be flawless? Of course not!? Ask Toyota 🙂 Will it be an improvement of quality? I think so.
The issue of Apple’s ‘censorship’ through iTunes is addressed as being evil. Well, for the most part it seems to me that Apple is protecting the satisfaction rating of their mobile devices customers. Suppose the censorship by Apple is not performed and you load crappy code onto your device and it breaks down to an unusable state? Is that Apple’s fault? No. Will they be blamed? Of course! It’s their product that crashes. Are there ethics involved? Naturally, and probably your ethical background can collide with that of Apple because your interests are different. It will be a never-ending discussion, and I am glad we can have such discussions, because it will keep Apple and other manufacturers on edge on what they do with their products.
When, then should things be open? When you talk standards. Communication, document formatting, hardware specifications, and the like are some examples of situations where openness is a must. Function descriptions should be open: i.e. What is a browser, what does it do? That should be an open specification. So that multiple parties can make their own (closed) implementation of it. So specifications and definitions should be as open as possible. The actual implementation of it can be closed.
In such a world, we are indeed producer dependent if we buy a certain product, but it will allow us to switch brands if our favorite make does not deliver. One of the conditions that have to be met in this model I am presenting, is free competition. You have to have multiple brands being able to compete with one another over their implementation of open specified products. Producers should still be able to patent or license their implementation, but not the specification.