The recent furore that erupted over Microsoft’s ‘shifting’ of focus for Silverlight and their somewhat wishy-washy attempts to clarify their position illustrates a fundamental disconnect between the Microsoft machine and the lives and careers of the developers that use their tools.
An enterprise CIO will cast his eyes over his domain and see a wide range of technologies in use and although he will use his enterprise architecture minions to herd technologies in some direction, he knows that there will be a lot of varying technologies being used – the decisions about which technologies to choose are complex and in many cases dictated by the vendor of the product chosen by the business. Likewise, a vendor technology strategist will build products that use many different technologies in order to support the varieties in the market and their big customers. Standing outside the technical trenches it is obvious that there is no preferred technology and besides, taking a bet on any particular technical bias is risky and best left to successors. It is difficult then for these people to realise or even understand the low level religious technical wars going on amongst the troops that are assembling one tiny piece over the overall solution.
But just because you can’t see it, it doesn’t mean that developers hedging their best on one or another technology does not exist. The ‘religious wars’ that we are familiar with, such as Java vs C#, Ruby vs Python and Lisp vs Common Sense are a ubiquitous testament to the dedication, passion and possibly blinding fanaticism that developers have for their particular chosen technology. The reason for this is quite simple, developing senior skills in any technology means that the average developer has to largely ignore others. Average to good developers (I am explicitly excluding ninjas) are unable to be considered masters at two wildly different programming languages, paradigms or frameworks. It is difficult enough as is, to become a master at a technology during a measly eight hour day, never mind going home and spending a single hour becoming at master at another.
While it may be sensible to have broad ‘generalist’ skills, and to a degree most good developers do, in the current employment market the best way to get a well paid job or contract is to have in-depth specialist skills on the technology being requested. A senior ASP.NET web developer used to leading teams and delivering complex solutions will find it difficult to be accepted to take the same role and for the same rate at a Java or Rails shop regardless of how much time has been spent as a Java or Rails hobbyist. Ultimately most developers (especially expensive Western developers) have a market value that is only as high as the value of their detailed technical skills from their last gig – regardless of years of other skills and experiences. And while we may blame the developers for their lack of foresight, we perpetuate this problem ourselves during interviews by asking specialist questions that only a developer with current, hands-on and detailed technical knowledge will be able to answer.
So the mistake that Microsoft made during PDC in seeming to support the ‘Silverlight is dead’ message by overplaying their plans for HTML5 was to ignore the developers that passionately believe that their medium term futures are dependant on Silverlight. Here in the City of London, it seems that Silverlight is gaining ground in financial services applications and many developers are whole hog, full time into Silverlight development (at very high contract rates I might add). Their value for their next gig will be determined by how much demand remains for the Silverlight skills that they have developed and a message about the death of Silverlight means that they will potentially hit the street as a junior to mid level web developer next time around. Microsoft, it seems, made assumptions about the fanatical support of their .NET developers without realising that that support is waning and their technology base is so broad that there are technology battles underway across Microsoft’s broad technology base.
I was interested to see Mike Taulty’s take (a well known UK Evangelist) in his post ‘Silverlight *versus* HTML5? Really?’ where he neatly lays out the argument for their being a place for various technologies but he makes the mistake of being a marketing person looking at the overall IT problem rather than as specialist developer. I especially like his comment ‘On the question of investment in Silverlight – yes, I’ve made that investment too.’ – with all due respect to Mike, the credentials used to find his next job will not be for his Silverlight coding skills, but for his awesome evangelist abilities.
An attraction to developers of more open source frameworks is that their demand is more natural and organic – not subject to marketing budgets, product positioning and acquisitions by mega corporations. For front end web skills, developers are comfortable putting a whole lot investment in JQuery because it is not subject to any vendors product focus and has a lifespan determined by developer support rather than how many developers are on the engineering teams at Redmond or how Microsoft intends to tackle the iPad problem.
In my twenty years in professional software development I have made a few major shifts in my preferred technology. It is difficult, frustrating and takes a while to get up to the same rate of delivery – but you get there in the end. So if a Silverlight developer were to take the huge investment to develop for HTML5, then why even bother going down the Microsoft route? Developers are finding that frameworks such as Rails and Django are fun to play with after hours, will get the same result and don’t have the risk that Microsoft will, yet again, cut their value at the whim of this financial years’ marketing objectives.