Sun’s second thoughts about Open Source Solaris
2000-06-30
Sun is developing a bit of a reputation for raising expectations to the highest levels then sheepishly admitting it cannot carry them through. It happened with the standardisation of Java, when Sun lodged the Java specification with standards body ECMA before withdrawing its application a few short months later. It looks likely to happen with Jini, the much-touted plug-and-play standard for devices which has thus far failed to deliver a product. And last week, it was the turn of Sun’s open source initiative, on the rocks only five months after it was announced.
What’s the problem? In this case, it seems to concern the scale of the task. There are nearly ten million lines of code in Solaris , and Sun did not feel it could just plonk it onto the mass market without “making it user friendly” first. It is understandable that massaging ten million lines of code would take a while, but a couple of questions remain.
The first is – what is so unfriendly about the code? Fair enough if this equates to adding a few copyright notices and generating an HTML code browser (à la Java). Less reasonable is if the code needs restructuring or additional comments to make it readable, or worse – are there parts of the code that are currently too convoluted to see? This brings us to the second question – surely a plan was put in place in January, so what has changed to delay the plan? It may be that, on investigation, Sun realised that the problems with the Solaris code were going to take longer to solve than they expected. Now, it may be that this is an attempt to overstate the issue – us commentators are always looking for a wound to rub salt in. But the fact remains that Sun is having to lower expectations about when and how it releases its code to the open community. Probably the biggest issue is this: Sun has realised that, given the profile of Solaris, there will be a horde of points-scorers ready to identify weaknesses, bugs, design faults and security holes in the code and Sun must do everything in its power to minimise any risks.
Sun’s code giveaway will come, but more slowly than was previously announced. Sun are likely to hit their Q3 target, but only for certain sections of the code. According to Anil Gadre, Sun’s vice president and general manager for Solaris software, “the other thing we are finding out is that maybe people actually wanted certain parts and not the whole thing.” Sun is therefor considering releasing the code in a piecemeal fashion. Gadre’s remarks sound a little like Sun is hedging its bets (use of the word “maybe” is the giveaway), and will be a good fig leaf if the time comes to stagger deliveries of the code. Sun will release the code sooner or later, but more with a whimper than a bang.
(First published 30 June 2000)