Politics or bugs behind Microsoft exec massacre?
Blood on the carpet
Ballmer has the HP-Compaq media circus to thank for distracting attention from the most recent Redmond reshuffle, which leaked out on Monday.
In Microsoft's mobile division, reorgs seem to take place with great regularity, and it's only three months since Ballmer streamlined the troubled phone side of the operation by ensuring Ben Waldman and Juha Christensen reported directly to Fester for head-to-head bawling-out sessions. And only a month since Pocket PC evangelist Derek Brown was promoted from the liaison work that made him a hit with user groups into Highly Important Obscurity.
But that shouldn't obscure the fact that the Chief Executive Klaxon has just carried out his biggest purge yet of his senior Microsoft executives, and this St Valentine's Massacre sees senior management at the Mobility Group comprehensively routed.
As we noted on Monday, Ben Waldman is dispatched to the Redmond piranha tank, following Phil Holden and Rogers Weed out of the Mobility Group. Holden and Weed have new roles elsewhere at The Beast, and Waldman's taking a year's leave.
Being declared "likeable" by Register editor John Lettice at CeBIT (GSM World, you dolt - Ed. And I liked Brad Silverberg back in 1994, which may be significant, see below) probably accounted for Ben, and it's made us think twice about letting you know that affable Microsofties exist.
Kevin Shields (not the founder of shoegazer rockers My Bloody Valentine, but a five-year Windows CE veteran) becomes the GM of the Mobility Group, reporting to Pieter Knook. Juha hangs on in there, still reporting directly to the Klax.
So what gives?
Little PC of Horrors
Well the Pocket PC 2002 release in October has proved extremely buggy and could be a candidate for the flakiest software to come out of Redmond since DOS 4.0, which shows how old we are. Don't take our word for it, however, but check out this chamber of horrors.
" The calculator application reports itself as 'caluclator' when you check the apps that are running" is one uh, feature that hasn't been resolved by a hurried PPC2002 Service Pack.
Is another open issue, "If the user tries to enter a search string in the FIND field, program aborts to the today screen. Soft reset doesn't help."
And this won't help much with those software downloads:-
"You cannot access FTP sites (either through IE or through WiniNet) that require username/password… always transmits anonymous now on Pocket PC 2002" runs another unresolved bug.
Pocket PC 2002 is backed up by an excellent marketing campaign, "Software Matters", which shows glitzy CE devices alongside, impoverished, Eastern Bloc-style monochrome Palms. But it doesn't matter how flashy the product is if it's fundamentally flawed. Maybe the slogan should have been "Bugs Matter", for in addition to the showstoppers, basic PIM and messaging errors are bedevilling users.
The runes point to Waldman rushing code to market, and the number of additional hardware-specific flaws suggest that the hardware OEMs (HP, Casio, Compaq and Toshiba) had insufficient time to perform testing and integration.
All of which ought to give Microsoft management cause for concern. PocketPC boasts far richer functionality than anything Palm can offer right now, and in the lull before it finds itself trumped by the slickly integrated communicators from the likes of Nokia, Ericsson and Sony, Microsoft ought to be capitalizing on the good reviews it generated with PPC2000. As it is, however, Palm tenaciously clings onto its PDA market share (in the sense that it's still considered a market).
With our mighty "caluclator", we shall slay Nokia
But we detect another reason for the slayings, and it's one that will be familiar to seasoned Redmond watchers. Recall that CE is now CE.NET, and with the platform reclassified as a strategic client for .NET web services.
The way that Microsoft has delivered code for handhelds works like this. The CE team delivers a milestone, and the clients - the Pocket PC team, the embedded division and the Mira team, for example - go off and tailor the code base for a specific device class. While the 2000 release looked very much in step with the underlying CE code, 2002 looks like a mutant version hacked together in a hurry. This we surmise from the fundamental bugs in the Internet stack (it reverses DNS) and manifested in the user level applications we detail above.
But rather than trying to tailor something acceptable to the market, the Pocket PC division appears to have caught .NET influenza, with management decreeing that it focusses on supporting the 'smart clients'.
With the Microsoft Smartphone 2002 still poisonous in the marketplace (the Beast appears to have given up signing any of the major Tier 1 vendors), and even loyal Microsoft customers phone like Samsung hedging their bets [Samsung confirmed our exclusive here], the company has reverted to stitching up deals directly with the carriers.
But the focus on the nebulous .NET marketicture isn't going to play well with loyal Microsoft devotees, who quite reasonably, want two things: a killer PDA and a killer phone. Having both a poor PDA and a poor phone on which to run "smart clients" or other .NET services isn't going to satisfy anyone but Bill, or at least not until there's a web services market mature enough to offer real value, and that's not going to happen anytime soon.
We're reminded of one of favorite trial memos, Brad Silverberg's pissy emails from billg" missive, which remains a pivotal document. Silverberg wrote "i simply do not want to spend my life in meetings struggling with the internal issues, getting pissy mail from billg or hearing from people who want me to do unnatural and losing things to 'protect' Windows."
It's an important reminder why mediocrity runs so deep in Microsoft's DNA, and why Redmond almost seems predetermined to produce poor software despite an abundance of so many gifted and motivated staff. When good stuff emerges - like Heljsberg's C# and CLR - it's almost always the exception. Compromises are the order of the day, and with the dead hand of the .NET boys ruling the roost now, we expect very little to get excited about from the Mobility Group for the forseeable future. Wake us up when something interesting happens, please readers. ®