IDC, the same company that predicted an overwhelming start for Windows Vista [pdf], is predicting that Microsoft will Sell 30 million units of their phone OS next year. Originally taken to mean Windows Phone 7, Microsoft later issued a correction to state that the number reflected all of Windows Mobile OS including version 6.5 which is currently in the market. Actually, being laughed out of the market would be a better description.

Microsoft Windows mobile operating systems for smartphones market share is in free-fall having dropped from nearly 20% in the quarter ending October 2009 to 6.9% for the quarter ending March 2010. It's losing 50% of its share every six months and the rate is accelerating. At a flat rate that's 1.5% share in 1Q2011. Including the accelerating rate of decline gives somewhere south of 0.7%. Negligible. Fringe. 0.7% is in a projected 500M unit market is 3.5 million units which at $15 a pop doesn't even start to pay for the cost of development, let alone any marketing money. And that's if they launch an awesome product in 1Q2011, which seems unlikely. 3Q brings 1/3 of that and past then it's zeros all the way down.

Microsoft is having IDC float these ridiculous numbers so it can push phone vendors and carriers into line - into giving up control of their user experience, into giving Windows Phone 7 preference over competing solutions, into giving up margin to get share. The message is simple: "You dare not ignore the overwhelming market power of Microsoft. You dare not argue. It is useless to resist us." The truth could not be more clear, nor more different. Microsoft is not in charge here - they're not even relevant with a <7% share. By giving up control of the user experience the phone vendors give up the differentiation that is the value they can add - their work that makes their platforms special to them is the thing that turns their product from a generic commodity to an object of desire. By giving up their margins they gain nothing but the opportunity to lose even more money - which is not their goal. By playing along with this the network providers likewise give up everything that makes them special. The smart executive will tell his Microsoft weasel: "You told me that story a year ago. Come back when you have product." The dumb executive will swallow this hook, line and sinker and fire up the co-marketing scheme without proof of product so keep an eye out for stocks to short because dumb executives run their companies into the ground.

Meanwhile, as you're all well aware Microsoft Friday reorganized its entertainment division. It's not an accident this happened on the Friday leading into a US holiday weekend any more than the announcement that the BP "Top Kill" failure was. What you may not know is that this is part of an escalating sequence of reorganizations that started in Windows Mobile nearly two years ago - and it's the third or fourth of such. Each time the reorganization goes higher up the tree and now this iteration has CEO Steve Ballmer directly in charge. Obviously there would be no need for a top-level reorganization if the product were on track, so expect the schedule to slip past Christmas and into next year or alternatively for the product to suck so hard that the level of vacuum is scientifically interesting.

Next year is too late. Nothing less than a perfect product will do, and even that is probably not enough. Nobody's going to believe a Microsoft funded IDC report in the face of actual shipping products with huge margins - not anybody who wants to survive anyway. Mobile is different. Two years is forever in mobile. A year from now to start to try to gain traction on a new OS and find developers willing to take a risk on the the continuous delays is just impossible.

In short it's game over for Windows Mobile and not too long thereafter for Windows as well. You see, there are these tablets...

This is a work of fiction - as all my slashdot coments are. It infers facts not in evidence. It describes inferred historical events. Dates are not precise. The bulk of it is however my firmly held opinion given considerable experience and study in the field. It is sourced from a /. comment previously written and I'm replicating and revising it in my journal both so I can find it more easily and so my slashdot friends can see it even though the article must fade from the main page:

Normally I'm not one to praise Microsoft's end results, but I'm not stupid. They hire the brightest minds from the best schools with strong foundations in classical IT art as well as contemporary vision and they work them to death because that hazy zone between exhaustion and physical failure is a special point where human brains integrate at miraculous levels. Microsoft has known this for twenty years and organizes its workers accordingly. These folks driven in this way can make an awesome mobile OS, they did, and I'd love to have a copy of the source for that bad boy. These Microsoft developers made a rock solid performant and genuinely innovative phone OS which is the core of Windows Phone 7. It's tiny, boots fast, suspends and resumes instantly, and pinches ergs like they're made of platinum. It has an intuitive touch-centric interface. It works flawlessly with all the latest technologies - hell, it'd make a great HPC OS if these jerks would think out of the box now and then. This was about two years and three reorgs ago. This is the mockup they'll trot out to the major phone vendors hoping to get them to push the platform - short a few apps but you can see the potential because it's beautiful, intuitive, responsive. They built an app store for it, and shopped the mockup around to app developers under NDA. Some of the AC posters here even have it and they're in awe of its incredible flexibility, its power, its potential - and they should be because this bare OS rocks. They built apps for that were complete a year ago, and they're continuously refining them in the hopes that the platform will launch someday. Microsoft floats an early 2009 release date to some of their preferred pundits even though it's not finished yet because that's how you feed a flackalyst.

It's a killer mobile OS but it's not a Windows yet. For six months they put some finishing touches on the version they intend to ship - integrating Bing search and Windows Live services into everything, building the Mobile Office apps for it, porting Silverlight, .NET for mobile and a bunch of other stuff. This is leveraging the platform so that it pushes all of the other Microsoft platforms because making products that can be extracted from their internicine application and server dependencies is not the One Microsoft Way. The shipping version then ran like a dog, leaked memory like a seive and crashed every few minutes. So eighteen months ago they rebooted the team and tried again. They got the same result, so nine months ago they reorg the group from higher up and try again. The new group can't get the thing to work right in nine months, so yesterday they reorg the entire entertainment and mobile division to be directly under Steve Ballmer and reboot their efforts yet again. This product was supposed to ship in early 2009. It is not even close to ready. It probably never will be because all of these internicine ties never did work well, are a moving target, and have reached an insurmountable level of complexity for a mobile device which must by definition be the ultimate in computer reliability and stability while remaining cutting edge in a dynamic market. It literally can't be done.

Even today Microsoft executives are shopping around that slick bare mockup that no end user is ever going to see to their phone partners at the manufacturers and carriers, playing the push/pull game. "You want this. You need this. You're going to want to start planning the marketing around this product right away. This is going to be a slam dunk! And look - it says Microsoft on it everywhere so you know businesses are going to eat it up. [Hushed]It has IE and Outlook." / "Of course, this iPhone killer isn't for everybody - it's exclusive to our special friends. Committed friends like you who know the value in long term relationships with a powerful member of the digital ecology. Partners who are willing to sacrifice any control whatever of the end-user experience. Existing paid-in Windows Mobile deals don't count because this isn't Windows Mobile - it's Windows Phone. For something this great we need a new level of commitment and we're not offering it to folks who care so little about their customers that they'll push that Android thing on them."

The whole time these Microsoft account execs know the real thing with a real app load bears as much relation to the thing they're showing as you do to a squid. They know the intended shippable version can't run at all on equipment with a BOM less than $1000 and even at that crashes several times an hour. They probably took meetings today pushing this thing, knowing that the latest reorg means there's no hope of a shipping product that works well in the next year. They don't know that in that year the target will shift again, pushing their target date out another year as the team refactors yet again. Even if they knew, they don't care. They're willing to believe the code monkeys can make it work because they make $400K + options, live in nice homes, and are going to milk this cow for all it's worth. If they had scruples they'd never have passed the first interview.

Meanwhile the phone partners, the carriers and the PC partners have wised up. They have to sell product today. They can't wait forever for Microsoft to pull their collective heads out the the dark place they're at and deliver a product to compete with the iPhone and the iPad and the iPod if they want to stay in business and WiMo 6.5 just won't do. Those that bought the "Real Soon Now" story for over a year lost serious opportunity to those more skeptical. In this business the two+ years from intended WP7 release in early 2009 to a current hope of mid 2011 really is forever. In one year Android has gone from "Android what?" to 28% share of sales and the growth arc if continued has total mobile domination by mid 2011. This is actually good news for all of us because once vendors let go of the "wait for WP7" idea they don't have to be Microsoft's special friend any more and can instead focus on giving us what we want right now: Progress. So HP buys Palm, Dell offers Android Phones and slates, all carriers that can't get iPhone like AT&T (who have an iPhone exclusive in the US) push Android like it's the Holy Grail and even AT&T carries Android because they must offer choice and there's only one iPhone.

We have one more Windows Mobile reorganization left to go. This is the one that knocks monkeyboy off his perch at the top of the tree. The guy who replaces him will know better than to let this cursed project live a single day. This train came off the rails years ago, and we're now only still watching the crash in slow motion because these things take time - but the outcome will be as certain that day as it is today and as it was two years ago. It's got fail written all over it.

Hidden in this post is a secret message to Microsoft on how to fix their stuff - as my posts often do. I've been doing that for years. I do that because it amuses me to taunt their employees who post here about their obviously predictable failures by linking to my old posts. I don't want Microsoft to win because they have a long history of being evil, but it amuses me to tell them how to win knowing that they'll ignore me so that I can later point out how they could have won if they weren't such fools. So far this is working and the outcome is to me hilarious. Everybody needs a hobby and this is mine.

For a long time now, Apple's total value in the market, or market capitalization, has grown relative to Microsoft. In April it was reported that Apple has passed Microsoft on the S&P 500, but many people pointed out that this was an adjusted figure, not raw market cap.

The difference between them has shrunk from over $100 Billion to only $3.1 Billion today. Given the rate of change it's reasonable to expect that Apple will pass Microsoft in this key metric before the end of June, perhaps as soon as next week. Possibly even tomorrow.

This is important because Microsoft's influence in the maketplace is disproportionate to the value of their products. It's driven largely by a long history of fear, a perception of invincibility. That another company can rise up and topple them from the height of America's largest tech company should have remarkable impacts on all the other tech companies. We may see more innovation because of this.

eWeek is reporting that HP's Windows 7 based tablet PC, demonstrated by Steve Ballmer at CES is, contrary to previous reports, alive but delayed to October. Other reports say that it will have WebOS from HP's recent Palm acquisition. Over on CNET, Erica Ogg isn't calling it either way.

HP's own site still has promo up. Certainly they are pushing Philip McKinney as a personality with vision.

So... An HP tablet is coming in October. Will it have W7? Maybe. Will it have WebOS? Maybe. Will you be able to choose from these? It seems unlikely given the history. Compaq showed a Windows tablet at Comdex in 2001. Nothing came of it, nor did anything come of the Compaq/HP tablets that followed. All of those came with Windows.

If it arrives, it's late. By then there will be 9 million iPads in the market, and dozens of Android on ARM alternatives to compare it to. Will it be a compelling product? We'll see. HP definitely has the manufacturing chops to pump out a bunch of these. They have first rate engineers. But will their partnership commitments result in a compromise product that's lackluster in performance, has an interface that's not suitable for the form factor, requires the requisite and power-sucking Windows antimalware suite? It's possible, especially if HP gets a really sweet W7/Server 2008 licensing deal in return. Microsoft can be really persuasive, and they have a large number of "special relationships" available to sweeten a deal. Whether they can sweeten a deal to be worth more than the $1B HP paid for Palm remains to be seen. To me it would be a shame if the net result of HP buying Palm was to kill WebOS in order to get a better deal on Windows. Palm was better than that even though they made some business missteps recently and found themselves cash poor at an unfortunate moment.

Clearly there are some CEO level executive negotiations going on between Microsoft and HP. Hopefully these will become open on Groklaw one day. For now we have to wait and see if HP is ready to compete in the new world. My guess: if they come out with a W7 only tablet it will suffer the same fate as all their other Windows tablets - unit sales that peak in the tens of thousands. If they give us a choice between Android, WebOS and W7, they've got a chance of making something that catches the wave - a chance to be a part of the new people-centric consumer electronics world. If they fail it, well, there's always all those other vendors who are willing to give us what we want.

We shall see.

A couple months ago my Achievements page clicked over from 32 to 64 +5 moderated comments. I decided to try to increment the +5 moderated comments counter on my Achievements page again with 64 more +5 comments. For NYCL or Bruce Perens this is a day's work but it took me a good while. Today I managed to achieve this goal and arrived at 2^7 comments moderated +5.

To commemorate the occasion, here it is. Lovely. I would have preferred this one. To my credit, most of the posts were going for insightful or informative and were modded correctly. In the whole time unfair moderation was very rare, and easily countered with a followup post at +2 bonus.

It's not very hard to do. Post at 1. If you post at 3 with karma and subscriber points and your first moderation is funny or insightful which most people default to +1, then the post looks like it's +5 moderated to the moderators, but (I think) those bonus points don't count - it has to have four legitimate up-moderations. Say the usual stuff. It helps to post early in a thread, and best to get the first one if you can do it and not get modded to oblivion.

Along the way I had one post that was so controversial it was modded over 30 times before sticking short of +5 but well in the visible range. I wish I could remember which one that was.

At one point I was getting so many mod points it was scary. I think I only used them up once.

Karma is neither cumulative nor persistent, or now I could be a real jerk and get away with it. That's probably for the best.

I guess my rambling point is that moderation does seem to work. I guess since it gave me the incentive to do this, maybe the achievements thing works too.

The new page design is still broken, and the home page now reports that it is too large to load into my BlackBerry so I'm down to slashdotting about 1/2 as much.

Was it really only nine months ago that Fusion-IO launched their IODrive(pdf)? Since then they've gotten VC funding and made strides in new product (pdf).

For those who aren't following along, they make an Flash memory data storage product that looks like a 320 GB drive, but rather than connecting through legacy SATA interface it connects through the much faster PCI Express (PCIe) interface. By using RAID techniques to "stripe" the data across multiple flash chips, an abstraction layer to hide failed memory cells and intelligent logic to provide wear levelling they take fallible flash chips and turn them into reliable static storage that has performance characteristics more familiar in RAM. Their first version pulled in over $10K for a 320GB card.

They weren't the first out the door with this technology, and they weren't the last. Now OCZ has stepped up with a PCIe flash storage device that offers up to 1TB of storage in one PCIe slot and delivers it at a claimed 500GB/s. Now PhotoFast has one that does the same at 1GB/s for under $5k. That's not 1Gb/s. It's really one billion bytes per second.

Combine this with the new Nehalem technology available in platforms like the HP DL370 G6 with NINE PCIe G2 slots. Add free SAN software with HA clustering, unlimited snapshots, unlimited storage and other popular features like openfiler. Mix in a little Infiniband QDR.

Now you can have for under $60k a box that delivers 6TB of storage that does >1 million IOPS and can deliver that at about 96Gbit/s, in 4U. For another $5k you can get a MDS 600 5U to attach to that box that holds 75 of these 1.5TB 3.5" drives. Add 75 of this Drive sled and you've got over 100TB of slower storage at 3Gbps for an upgrade cost of under $20K. Even with 24/7 unlimited systems enterprise level support for five years you're looking at less than $100K. And it scales to infinity.

So here we are, with 6TB of insanely fast storage and 100TB of nearline storage, HA, thin provisioning, iSCSI and remote admin delivered at insane bandwidth. In 9U, for under $100k and burning less than 1000W, with unlimited LTU and unlimited support converged with current network architecture. And the speed and performance of the flash devices is more than doubling every nine months at the same time as the price goes down by more than half.

8Gbps FC San just got here. There isn't even a FCoE standard approved yet and when it is, it's not as fast as this by an order of magnitude.

Is the classic FC SAN dead?

Toshiba will show a 0.5T SSD in a 2.5" form factor at CES according to CNET. The drive is expected to cost $1,652 in "sample quantities". The drives are reported to perform 240MBps max sequential read and 200MBps max sequential write using MLC flash memory. IOPS information is not yet available.

It may be time to recant my bias toward SLC in relation to MLC, at least in terms of laptop drives. The preference IRT database cache and server drives will have to wait until full benchmarks are available. They seem to have sufficiently exploited the benefits of parallel ICs to defeat the performance advantage of the underlying technologies.

The interface is not specified in the article but it's probably SATA 2.0. The SAS version is probably 3-5 years out, as is usual.

SO just found this game. It's been what, three days, and already there's a shoe throwing Flash game. The responsiveness of modern culture continues to amaze.

As of now the top countries shoeing the US president are The United States, France, Australia, United Arab Emirates, Saudi Arabia, Turkey, Egypt, United Kingdom, Germany and Pakistan.

I usually give positive moderations lots of slack - I may think a joke or an article is lame, but if somebody wants to moderate it Insightful or Funny, fine - I'll mark those moderations as Unfair/Unfunny if they're promoting obvious trolls or whatever, but that's not common.

Negative moderations are usually obvious also, but the one that I don't give much slack is "Redundant". If an article really was duplicating existing content at the time it was written, or is just adding a content-free me-too, then it's redundant, but if it's a +1 article written two minutes after the main slashdot article, and somebody posts something similar but much more insightful an hour later that makes it up to +5, the first one is still Not Redundant. Maybe it's Overrated, maybe it's Flamebait, and I'd let those moderations through, but I'll call a "Redundant" as "Unfair" if it wasn't redundant enough.

"Flamebait" gets a lot more slack - sometimes there are articles that I strongly agree with (even if I've written them myself :-) that are aggressive enough that they get Flamebait, and I'll usually let those stand - but I try to ding any moderations where the moderator's calling something Flamebait just because they disagree with it.

The one meta-moderation I have trouble with is when somebody rates something as "Funny" that looks like it was intended to be serious, not funny (and wasn't accidentally funny either.) Does marking the moderation Unfair undo the moderation, decreasing the posting's status? Or does it just ding the moderator's karma, which is fine...?

I have a section for "The Book of Sean", my brain dump onto the web. I also have photos from my various nerdy projects.

ok, so it needs a body.

Blogging is lame enough - it's a way of nagging people who don't write their own HTML into at least writing text and links. Guess I can't flame them too much, given that I've done almost no edits to my web page in years :-) So here *I* am, not even getting my own Blogging software for my web pages, much less writing it myself - I'm just using /.s.

• Designate a master in the /etc/hosts file
• Create an xterm init script that basically calls: X -query master
• Strip down runlevel 4. (Including xfs)
• Add xterm to runlevel 4.
• Tweak the /etc/X11/XF86Config file. Change FontPath to FontPath="tcp/master:7100"

On the server, tweak /etc/X11/gdm/gdm.conf (next journal entry.)