Please welcome the XAML platform team to Windows!

image

I got word of a leaked email early yesterday that confirmed what I had been told in passing gossip – the XAML team being disbanded.

This morning I awoke to not just one email but five of them from my various sources all attached with mixed opinions on what it means. Here is my famous (internally in Microsoft, I would send these style of emails about Adobe and their competitive threats) "What Just Happened" response.

We’re pleased to announce the transition of the XAML
platform team from the Developer Division to the Windows team.  While the
team has been working side-by-side with the Windows team for the entire project,
this step brings them into our team formally. 

It is time to start moving the battleships into the attack formation. In that time to start the consolidation into the new ux platform we are about to remake again.

On the upside it means you have a consolidated outcome likely to hit our hard-drives in the next 2-3 years on the downside you have what I would call a technology freeze in effect. If the new iteration of WPF or Silverlight does not comply with the vNext vision, do not hold your breath for a new announcement anytime soon that does not involve Windows 8 future(s).

The team will continue their work on Windows 8 as planned
and will join our Developer Experience (DEVX) team. This transition allows us
to bring together our platform development team in a single-management
structure.

That doesn’t sound to bad, I mean on the surface its just a single management restructure. A day in the life of a Microsoftee where every fiscal year or often more than once you are given new managers because the strategy – scratch that – tactics have changed. That in itself is probably your biggest hint of all around the word commitment, this is not just a case of waiting for a restructure to occur once every 5 years – it happens often.

To clarify, do you keep swapping your generals around in war to the point where the troops effectively stop caring who they are reporting to? Probably not a smart idea but nonetheless.

The dev, test, and pm leaders who will be leading the
team reporting to AlesH, YvesN, and LindaAv are:

• Sujal Parikh, Development Manager 
• Eduardo Leal-Tostado, Test Manager 
• Joe Stegman, Group Program Manager

The leads and individuals joining our team are receiving
this mail and have received communication on next steps.

If most of you who have been involved in the Silverlight ethos are reading then the name, Joe Stegman will probably stand out the most. Joe’s background in the .NET space goes back a ways but in the end Joe’s really been one of the guys under the hierarchy crust of commitment pledges keeping things in the development side of things in check. Officiating his role further in this equation for me is a bit puzzling as it’s kind of the same thing different org tree?

These changes in leadership and organization are
effective today.  For the purposes of finishing out the fiscal year and
the performance review process the team will operate under the existing
management structure.

That is a swift maneuver. Nothing surprising though.

Now onto Soma’s email to the troops which kicked the previous email off (Notice how VP’s etc all pile on from one another with "what he said was.." like somehow being apart of the thread is being seen as a role of importance. Classic Microsoft Victory Email formula, just once I’d like them to send out just one email outlining the change. This is what I’m talking about when I say Microsoft Culture is retarded.

I digress.

MICROSOFT CONFIDENTIAL

Over the last couple of years, our Client and Mobile team
has done a fantastic job of building a number of XAML related technologies that
have been a huge value add to the Microsoft client platforms and an
instrumental part of delighting our developer customers.  The agility and
customer focus that the team has demonstrated over the years has been a
pleasure to watch.

Ooops. Btw this was supposed to be confidential. So do not tell anybody.

Over the last couple of years? From memory, I recall WPF being around for more than a couple of years? Never mind, I keep forgetting everyone internally has forgotten about WPF.

Soma is kind of saying, "thanks for the hard work troops, you managed to outpace most products in Microsoft with your constant brilliance around the word agility, which being said here comes the but to that placating statement.

Today, we are making some organization changes to bring
our platform technologies under a single management structure.  These
changes are centered around three focus areas:

• The team working on XAML technologies for Windows will
move to Windows.

• The team working on XAML technologies for Windows
Phone, Xbox and browser plugin will move to Windows Phone. 

• The Client and Mobile tools teams, including Windows
Phone tools and XAML tools, will stay in DevDiv.

These changes are all effective immediately.  From a
performance review perspective, we will do this year’s performance review under
the DevDiv organization model.

"Today marks a new day troops, for we storm a new beach" is kind of the response to that next piece.

Firstly you have parts of the XAML team(s) parked inside the Windows organization. First impressions on twitter are that "Way to go! Means XAML and Windows are finally going to get along and create awesome XAML experiences"

Have I not shown you the Annie video? Moreover, have I discussed the Orphan Syndrome? "My dad’s going to come for me, he’s rich you…you..you just wait and see"

To me that read as being a case of cherry picking parts of the team to socket into the windows division and their new coding charter will come next. If it involves XAML it will be based around what XAML vNext is likely to be – HTML5 meets Jupiter.

Same with the Phone team, it is what I would call "please standby for further orders" moments.

As for the tooling teams, well you got Cider and Expression Blend team is what that really comes back to. Given most of the Client employees have left, I am not sure what that means suffice to say I am not holding out for a new release for Sketchflow for starters and I am guessing that the Blend teams are not exactly getting high fives for poor sales and download rates to date. If I were in that team, I would be updating my LinkedIn account quickly.

I want to thank Kevin Gallo and the team for all the
great work that they have done over the years.  Moving forward, I’m very
excited to bring the client platform efforts closer to the platform
teams.  There is a lot of very exciting and critical work underway as part
of our next wave of platform releases and I am very eagerly looking forward to
seeing the team’s work in the hands of our developers and customers.

Remember when Mary Jo posted a while back on how Kevin Gallo would be taking over the reins of Scott Guthrie. That’s probably the quickest promotion and I’m not ready to say demotion but I’m not ready to say continuance either – that I’ve seen?

The positive part there is the "next wave of platform releases" that sounds a lot like a continuation of what we have in front of us. Make no mistake there will be a Silverlight 5 and a WPF vNext released next fiscal, its already got most of the code done and it would be foolish to not release those when they can – especially after MIX2011.

Releasing those two would also buy you time for the next 2 fiscals at most. As by doing this you create this calming effect around "see, we’re still working on it..honest" to placate the developer hordes.

That is up to you, you can buy into that sure, and it is hard to debunk given there is not much visibility behind what we are likely to see next – especially given this is tradition within the Microsoft roadmap(s).

For me personally, I’d like to corner Microsoft If I could into giving more concrete assurances that whatever the next wave of bets are that they are either backwards compliance or have parity around what we have today in terms of conceptual features today.

Features for me are not will I be able to still hit F5 without changing code. Features for me are the concepts that are on the table today, around how one manages the out of browser and in-browser functionality – everything from casual gaming through to enterprise ready features (printing, isolated storage, data binding etc).

Going forward.

I think what has happened in this email is the equivalent of me saying "I really like this car, now can we take the wheels and put them over there. In addition, can you take the steering wheel and dashboard and put it over there. Lastly, can you take the engine and well. Just leave it in place for second; I’ll get back to you later on where we can put that next"

It’s clear there is a consolidation happening that I think we can all agree on early. How will the consolidation impact the average .NET developer is likely to be dramatic enough to warrant some applications having to have code refactored down the track – you will not escape that sorry.

Does this mean .NET is dead? Who actually knows what .NET vNext will be so it is hard to simply say "yes" and it could very well be a reset of .NET to fix a lot of pent up frustration in the way it sticks together today.

What I am certain of is WPF is definitely officially done. The chance of WPF going beyond what it is today is slim. Some journalists etc. will gloss over this as its not news but let me be clear in saying at Microsoft we really had no clue just how deeply seeded this product became.

In Australia it’s used quite heavily and it’s something I personally noticed whenever I used to travel around the country meeting Microsoft customers (both as an Evangelist and Product Manager). I used to send emails internally stating "I think we underestimated is usage, as it definitely appears to have more devs using than Silverlight" which was later brushed aside as being "Not realistic".

I think post September the announcements that are to follow will give these warnings probably some second thoughts around what parking WPF in the retired bay is likely to mean for Microsoft when it comes to the words "trust" and "commitment"

The product and developer satisfaction surveys for the last few years haven’t been something you’d brag about internally which for me indicates a strong sense of "fatigue" within the ranks of our beloved .NET developer communities.

It’s now one thing to announce what the next version of .NET will be its entirely different thing to convince and sell these fatigue customers that this is defiantly the bet this time. Silverlight, WPF and WinForm are available today and millions are shipping software solutions using them.

Microsoft now has to figure out a way to convince the millions that the "Windows 8" wave of vNext will fix all of these problems and more – and – will not require a reduction in feature parity along with extra boost in tooling.

If I know, my Microsoft and I like to think I do, good luck J

Full email below:

From: Julie Larson-Green
Sent: Monday, June 20, 2011 9:35 AM
To: Grant George; Jon DeVaan; Julie Larson-Green; John
Cable; Yves Neyrand; Craig Fleischman; Bambo C. Sofola; Scott Herrboldt; Greg
Chapman; Julie Bennett; Jeff Johnson; Ales Holecek; Mohammed El-Gammal; Chuck
Chan; Michael Fortin; Eric Traut; Jensen Harris; Linda Averett; Alex Simons
(WINDOWS); Gabriel Aul; Dennis Flanagan; Iain McDonald; Samuel Moreau; Dean Hachamovitch;
Michael Angiulo; Antoine Leblond; Tami Reller; Chris Jones (WINDOWS LIVE);
Jonathan Wiedemann; Ulrike Irmler; Adrianna Burrows
Cc: XAML Team; Kevin Gallo; S. Somasegar; Terry Myerson;
Sharman Mailloux Sosa; Brad Fringer; Steven Sinofsky
Subject: Please welcome the XAML platform team to
Windows!

We're pleased to announce the transition of the XAML
platform team from the Developer Division to the Windows team.  While the
team has been working side-by-side with the Windows team for the entire project,
this step brings them into our team formally.  

The team will continue their work on Windows 8 as planned
and will join our Developer Experience (DEVX) team. This transition allows us
to bring together our platform development team in a single-management
structure. 
The dev, test, and pm leaders who will be leading the
team reporting to AlesH, YvesN, and LindaAv are:

• Sujal Parikh, Development Manager 
• Eduardo Leal-Tostado, Test Manager 
• Joe Stegman, Group Program Manager 
The leads and individuals joining our team are receiving
this mail and have received communication on next steps.
 
These changes in leadership and organization are
effective today.  For the purposes of finishing out the fiscal year and
the performance review process the team will operate under the existing
management structure.

There will be an informal Q&A session today to
welcome everyone and answer any questions that folks might have.
• XAML team welcome – 2:00-3:00 in building 37/1701

Please join me in welcoming these folks to our
organization!
Julie
 

 

From: S. Somasegar 
Sent: Monday, June 20, 2011 9:16 AM
To: Client and Mobile Team
Cc: Developer Division FTE; Steven Sinofsky; Julie
Larson-Green; Terry Myerson; David Treadwell
Subject: Bringing together client platform efforts

MICROSOFT CONFIDENTIAL

Over the last couple of years, our Client and Mobile team
has done a fantastic job of building a number of XAML related technologies that
have been a huge value add to the Microsoft client platforms and an
instrumental part of delighting our developer customers.  The agility and
customer focus that the team has demonstrated over the years has been a
pleasure to watch.  

Today, we are making some organization changes to bring
our platform technologies under a single management structure.  These
changes are centered around three focus areas:
• The team working on XAML technologies for Windows will
move to Windows.
• The team working on XAML technologies for Windows
Phone, Xbox and browser plugin will move to Windows Phone.  
• The Client and Mobile tools teams, including Windows
Phone tools and XAML tools, will stay in DevDiv. 

These changes are all effective immediately.  From a
performance review perspective, we will do this year’s performance review under
the DevDiv organization model.

I want to thank Kevin Gallo and the team for all the
great work that they have done over the years.  Moving forward, I'm very
excited to bring the client platform efforts closer to the platform
teams.  There is a lot of very exciting and critical work underway as part
of our next wave of platform releases and I am very eagerly looking forward to
seeing the team’s work in the hands of our developers and customers.  

The follow-up emails will provide more details on the
changes to those impacted.  Please join me in wishing Kevin and the team
all the very best as we move forward.  If you have any questions about
this change, please let your manager or me know.

-somasegar

Related Posts:

  • Fallon Massey

    Well gag me with a spoon!

    You really got your hands on a golden one this time, although it’s hard to sniff out the good news.

    It appears that XAML will be there in the future, or at least a mutated version.

    Silverlight plugin going over to the WP team sounds like elephant graveyard stuff to me, but I could be wrong.

  • pm

    So what happens to the .net focused P&P Teams? ( Prism, Ent Lib, Unity, etc. )
    What happens to people like Pete Brown, Tim Heuer, etc. who have put so much into their work?
    Does Pete Brown’s structure even still exist?

  • Niclas Lindgren

    It could also mean this..

    http://arstechnica.com/microsoft/news/2011/06/windows-8-for-software-developers-the-longhorn-dream-reborn.ars/

    Which would be essentially what everyone has wanted since SL was first seen..

    It would be a leapfrog in the OS platform, the pipe dream of longhorn for real.

  • K.M.

    @ Niclas Lindgren:
    Yeah it could be, but if I were to place a bet, realizing Microsoft’s recent history, I’d probably have better odds on the pessimistic version.

    If the indications of an internal civil war are close to correct, and it certainly appears like it, and WinDev has the advantage, then maybe the pessimistic view is more correct.

    I *hope* you’re right, but I’m personally not going to plan on it.

  • Pingback: A XAML is lényeges, nem csak a HTML5 - Hírcsatorna - devPortal()

  • John

    Sometimes I think to myself – “Gosh, those guys worrying about Microsoft direction are so lucky!”
    I have to wait when our biggest client finally transitions from Win 2000 to even think about using WPF in main, cash generating applications. So I believe that WPF, Silverlight and .NET will be just fine and demand for them will even increase during the following 5 years.

    And after that – Microsoft will die anyway:)

  • Niclas Lindgren

    @John
    Yes, And I think what you are talking about is why things don’t get picked up. It takes 2-3 years for any business (if you are lucky it is that fast) programs to pick up new techonology. However for the blogosphere it is here and now they measure success. iPhone 1 wasn’t a real success why it came either not even iPhone 2, yet everyone expect WinPhone 7 to be a it in no time, not even time needed to get the retail lines fed information.

    Perhaps this incredible trust, from the blogosphere , that MS will wipe the floor is their biggest obstacle

    @K.M
    If you look back all the way to PDC 2008, the MS strategy actually lines up nicely with that that articale describe, which is based on the same info as SB uses.

    I find it really incredible that the Hyper-V solution got traction so quickly in a VmWare only world, Azure is making noises and Windows 7 delivered.

    Google Wave anyone, or Google Buzz? Apple iCloud… Apple ToS…

    Everyone fails, that is how we learn, when we don’t allow people to fail anymore we won’t have progress…

  • D

    “It’s clear there is a consolidation happening that I think we can all agree on early. How will the consolidation impact the average .NET developer is likely to be dramatic enough to warrant some applications having to have code refactored down the track – you will not escape that sorry.”

    “Does this mean .NET is dead? Who actually knows what .NET vNext will be so it is hard to simply say “yes” and it could very well be a reset of .NET to fix a lot of pent up frustration in the way it sticks together today.”

    Thank you Riagenie!

    XAML is dead because XML is dying a slow death. JSON has taken the forefront for communication over the wire and with node.js becoming a first class, MS citizen (http://www.theregister.co.uk/2011/06/23/microsoft_putting_node_js_on_windows_azure/) and it should be if azure is going to compete with amazon’s cloud, then javascript is everywhere. Remember WPF/E, it should now be JS/E. Since MS won’t stick to its guns and push its own tech and make it work for the masses(apple seems to be able to do this though), well the reality then is XAML is dead/dying. Why do you need that verbose of an xml dialect for the control set going forward? Do you really need that many hooks for properties and methods to be set up for a control or do they provide baseline hooks for core HTML/CSS/JS and a code layer for the more advanced features? The translate engines underneath the hood are targeting HTML markup, CSS, and Javascript not something else. This targeting must happen if IE10 will outperform the other browsers on Win8 unless they plan on inventing their own strain of HTML tags at which point everyone leaves IE in the dust. No, the dream appears to be to create controls using C++ for the client as well as standards based HTML tags that the OS will work with. XAML isn’t a necessary layer there and why throw good money after bad in retooling blend to support output to HTML/CSS/JS. Frontpage – err – expression web does that already. The “vision” is a return to basics with control sets that are more akin to a, dare I say, winforms experience? At least in my opinion. If you can convert HTML/CSS/JS to a native app (think PhoneGap), then the same will need to be done for a Win8 app. That is if they want to continue to compete for the developer share.

    As far as web.NET goes, ASP.NET server side controls already render to HTML/CSS/JS so bump up to the next level of HTML5 – think how MS fixed the menu control from rendering out as a table to finally an unordered list in .NET4. ASP.NET5, of which nothing has been said, should do this and how nice even the numbers match(sarcasm). If they don’t do this, then there are ways currently to jump into the rendering stream of the control and output what HTML you want, just requires some work. In terms of server controls, nothing new has really come about since 3.5 and the listview control. If MS doesn’t want to be tied to pushing their framework and IIS (which doesn’t make sense because you’d think they’d want to sell server OSes, but perhaps no longer as they want to sell cloud services and put the private hosters out of business?) then toolkit the new html5 tags/server controls. Use the ajax control toolkit model and that way if they want to walk away from webforms, like they did with ACT, they can and yet folks can still enhance them on their own. Of course one could develop webforms pages just like classic asp to get around all of this as well – sigh – regression I mean progress 😉

    MVC and webapages are ruby and php spins – folks can switch to the real deals or manually type in their HTML5 goodies. Outside of a major need to revise the razor view engine, things are status quo.

    I would think major shake ups are in the works for control vendors and web hosters. C++ controls may be the bone they throw the control vendors, but don’t see any doggie treats for web hosters – maybe saying they support a flavor of ASP.NET will be enough? Won’t know until September and even then the full picture most likely won’t be revealed. What is your purpose MS? Serve the developer or serve yourself? Funny how the latter won’t exist without the prior…

  • Niclas Lindgren

    @D XAML isn’t so much about XML as it is about designer/developer workflow and databindings, something that HTML/CSS and Javascript is severly lacking.

    The name of the technology is not so important it is what features it has, if HTML/CSS/Javascript gets all the good parts of XAML/C# then of couse you can give up XAML. People aren’t complaining because they don’t have a purpose, they are complaing because they know that HTML/CSS/Javascript as a programming model is not as productive as the XAML/SL/C# model is.

    People aren’t complaing about techonolgy they are complaining about a possible loss of productivity and tools, which is actually a testament to how much people like the VS tooling today.

  • D

    @ Niclas Lindgren:
    I completely agree. My larger theme is that in less than five years, they are potentially undoing everything they have promoted and marketed and sold. Things change, but how many iterations does MS get before devs leave the platform? How does one plan, therefore, to continue implementing MS technological solutions? Place your bet. Thanks.

  • reymond

    Repost of my long conspiracy theory from SilverLight.net forum (bear with me).

    1. Client-side .Net annoys WinDiv
    WinDiv views client side .Net probably the same way Steve Jobs views Java. It’s less performant, brings extra work now that you’d have to address both native and .Net APIs and align their functionality and slows down the pace of Windows development. To them they’d like to go w/o that burden.

    2. SL RIAs give WinDiv more troubles than HTML/JS
    I don’t for one second believe WinDiv is as committed to HTML as Google and Mozilla. Why would they be? If a standard web platform gets rich and powerful enough then who still pays to use Windows. Grab a free Linux and throw a free ChromeOS on top of it. You are all set. Who needs a client-side native Windows anymore?

    They then take a close look at the viability of HTML. They see a bunch of disjoint tags, scripts and style sheets. They see a spaghetti coding architecture. They see a time table of 2014 or so for next draft and 2020 or so the final spec. They see how underwhelming GoogleDocs and ChromeOS are. They realize, correctly, that HTML as a platform is an emperor w/o cloth or a paper tiger if you will. HTML in this shaky state is no threat to Windows.

    SilverLight on the other hand is all business. It’s rich, robust and improving. For the first time SL makes desktop UX available to Web Apps. If SL truly gets broadly accepted (e.g. w/ the help of MoonLight) then people may not need Windows to run rich apps anymore. Grab yourself a free browser and a free SL plug-in and you are ready to run RIAs. SL actually poses bigger headaches to Windows (a screw-Windows product) than HTML. Imagine how much better ChromeOS and GoogleDocs could be if they were done in, say, SL5. To make matters worse SL is controlled by DevDiv needing not to answer to WinDiv, which gives WinDiv extra motives to dislike SL.

    If we put these elements together we could see why things are in the current shape when WinDiv is calling the shots. They may not mind .Net being an enterprise framework but they dislike it at the client-side so they are happy to shut it down. They then notice there are lots of HTML koolaid addicts out there so they say, “Oh well let’s throw them a bone of ‘immerse’ apps or sth. Put on a smiley face and tell them we love HTML5 to death (except that WebGL stuff getting in the way of our DirectX). What do we worry about? HTML is no threat. Being friendly to those web guys may help sell more Windows.”

    It ends up like this: If you wanna go for reach over rich then Win8 offers you some HTML hacks to play with. If you wanna go for the other way then Win8 offers a new native approach (in a form of next-gen C++ based framework perhaps). If you want both reach and rich, AKA SilverLight RIAs, FUGETABOTIT. SL5 is the last plug-in version. In future SL will be confined to Windows Phone platform where there’s no conflict with WinDiv (not yet).

    As for you SL RIA developers, well, you guys are in the way. Our apologies to whack you.

    “It’s not personal, Sonny. It’s strictly business.” Michael Corleone / ‘God Father’.

  • matware

    When i first heard about all this I laughed at the massive freak out that was going on, I couldn’t understand the fear it was generating. So I’ll share my perspective and maybe it will calm some nerves.

    1. XAML scares the pants out of the masses of designers, they don’t use it, most they barely use Flash and MS’s hope that there would be a flood of really sexy looking apps from all those Flash designers was a pipe dream. To murder a saying, If the designers won’t come to windows, windows will go to the designers. The web is the home of designers, so MS making HTML an integral part of the OS development experience is blinding obvious. Ask yourself what percentage of IPhone / Android apps are just HTML HMIs. Oh look now they run on Windows.

    2. Windows needs more than just cosmetic updates and additional services, it needs a rich HMI team. I’m running Windows 7, that start bar looks nice, but I’m always one mouse click away from an application that was written in 1995. WinDev has needed a more than cosmetic HMI update for the last 8 years and it needs people who don’t think that notepad.exe is fine. I always ask myself how the hell one person doing Paint.Net can make such a great app, when WinDev and their X developers working MS Paint can produce such a turd. Given my experience with Win32 development, the answer is because it’s hard to produce good looking apps in Win32 they a long time to develop and test. Having on hand WPF developers working on expanding the built in tools will allow them to point at the results and say… do that in Win32, I dare you.

    3. WPF/Dot are aren’t suitable for writing massive chunks of an OS. Or they weren’t until Dev took on the task of writing massive chunks of VS2010 in WPF, they found and fixed untold numbers of bugs. I suspect that the Dev team eating their own dog food (for the first time?) made some of the consumers of their work (Win/Office) actually seriously consider WPF. Up until then nobody had actually made a serious app with it and I’m sure WinDev just looked at the whole .Net stack as the VB Next. Now they have done that they may have something to contribute to the Win team when it comes to how to build ‘real’ software with it rather than just pretty little internal toys.

    4. DotNet apps will run on ARM, so how do you maximize reuse? The more .net apps there are in the Windows 8 core install, the more apps you have to build and test. If MS is planning an appstore to come out with Windows 8, then web apps and .net apps are the simple way to cross architecture happiness. If they are also seriously pushing touch interfaces on windows 8, then all those fiddly little configuration forms better scale well to 2560×1440 and still be finger friendly otherwise they’ll be laughed off the stage. Again, it means big chunks of Windows 8 being written in something like WPF and something architecture agnostic like .NET, either that or HTML and JS :-p

  • Pingback: HTML5 and Silverlight (It’s been a long summer) « Affirma Consulting Blog()

  • ac

    I do a lot beside programming so I don’t jump on all the latest buzztech. Last tech I jumped on was C# when it was at 1.1 and have evaluated related things like WPF/SL. I took a decent look at WPF and Blend something about it just didn’t feel right and the cold startup time for blank window app was not just 2x slower than say Win32 Delphi blank app. 100x slower maybe? Hot startup is better but that doesn’t count unless the OS loads the framework GUI bits on boot. To me that immediately screams “AVOID”, since I press F5 alot during my C# development and my code isn’t running within ~150 milliseconds then it gets very old very quick to do quick iterative development and debugging..

    Recently I re-evaluated Silverlight simply by doing a test run of some hello-world level sample apps and the feeling was that computers need to be 15 Ghz with SSD’s in order to have that stuff even barely usable.

    You might say that my “evaluations” are non-sense but so far it has worked well for me in avoiding spending too much time on stuff that wasn’t good to begin with.

    So now the question is, what’s the “evaluation” for WinRT/Metro? API Latencies (round trip from pressing key/touch to displaying change on the screen through your logic), cold start perf, execution jitter (eg. fast touch-drag where your code interacts immediate-mode style instead of declaratively)… ? I call these the fundamentals of a development platform. Nice features are needed on top of these to make the whole thing attractive.

    Last exciting stuff I read from MS was Singularity OS. C# style code with less latency for driver calls and secure. WinRT and COM with 50 instructions overhead? I have not digged around yet so as to see how often these overheads are going to be paid but if as end user I’m going to try two tablets and one is laggy to my touch and stuff jerks around/doesn’t react smoothly then that’s all it takes to tell that if the platform can’t get the basics right, then it’s unlikely they get anything else right. And if I as user would not buy the platform, why would anyone want to develop on it?