Category Archives: microsoft

Jewels from the loft: launch of Delphi, Netscape’s Constellation, HTML to die, Longhorn for developers

It’s the Easter holiday in the UK and I’ve suffered a bout of spring-clean fever. It is time, I decided, to clear out a mountain of old books and magazines.

A job like this always prompts reflections, the first of which is the sad decline of print journalism in the field of software development. It hurt to send piles of Byte, Exe, Dr Dobbs’s Journal, Application Development Advisor and others off for recycling.

image

A few things caught my eye. Exe June 1995, and there is a young Anders Hejlsberg talking to Will Watts about his new creation: Borland Delphi:

Before Delphi, you always had to make a choice. Do I go for the performance of a native code compiler, or the ease of use of a visual development environment? Do I go for a powerful object-oriented language, or a proprietary 4GL client/server tool? What programmers really want is all of the above, in one package. That’s what we set out to do.

What is striking about Delphi is that this was not hype. It delivered on that promise. It was better than its obvious rival, Microsoft’s Visual Basic, in almost every way (I will give VB a point for sheer ubiquity, especially in VBA guise). Delphi is still with us today, not bad after fifteen years. However, it never came close to VB’s market share, which shows that quality has never been the sole or even the most important determinant of sales success.

Next up is Byte, March 1998. “Reinventing the Web”, the cover proclaims. “XML and DHTML will bring order to the chaos”.

image

Inside there is a breathless description of how XML will change everything, and a quote from Jon Bosak:

HTML, this so-called ‘hypertext markup language,’ implements just a tiny amount of the functionality that has historically been associated with the concept of hypertext systems. Only the simplest form of linking is supported – unidirectional links to hard-coded locations. This is a far cry from the systems that were built and proven during the 1970s and 1980s.

Indeed. “We need to start replacing simple HTML with more powerful alternatives”, the article concludes. “The migration to XML must begin. The future of the Web depends on it.”

Here’s one thing that mostly did not work out as planned. The W3C tried to retire HTML, failed, and is now belatedly engaged in specifying HTML 5.

Byte March 1997 is also intriguing. Netscape’s Marc Andreessen smiles out of the cover.

image

Jon Udell, in the days before he disappeared into some Microsoft corridor, writes about Netscape’s “Constellation: the network-centric desktop”:

Netscape’s Constellation takes a less Windows-centric approach and puts more emphasis on location-independent computing, regardless of the platform. No matter what kind of system you’re using or where you are, Constellation presents a universal desktop called the Homeport. Although the Homeport can appear in a browser window, Netscape usually demonstrates it as a full-screen layer that buries the native OS – certainly one reason Microsoft is not embracing Constellation.

Netscape got a lot of things right, a true pioneer of what we now call cloud computing. What went wrong? Well, Microsoft went all-out to conquer Netscape by removing its browser dominance. Microsoft’s weapon was the free Internet Explorer.

It is all a pre-echo of what is happening now with Google and Microsoft, the difference being that Google has huge financial power thanks to its marriage of internet search and internet advertising. Unlike Netscape, Google is winning.

This blog is long enough; but I’ll give a brief mention to another jewel from the archives: a book given out at PDC 2003 entitled Introducing Longhorn for Developers.

image

It describes Microsoft’s vision for Longhorn: a radical new application model for Windows, building on XAML, WinFS and “Indigo”, the communication framework. It bears little resemblance to what eventually appeared as Vista, which is a shame as it was compelling in many ways.

Microsoft accused of failure to observe Open XML standards process

XML specialist Alex Brown, who was involved in the ISO standardisation of Microsoft’s Open XML – still perhaps best known as OOXML – says Microsoft has failed to honour the commitments it made when the standard was approved. In particular, it seems little progress has been made between Office 2007 and Office 2010. The key problem is that Microsoft implemented Open XML before it was standardised. There were numerous changes made during the standardisation process, but what to do about the existing implementation? Loosely, the existing unacceptable format was given a “Transitional” status, while the more satisfactory, corrected format was called “Strict”. Microsoft promised to implement the “Strict” variant as soon as it could. Brown adds:

I was convinced at the time, and remain convinced today, that the division of OOXML into Strict and Transitional variants was the innovation which allowed the Standard to pass. Enough National Bodies could then vote in good conscience for OOXML knowing that their preferred, Strict, variant would be under their control into the future while the Transitional variant (which – remember – they had effectively rejected in 2007) would remain purely for the purpose of accurately specifying old documents: a useful aim in itself.

It is now two years since Open XML was approved, and Microsoft is on the brink of releasing a new version of Office. So does Office 2010 implement Open XML Strict? Apparently not – it’s the Transitional version. That is bad enough; worse still, according to Brown, it does not even conform correctly to that:

It is also a worrying commentary on the standards-savvyness of the Office developers that the first amateur attempts of part-time outsiders find problems with documents which Redmond’s internal QA processes have missed. I confidently predict that fuller validation of Office document is likely to reveal many problems both with those documents, and with the Standard itself, over the coming years.

Note that Brown is basing his remarks on the preview of Office 2010; we have not seen the final release yet. I can believe that Microsoft may fix some issues, but it looks vanishingly unlikely that Office 2010 will implement the “Strict” standard which ISO approved.

Brown’s remarks shed light on something I noticed when reviewing the preview:

As for Open XML, it’s notable that Microsoft neglects to mention it at all in its Reviewer’s Guide, even though this is supposedly the release that will fully implement ISO/IEC 29500. It is odd how this has gone from a cause to campaign for, to not-worth-mentioning in just over a year. To be fair, few users ever cared about XML formats themselves: it is only when documents get scrambled or fail to open that such things become important.

No wonder Microsoft said nothing about it, if in reality it has lost interest in conformance.

I think it is a good thing for Microsoft to standardise its Office formats. Selfish manipulation of standards committees on the other hand is not acceptable. One thing is for sure: if Brown is right and

without a change of direction, the entire OOXML project is now surely heading for failure.

then the company will only have itself to blame. Its nightmare will re-emerge: entire governments mandating OpenOffice for the sake of  standards conformance.

That said, and despite the hype, I regard Office 2010 as a minor release. 64-bit Excel, a few tweaks, and a first foray into browser-hosted versions. Microsoft often displays this pattern, following up a release with major changes – Office 2007, for example – with one that is really just a refinement of what went before. It is not impossible that somewhere in the corridors of Redmond a team is working on a new Office that does a much better job with the Open XML standard.

Over to Microsoft – serious about Open XML? Or just doing the minimum necessary to protect a lucrative market dominance – maybe a bit less than the minimum?

Update: Microsoft’s Doug Mahugh has replied to Brown’s comments here. I am writing separately about this.

Building for multiple mobile platforms with one codebase

Individuals may have strong opinions about the merits of Apple iPhone versus Google Android versus the struggling Palm WebOS versus the not-yet Windows Phone 7; but sit them round a table to discuss app strategy and those diverse platforms change from a debating point to a problem. Presuming a web app won’t cut it, how do you target all those devices without the unreasonable expense and complication of managing multiple projects? The native languages are all different; Objective-C for iPhone and iPad, Java for Android and RIM BlackBerry, JavaScript for WebOS, C# for Windows Phone 7.

There are three possibilities that come to mind. One is that all the platforms will eventually allow you to write in C or C++, making this the unifying language, though you still have some fancy footwork to do overcoming library differences. Android now allows this via the NDK, and Palm via the PDK. There is currently no alternative to Java for Blackberry, and Microsoft says native code won’t be possible on Windows Phone 7, but well, you never know.

The second is Adobe Flash. This is an interesting one, because Apple prohibits Flash on the iPhone, but Adobe has a Packager for iPhone that builds native iPhone apps from Flash projects. Another issue is that although Flash is available or promised for all the main non-Apple devices – Apple’s gift of a selling point to its rivals – it is not Flash alone that does what it needed, but AIR, the “desktop” or out-of-browser runtime. This has been previewed for Android and promised for other devices including Blackberry. AIR for Windows Phone 7? Maybe, though I’ve not seen it mentioned.

A third idea is a clever framework that does the necessary cross-compilation under the covers. This cannot depend on deploying a runtime, nor compiling to native code, because these approaches are blocked by some mobile platforms. Rhomobile has the Rhodes framework, where you code your app in HTML and Ruby and compile for devices including iPhone, Windows Mobile, RIM Blackberry, Symbian, and Android. Rhodes includes an MVC (Model View Controller) framework and an ORM (Object Relational Mapper) to wrap database access. There is also a RhoSync server component to enable offline data with synchronisation back to the server when reconnected; and the RhoHub hosted IDE for buildings apps with a web browser.

Rhomobile tells me that Palm WebOS support is in the works. They also promise Windows Phone 7 support, which intrigued me because Rhodes says it compiles to “true native device applications”. Has Rhomobile gotten round Microsoft’s opposition to native code? Apparently not; VP Rob McMillen eventually told me that this will mean a .NET IL (intermediate language) implementation.

The Rhomobile approach reminds me of AppForge, a company which produced the well-regarded Crossfire add-on for Visual Studio and compiled Visual Basic to a wide variety of mobile platforms. Unfortunately AppForge was acquired by Oracle, and its new owners showed callous disregard for existing customers. Not only did development cease; it also became impossible to renew existing licenses. Thanks to an activation component, that also blocked new deployment of existing applications – every developer’s nightmare.

That said, there is no activation requirement for Rhodes that I know of, and the framework is open source, so I don’t mean to suggest it will suffer a similar fate.

What about Java? On the face of it, Java should be ideal, since multi-device support is what it was designed for. It is a measure of how far Java has fallen that we hear far more about the lack of Flash on the iPhone, than the lack of Java. Microsoft says yes to Flash on Windows Phone 7, though not on first release, but nothing about Java.

Java as a mobile runtime needs a strong dose of lobbying and evangelism from its new stewards Oracle if it is not to fall by the wayside in this context. Hmm, AppForge.

Bill Buxton, Windows Phone 7 hits at Mix10, say attendees

I recall seeing conversations on Twitter about whether it is worth going to conferences, especially vendor-led ones such as Microsoft’s Mix10 which took place earlier this month. It’s expensive, it’s marketing, and many of the sessions are available for free online soon after they take place.

In that context it’s interesting to see some off-the-cuff comments from attendees about what they liked at Mix. Networking comes high on the list, something you cannot get from a distance. There is general boredom with slide-driven presentations, and enthusiasm for interactive panels. One attendee says:

I always find the panels more exciting, and I like the open dialogs where people can ask questions and get real world opinion and advice

As for the content, the name that comes up most often is Bill Buxton, the design expert who joined Microsoft specifically to introduce greater design-consciousness into a company whose track record in this respect is poor. It was a fascinating appointment, and I believe you can see traces of his work in Windows 7 and much more in Windows Phone 7. I talked to Buxton briefly about this and wrote it up here. When I say “his work”, I don’t mean hands-on screen designs, rather a different approach to product development.

Buxton apparently made a considerable impression at Mix10. His session An hour with Bill Buxton (the link has the video) is also one that is interactive, and worth watching if you have an interest in where design fits into the development process. The fact that Microsoft-platform developers are engaging with his ideas is encouraging, given the usability issues that spoil so many Windows applications and machines.

Overall it also seems that Windows Phone 7 went down well – remarkable when you consider that Windows Mobile was all-but dismissed as irrelevant a few months ago – though I did smile at the Microsoft employee (I think) who shows an iPhone with a Mix scheduling app installed. Its special feature: shake the iPhone and it displays a Windows Phone 7 screen, handy if Ballmer appears in the corridor:

image

One final point: an attendee says IronRuby was shown running on Windows Phone 7 in one of the sessions, something I was told elsewhere was not possible.

Windows Phone 7 emulator yields its secrets

So that didn’t take long. Microsoft has made a Windows Phone 7 emulator available to all, explaining that it is a virtual machine running the real device OS. While it works fine for debugging applications, most of the phone UI is mysteriously absent, the exception being Internet Explorer (needed for testing web applications). However, Dan Ardelean has done some nifty work with the emulator image, discovering that a lot more of the UI is present and can be exposed if you make a few modifications. He posted his modified ROM but has since removed it because of copyright concerns. I doubt it will be hard to find.

image

Applications available in the hacked emulator include Mobile Office with Word, Excel, PowerPoint, OneNote and Sharepoint Workspace (undermining Microsoft’s claim that this is only a consumer device), games including a version of Times Reader – full circle for this one, which started out in WPF, moved to Silverlight for Mac, deserted and went to Adobe Flash, and may now reappear in Silverlight – and music/video via a Zune section.

image

My tip to Microsoft: if you don’t want these things to leak, best not put them in the publicly-available image.

Ardelean has also figured out how to amend the native C# port of SQLite, satisfying demand for a local SQL database engine. SQL Server Compact Edition is actually present in Windows Phone 7, but not available to custom applications.

Visual Studio software factories to emerge from Microsoft deep freeze

Last night at the Microsoft Mix party in Las Vegas I happened across Michael Lehman, a senior architect, who told me he had been working for the last six years on a Visual Studio add-on called Feature Builder. This turns out to be the evolution of the very same project which Microsoft’s Jack Greenfield told me about back in March 2007, at an architecture conference in the UK, though back then it had a more exalted status (in Greenfield’s eyes at least) – he described it as a platform rather than merely an add-on. Now Feature Builder is to appear as an extension in the Visual Studio 2010 Gallery, only discoverable by those who seek it out. If enough developers find it and like it, it may end up as part of a future Visual Studio release.

It’s all part of Microsoft’s desperately confused architecture and modelling story. Note that Greenfield in 2007 denigrated the UML:

The UML is a collection of useful abstractions. Unfortunately it’s been peddled as a universal modelling language, but the U never stood for Universal. We subscribe to Michael Jackson, author of Problem Frames, who says that there is no such thing as a universal solution. It’s a childish approach. This is where the Universal Modelling Language marketing pitch fails. UML was never properly extensible. It also has the problem that it was designed by a committee.

So what do we have actually shipping with Visual Studio 2010? Standard UML modelling, the best Microsoft has yet come up with.

Reading between the lines, this likely means that the UML faction within Microsoft outvoted the non-UML faction. However, each release of Visual Studio seems to have its own unique approach to modelling and architecture tools, so it would be no great surprise if Visual Studio 2012, say, replaced them with something different.

As for Feature Builder, it does sound interesting – read the Greenfield interview referenced above for why it has potential. Lehman says it is a simpler approach than previous software factory tools from Microsoft, and named Feature Builder to avoid association with past efforts.

Update: Lehman has blogged about the project here, and you can download the preview here.

Microsoft playing HTML 5 standards game alongside Silverlight game

I’m at Mix10 in Las Vegas where Microsoft has been showing off the latest preview of IE9 – you can try it here, provided you have Vista SP2, Windows 2008 or Windows 7.

The two themes are performance, with GPU-accelerated HTML and graphics and a new Javascript engine that compiles code in the background, and standards support. This latter was not a surprise to me, as I’d heard the well-informed Molly Holzschlag praise Microsoft’s commitment to HTML5 at a workshop here on Sunday – see this earlier post.

During the keynote, we saw IE9 playing a video using the HTML 5 video tag – no Flash or Silverlight needed. Microsoft also showed that in this instance IE9 performed better than Chrome thanks to better hardware acceleration. Although one should always mistrust one vendor’s demonstration of another vendor’s product, it should not be surprising that Microsoft is able to deliver a browser that is better optimised for Windows.

Video, hardware accelerated graphics, audio element support, fast JavaScript: there is considerable overlap with the features of the Microsoft Silverlight (and Adobe Flash) plug-ins.

The plug-in approach has advantages. It offers consistency across browsers, and enables rapid evolution without the hassles of standards committees. The multimedia features in Silverlight and Flash are well ahead of those in HTML 5 – Holzschlag nailed this when she described today’s HTML 5 demos as reminiscent of Flash demos a decade ago.

Still, if you can do without the plug-in you end up with cleaner code, removing the awkward transition between what is in HTML and JavaScript, and what is in the plug-in. There is also a better chance that your code will run on Apple’s iPhone and iPad, for example.

The question though: can Microsoft do an equally good job of supporting HTML 5 throughout its platform, as it will do with Silverlight? This is where I’m doubtful. The Visual Studio and Expression tools will continue to drive developers towards Silverlight rather than HTML 5.

It’s notable that shortly after Microsoft’s IE9 demos at Mix, we saw demos of fun technology like code-name Houston, develop databases in the cloud using just your browser and … Silverlight.

No native code development on Windows Phone 7 says Microsoft – so what about Flash?

Windows Phone 7 is a managed code platform, we’ve been told at Mix10 in Las Vegas. Development is via Silverlight or XNA; there is no native API.

Of course there is a native API; the question is more about what code is allowed to access it. Still, in the press briefing the spokesman was clear that native code development will not be supported.

What about projects like Adobe’s Flash runtime, which both Microsoft and Adobe have said is planned, or at least (in Microsoft’s case), not blocked – although we already know that Flash will not be available in the first release.

All my spokesman would say is that nothing has been announced about that.

My suspicion is that in reality certain privileged vendors will be able to, in effect, extend the operating system with native code libraries. Adobe could be one of those; so too could a company like Rhomobile, which has a cross-compiler for a variety of mobile platforms. So I doubt that Microsoft has yet given us the full story here.

Update: The latest on this is that Microsoft’s Charlie Kindel says that Adobe will have special native access for Flash, but that no other vendor will have that privilege. This still does not make sense to me. Let’s suppose that Windows Phone 7 is a big success. What justification could Microsoft have for supporting the Flash runtime but not the Java runtime, for example? I suspect that Microsoft is chasing the Flash checkbox to one-up Apple; but if Adobe gets native access, others will no doubt follow.

Windows Phone 7 developer story unveiled at Mix10

I’m in Las Vegas for Microsoft’s Mix10 conference, where the developer story for Windows Phone 7 series is being unveiled. According to the press release, the tooling for Windows Phone 7 looks like this:

  • Visual Studio 2010 Express for Windows Phone (free)
  • Windows Phone 7 Series add-in for Visual Studio 2010 RC
  • XNA Game Studio 4.0
  • Emulator
  • Expression Blend for Win Phone CTP

Essentially, you are meant to use XNA for games on the device, and Silverlight for other kinds of application.

Another part of the announcement describes new services for developers – Microsoft Location Service to provide location information, and a notification service to “push information to the phone, regardless of whether or not an application is running”.

Applications will be marketed through a new Windows Phone Marketplace.

I’ll report more details as they emerge, here and on The Register.

Update: Microsoft has added that Expression Blend 4.0, for Silverlight 4.0 support, will be a free upgrade from Expression Blend 3.0.

Silverlight 4.0 RC, Expression Blend 4.0 beta, VS 2010 add-in available for download here: http://silverlight.net/getstarted/silverlight-4-beta/ 

The MSDN documentation for Windows Phone 7 is here: http://msdn.microsoft.com/en-us/library/ff402531(VS.92).aspx

Functional programming, NOSQL themes at QCon London

One reason I enjoy the QCon London software development conference is that it reflects programming trends. Organiser Floyd Marinescu described it as by practitioners for practitioners. In previous years I’ve seen themes like disillusionment with enterprise Java, the rise of Agile methodologies, the trend towards dynamic languages, and the benefits of REST.

So what’s hot this year? A couple of trends are striking. One is functional programming. Don Syme, Principal Researcher at Microsoft Research and co-inventor of F#, gave a lively presentation on functional approaches to parallelism and concurrency. He shows screen after screen of equivalent F# and C# code, illustrating how F# is more concise and expressive, as well as being better suited to concurrent development.

F# is one of the languages included by default in Visual Studio 2010, which should be released shortly.

I asked Syme what sort of problems are not well suited to F#. In his reply he described the state of play in Visual Studio 2010, where you can easily create F# libraries but there is no designer support for user interface code, such as Windows Forms or Windows Presentation Foundation. That is merely a tooling issue though.

Syme’s point is that functional programming, and F# in particular, is ideal for today’s programming challenges, including concurrency and asynchronous code.

If nothing else, he convinced me that every .NET programmer should at least be looking at F# and learning what it can do.

The functional programming track at QCon is not just about F#, of course, though in some ways it seems to be the functional language of the moment.

The other theme that has made a big impression is NoSQL, or what the QCon track calls “Non-relational database managers and web-oriented data”.Geir Magnusson from Gilt Groupe talked about the challenge of running a web site which has extreme peaks in traffic, and where every user needs dynamic data and transaction support so simple caching does not work. They were unable to get their relational database store to scale to handle thousands of transactions a second. They solved the problem with an in-memory non-relational database.

In another talk, the BBC discussed their use of CouchDB for highly scalable web sites.