The Daydream Blog

Microsoft’s Windows marketing seems to have been extremely muddled of late. Windows Vista has been a publicity nightmare for the company, garnering poor reviews, demands for XP’s lifetime to be extended, users downgrading and major partners and customers holding off from upgrading. Some major clients have even decided to take the opportunity to switch to the Mac.

What has been most surprising throughout the first half of this year, was Microsoft’s unwillingness to come out in support of its most unloved product. Key Microsoft staff, including Bill Gates himself, spent considerable effort talking up Vista’s successor, Windows 7. This led to conjecture that Windows 7 would be released much earlier than expected. It may also have led to many enterprise customers delaying their Vista upgrade programmes.

Windows 7 talk went quiet midyear, with Microsoft promising to come out fighting for Vista. The first fruits of that campaign is the Mojave Experiment. Irrespective of the scientific and content value of the experiment, from a marketing perspective it is a disaster. Essentially a Pepsi challenge, Vista is put into a blind taste test against not its current competitors, but against the 7 year old product it was supposed to replace.

With Mojave barely having had time to make an impact, Microsoft is now talking about Midori, a complete replacement for Windows altogether.

What is really surprising about the situation is that Microsoft’s muddled marketing may actually be leading to a reduction in their Vista sales, as customers wait on Windows 7. Has Microsoft created an Osbourne effect entirely of its own making?

With its stock flat-lining, Microsoft’s mind share is taking a significant battering. Tellingly, start up software companies no longer plan to be taken over by Microsoft, nor worry about a competing Microsoft product. Instead Google, Apple, Amazon, eBay and others are the gorillas in their respective fields. The Yahoo! take over debacle clearly demonstrates how “Microsoft is dead” in terms of its influence.

I recently concluded that the role of CEO of Microsoft would be a much more exciting role than succeeding Steve Jobs at Apple. There is so much more potential to improve a company.

As an unashamed Mac zealot, it is actually disappointing to see Microsoft in this situation. However, the prospect of a resurgent Microsoft in the future, in a more competitive market place will only be good for consumers in the long run. Its current problems are a surprising, sad, but necessary step along the way to Microsoft’s rehabilitation.

“The most important election in our lifetime” will be about whether or not the US takes a lead in energy policy. Obama talks of a 10 year goal to make the US independent of Middle East and Venezualan oil. It is a task, that if backed by the government, entrepreuners, business and the people of America, could be as inspirational and motivational as JFK’s target of landing a man on the moon. It would also mean taking the biggest step forward, so far, to tackle climate change.

A major European newspaper company has decided to move its entire workforce over to the Mac. Estimates for Axel Springer range between 10 - 12,000 desktops, which will make them Apple’s second largest client, behind Google.

There are continued criticisms of Apple’s lack of focus on the enterprise. With the iPhone, Apple made a concerted effort to answer their critics with one big software update. Whereas with the Mac, support for VPNs, Exchange, workgroup sharing, security keycards has been a slow, but steady process. The main outstanding issues are the lack of a public product timeline, guaranteed configurations available for 3-5 years and the lack of some key third party enterprise software.

With the switch to Intel, Apple’s product timeline is now at least translucent, if not transparent. With Apple’s financial stability, I believe that Apple can afford to be less secretive about its future product road map. Certainly with the iPhone and Apple TV, it has been willing to pre-announce products and deliver a roadmap. Whilst this is less critical when Apple’s minor updates can be accurately predicted from Intel’s plans, Apple should bring more openness to its Mac plans. For example, the MacBook Air was only a rumour until just a few days before the announcement. An enterprise client looking to make a large laptop purchase would have been very disappointed to have that announcement sprung on them just after placing a large order. Pre-announcement of better Exchange support in Snow Leopard is a step in the right direction.

With Mac OS X’s tight integration with Mac hardware, having a guaranteed configuration is a minor issue. IT departments are unlikely to face major issues with even major updates to the Mac line. The only concern here will be employee envy, where the newest staff may end up with the newest, fastest, computers.

The lack of third party software is a chicken and egg situation. Apple has slowly but surely been adding enterprise class software to Mac OS X Server, but there is still a lack of key software. Namely middle-ware, such as IBM’s Websphere MQ, TIBCO Rendezvous or Orbix; CRM software and HR software. As the Mac’s presence in enterprise grows, this is really an opportunity for small Mac software firms to become big Mac software firms.

Note this article is intended for developers and may be of limited interest to our wider client base.

Kevin Hoffman has given a critique of the now infamous criticism of iPhone development. (Another Microsoft Developer Falls Victim to Framework Myopia [The .NET Addict's Blog])

The key points in Kevin’s post are his explanation of how experienced Windows developers should handle their migration to Mac or iPhone development. It reminded me of a post I’ve been planning to write for a long time that most easily be summarised as:

To learn Cocoa, you must un-learn your past development experience.

Cocoa assumes an MVC paradigm. Interface Builder and NIB’s are very weird compared to other approaches. The frameworks are heavily designed for sub-classing to create your own behaviour, rather than having lots of settings to tweak behaviour.

Every time you try to learn a new aspect of Cocoa, you find yourself struggling to do something that feels like the “right way”. A few days later you come out with 3 lines of code that feel completely natural but felt incredibly painful to reach. You will be met with countless, “aaah, that’s how it works and it makes complete sense” moments. You spend days wondering why this bit of Cocoa is so gnarly, and seemingly unintuitive, until you reach another moment of understanding.

People often say that Cocoa has a steep learning curve. It does not. It has a steep un-learning curve, followed by a gentle re-learning curve. Getting yourself out of your previous development mind set can make the whole experience a lot more pleasant.

What is disappointing is that none of the documentation or tutorials spend much time comparing and contrasting with other approaches, that would make the process of unlearning easier.

The next MacMacDev for Edinburgh has been announced for Thursday 10 July 2008, starting at 19:00. The venue is unchanged and is at Baroque, 39-41 Broughton Street, Edinburgh, EH1 3JU. Full details can be found here.

There’s also a new mailing list for the community, full details can be found here.

Hope to see you there.

Note this article is intended for developers and may be of limited interest to our wider client base.

Last week saw the release of Cornerstone, a new Subversion GUI client for the Mac from Zennaware. After initial testing, I am quite excited about Cornerstone, as it breaks new ground for Version Control Graphical Clients.

Historically, graphical version control clients have simply put a GUI on top of the command line tools. It is disappointing that there is little or no thought given to the problems that developers want to solve with their source control systems. Developers of these applications seem to completely forget the issues they themselves face.

Cornerstone’s most impressive feature and one that I have discussed extensively in the #macsb IRC channel is it’s timeline. You can easily go forward and back through the history of a file and see changes for each subsequent version. Whilst this falls short of my wishes for a source control timeline, it is a massive step in the right direction.

Cornerstone’s source code diff tool looks great, but still cannot identify when a method has simply moved within a file. If you change the method as well as move it, which is quite common, the changes are not clearly highlighted in Cornerstone, Xcode or any other diff tool I have come across on the Mac. I had already decided that a future version of Differencia will act as a source-code aware diff tool.

Whilst there are a long list of feature requests I have already for Cornerstone, after only a day’s use, it is a huge relief to find that someone is taking source control seriously.