Monthly Archives: May 2007

drm, internet, multimedia

Amazon and the future of music downloads

1 Comment

What’s the significance of Amazon’s announced DRM-free music download store?

Amazon is a major internet property for which I have a lot of respect. It had to decide between Microsoft DRM (“Plays for Sure”) or none, and it went for the latter. I think this is the end of the road for Plays for Sure. Apparently nobody can compete with iTunes. There is Zune, of course, but Zune isn’t even Play for Sure.

So it is DRM-free vs iTunes. As for iTunes, it will do both. Where we go from here?

Amazon’s problem is limited inventory. Of the major labels, so far only EMI is willing to go DRM-free. Users who purchase music downloads will stick with iTunes if they can’t get the songs they want.

Amazon’s prospects therefore hinge on whether or not other major labels follow EMI. That in turn will depend on how it works out for EMI. If it is seen to be growing its market share because it offers a better download product, others will abandon DRM and Amazon downloads can prosper. On the other hand, maybe EMI is devaluing its product. Perhaps the public will now perceive EMI music as free music, and actually buy less of it. In this case DRM, and Apple, are the winners.

While Apple has won the DRM-encumbered download war, it won’t necessarily have an easy ride in future. There are several interesting issues. One is how many users will bother paying for music at all. I’d love to know the age profile of iTunes customers. When I asked a teenager whether she ever paid for downloads, she just laughed. Yet music execs have told me that most customers buy less music as they age. If you follow the logic through, it implies that iTunes may be squeezed by an up-and-coming generation that doesn’t regard music as something you need to purchase, and an older generation moving into a time of life when they stop acquiring music.

A second factor is whether the music labels will continue to tolerate Apple as a middle man. In this respect, there’s an intriguing Reuters report which says that “the paid download video market is a dead end”. The report is a bit confusing, but seems to predict that free online video supported by advertising will win over paid-for downloads. The fundamental question is whether content providers will continue to let their customers interact with iTunes, giving Apple a cut of the proceeds, when they could interact with their customers directly. Although this report is about video, similar considerations apply to music.

Perhaps we should identify three phases in music “ownership”. Phase one was when you purchased a physical item – vinyl or CD. Phase two was when you downloaded music file by file. Phase three is when you just play music, leaving it to the system to work out whether it is played from a locally cached file or streamed from the internet.

Phase three is the one that makes sense in the digital era; phase two is a short-lived transition period. Phase two exists for two reasons. First, when connectivity and bandwidth is limited streaming does not work well. Second, it reflects the difficulty we have mentally adjusting to new technology. Paying for a download is physical media thinking translated to the Internet age.

That’s why I still think the subscription model is the only one that makes sense, long term. Either that, or I suppose everything may become free, which is the subscription model with zero fee. Amazon’s store may have some success for a year or two, but in due course nobody will pay for individual downloads.

 

software

Open Office chokes on Open XML spec

1 Comment

I’ve been looking at the notoriously lengthy specification for ECMA-376, also known as Office Open XML. You can download it here in both PDF and Open XML (.docx) format. I grabbed both.

At 5219 pages the Markup Language Reference is a seriously large document. The PDF is about 34MB, and the docx about 14MB. I started with the PDF, which opens easily enough, but is desperately slow to search, prompting me to try the alternate docx version. This dialog amused me:

There are too many spelling or grammatical errors in Office Open XML Part 4 - Markup Language Reference_final.docx to continue displaying them.

The docx took longer to load than the PDF, but searching is indeed quicker. Frankly it’s impressive that it is usable at all. I was trained to avoid long documents, on the grounds that they are prone to corruption, and that if they do corrupt you lose more work. That was when we thought 100 pages was long. I cannot think of any good reason why this document is not broken into smaller pieces

Still, it makes an interesting test case. I wondered how Open Office would cope with a document this size. I saved it as a .doc – 68MB – and loaded it into Open Office 2.1. This took several hours (I left it overnight). Once loaded, Open Office repaginated the document to 7453 pages. Searching it was pretty quick though, if anything faster than Word. Finally I saved it in Open Document format. 15MB. Note that both Open Document and Open XML are zipped formats, which explains their smaller size.

Open Office takes 31 minutes to load the Open Document version – quicker than loading the .doc, but not tolerable for normal work. By contrast, Word 2007 can load the .docx to a usable state in around 5 minutes. It all suggests that Word 2007 and/or Open XML is superior for very long documents. I’m using Vista by the way, with 3GB RAM.

What about the content? So far I’m impressed. The entries I’ve looked at have been clear,to the point, and include example code. No doubt there are dark corners, but this strikes me as a good effort.

 

Technorati tags: , , , , ,
software, windows

Microsoft vs Open Source: only one loser

3 Comments

Microsoft, of course. Fortune reports that Microsoft will seek to extract royalties from users of open-source software. That would be monumental folly. Here’s why.

First, the company already has an image problem. It’s the “Evil Empire”, the vendor we love to hate. Litigating against free software would be appalling PR.

Second, let’s consider who would lose out if Microsoft succeeded in making widely used open source operating systems or applications illegal. Clearly, it would be the users of that software. But these users are in many cases also Microsoft’s customers. Windows on the desktop, Linux and Apache on the server, for example. Anyone who uses the internet uses open source software. If Microsoft litigates against open source, it will be litigating against its own customers.

Third, Microsoft won’t succeed. I don’t find it difficult to believe that:

…FOSS infringes on no fewer than 235 Microsoft patents.

as Fortune reports. But what if big patent holders like IBM decide to trawl their files looking for ways in which Windows or Office might infringe a patent or two? I’d be astonished if they came back empty-handed. This is not a game that Microsoft can win.

Fourth, the free and open source software movement is good for all of us. It’s lowered prices and fostered innovation. That’s a problem for a company that decides to attack it, because everyone will want it to fail.

Fifth, Microsoft has a dismal record in the courtroom.

Sixth, major legal confrontations are a huge distraction. They drain productivity. They divert energy and attention away from what the company is good at.

If Fortune is to be believed, Microsoft has been listening too much to its lawyers, and not enough to its customers.

Microsoft can thrive alongside open source. The way to do so is to create great software like Silverlight. Not by embarking on unwinnable legal contests.

Despite the above, I can understand (though not approve) that Microsoft may wish to mutter about its patents now and again, to spread a little FUD and dissuade customers from a switch to Linux. This may be no more than that. Otherwise, it is making a costly mistake.

 

Technorati tags: , , , , ,
software development, windows

Office Open XML needs wrapper classes

I’ve been writing a hands-on style article on using the Packaging API and Office Open XML. It all works smoothly, but its glaringly apparent that Microsoft needs to provide some wrapper classes for Open XML documents to make the API more usable. You can download some code snippets, which is a start, but these only scratch the surface. After all, Office developers are used to the COM automation API which makes it relatively easy to create and manipulate documents. By contrast, to get anywhere with Open XML you have to understand the raw XML. It is still miles better than working with RTF, or with the binary Office formats, but more difficult than it should be.

By way of illustration, look no further than the official code snippet for XLInsertStringIntoCell, which does what its name suggests: writes a string into the specified cell of an Excel spreadsheet. It is over 200 lines of code and comment (in fairness, more comment than code). That would be just 2 or 3 lines in VBA. Of course, once you have the wrapper function, it is just as easy. Unfortuntely there is a lot to wrap, but it is not necessary to be comprehensive. A simplified DOM that enabled the creation of basic formatted documents and spreadsheets without having to write thousands of lines of code would meet most needs.

I realise that you can write a “Hello world” document fairly easily; I’ve done it myself. But “Hello World” is not particularly useful. After all, you could output plain text or CSV, and Word and Excel will open them happily enough. The point of Open XML is to enable documents that have a little more to them: headers and footers, specified margins, rich formatting with fonts and paragraph styles, tables and graphics. Doing all that in Open XML is not trivial.

I am not the first to think along these lines. For example, here’s some code posted to CodePlex for working with Excel. Good stuff, but personally I’d like to see some official libraries, or even a well-run (WIX-style) officially endorsed open source project for this. It’s badly needed.

 

Technorati tags: , ,
software development, windows

Developers still miss VB6

11 Comments

A couple of years ago I wrote a piece on why Visual Basic 6 was frozen.

The topic is still of interest, and some reason reddit.com picked this link up recently, so the article has thousands of new readers.

If nothing else, it proves that developers still miss the old Visual Basic. Perhaps not so surprising; as I pointed out, it once had a reasonable claim to be the most popular programming language. That would not be true now; C# seems to be more popular than VB.NET, certainly among professionals, and I suspect Java is the number one overall (though these things are hard to measure intelligently).

Would I write the same article today? More or less, though the arrival of Vista and Office 2007 would make me state more forcibly that neither COM nor the Win32 API is dead. I still think that maintaining old-style VB would not have been feasible for Microsoft, except like FoxPro as a legacy thing and sadly now a dead end.

It’s also worth noting that VBA lives on, even though Microsoft is focusing on VSTO in its place. Except on the Mac, which is another story.

PS: I’ve fixed the comment feature on the article, so you can now have your say.

Technorati tags: , , , , ,
internet, web authoring

Sweet harmony between W3C and WHATWG

The new-ish W3C working group set up to create a new version of HTML has voted to adopt the work of WHATWG as its starting point, in particular the work on HTML 5. Here are the details from the co-chairs. This will speed up the process, and more significantly, brings together the W3C and the WHATWG. WHATWG was set up by browser vendors (Microsoft excluded) out of frustration with the W3C process and its abandonment of HTML in favour of XHTML.

Does this now make WHATWG pointless? That’s the obvious conclusion; but the group may not want to disband itself.

Thanks to Simon WIllison for the link.

 

Technorati tags: , ,
security, software, windows

Why you should keep UAC enabled on Vista

Ian Griffiths has a nice post on why you should not disable UAC, even if you are are a developer.

I’ve followed that advice and it works for me, though there are still one or two apps where I have to Run As Administrator.

That does not include Visual Studio 2005. Despite the warning which it issues, I find it works for me without it (I realise there are scenarios where this won’t be the case).

The intriguing thing is that (as Griffiths notes) even Microsoft is not solidly behind UAC. I’ve commented on this before.

Since there is still a myth that running Vista with UAC enabled results in an avalanche of intrusive dialogs, it’s worth popping up from time to time to say that it is not so.

Windows security affects all of us, even if you do not like Windows or use it. UAC (and IE7’s protected mode, which depends on it) is a step forward and worth supporting.

 

Technorati tags: , ,
borland, software development

JBuilder 2007 comes to the Mac

Codegear has announced a new JBuilder 2007 release which includes Mac, Vista and RedHat support (the earlier release only ran on Windows). It is to be made available later this month (May 2007).

There are three editions, Turbo (free), standard and Enterprise. Enterprise has “Team Server” features, with tracking and source code management; it’s not clear from the release how this ties in with existing team offerings from Codegear/Borland. The standard edition replaces both Developer and Professional editions in the previous range; the release says there is: 

Special upgrade pricing of $250 and new user pricing of $499

You would have thought this would be a free upgrade for existing JBuilder 2007 users, since cross-platform support should have been there from the beginning, but the release doesn’t say that it is. I’d like clarification.

This of course is the “JBuilder” based on Eclipse. I was interested in a discussion on one of the JBuilder newsgroups about Eclipse updates. Eclipse is a platform for add-ins, each of which is constantly being updated. The idea is that you run the update manager from time to time to get the latest version of each add–in, or perhaps install new ones. There are multiple dependencies with obvious potential for conflict. Borland’s JBuilder is a tailored build of Eclipse, and in consequence it is apparently dangerous to use the update manager. One user complained about this and drew the following comment from JBuilder expert David Orriss:

Do not try to use the Eclipse updater in JBuilder 2007. It can lead to problems, as you have seen. I’ll agree that it could have been documented better, but to try to effectively block the updater [which] (via plugins or code modifications) causes problems in the Eclipse platform.

It is a significant point. On the plus side, one of the attractions of JBuilder 2007 is that it offers a consistent, supported build of Eclipse unlike an uncontrollable open-source installation. On the minus side, blocking the update manager blocks the key Eclipse benefit: its extensibility and continuous improvement.

 

internet, multimedia, software development, web authoring

Mix Unmixed

Microsoft had a good Mix07. Let’s start with the provisos. Silverlight with .NET may have been announced, but it’s a long way from delivery, with alpha code just posted and no date set. Adobe already has a widely deployed cross-platform runtime with an embedded Javascript engine, complete with JIT compiler.

Next, Microsoft is miles behind its rival in the design world; it’s Expression tools are just now appearing, while products like Illustrator, Photoshop and Dreamweaver are de facto industry standards.

Third, Mix07 had its share of hiccups, not least the bizarre closing plenary. The theme seemed to be the future of advertising: XBox maestro Robbie Bach entertained us for a short time with clips of in-game advertising, following which came a lengthy and less than sparkling panel debate, only redeemed by pithy comments from Economist Publisher and Managing Director Andrew Rashbash who reminded us that editorial independence still matters, which is a relief.

That wasn’t enough for most delegates. A little way into the debate I became aware of a distracting bonging noise from somewhere behind me. It was the doors clanging as attendees headed for the exit.

I have never seen so many people leave a Microsoft plenary, and I can’t shake off the suspicion that something else was planned for this slot, but pulled at a late hour.

Never mind, this was nevertheless a good conference for Microsoft. It is all about a one-two Silverlight punch. Punch one is online video. You know the story: streaming internet video used to belong to Real, Apple and Microsoft until Adobe sneaked in with a marvellous “it just works” implementation in the Flash runtime. How can Microsoft now compete? Two ways: price and quality. On the price front, it is giving away space on its streaming servers, a more than generous offer that is likely to be widely taken up.

Another factor is codecs. Flash has two, H263+ and VP6. H263+ is cheap to implement, thanks to to support in FFMPEG, but the quality is poor. The newer VP6 codec, from 0n2, is equally high quality but according to Microsoft’s Forest Key, less efficient:

We are 20% better. At constrained data rates, or at HD data rates, because it’s computationally more effective, we can do a significantly better job . We can do HD on a significant number of machines. You will be able to do HD video with Silverlight. That is something that Flash can’t claim.

On2 may dispute this, judging from the claims on its site; but even if Microsoft is only on a par, that might be good enough, bearing in mind the low cost of encoding and delivering:

Expression Media Encoder is a batch processing tool for delivering media to Silverlight. It’s an enterprise scale product. We have a streaming server that is part of Windows server. It is very cost effective. By comparison Flash has a very expensive SKU.

says Key. Another plus for Silverlight video is ease of development. Program Manager Wayne Smith closed his demo with a jigsaw puzzle video, by which I mean a jigsaw image, pieces strewn everywhere, with each piece playing a segment of video. This is not useful in itself, but it nicely illustrates that in WPF video is just another graphics brush. This makes it easy to integrate video into an interactive application, with overlays, user configuration options, multiple simultaneous videos playing, and so on. Welcome to interactive broadcasting.

I am not personally a video person; I’m more interested in the programming side. The second Silverlight punch is the announcement of official cross-platform .NET, something I’ve speculated about for a long time, since before .NET 1.0 was released. Why is Microsoft doing it now? In one sense it’s an admission of failure: there will never be a Windows-only internet, thank goodness. For the rest of us it is good news.

Will Microsoft compromise Silverlight to keep the full WPF better? That must be a risk; but Key insists not:

That’s not our concern. We’re going to make Silverlight as good as possible. We shouldn’t artificially sabotage Silverlight to keep differentiation.

Overall it’s a good story, and accounts for the generally enthusiastic reception which Mix delegates gave to the opening keynote, which was as good as the closing plenary was bad. Pay special attention to what the guys from Major League Baseball talked about, as this is where it comes together into a compelling deliverable.

With Silverlight and Expression, I now think Microsoft will make real impact.

Technorati tags: , , , , , ,
Uncategorized

Canon Exposed

I attended a briefing/party in Monaco to mark the 50th Anniversary of Canon in Europe. At the exhibition I saw this display of the insides of a Digital Ixus – picture taken with an earlier model of the self-same camera:

At the event Canon announced the IXUS 950 IS, with 8.0 megapixels, 30fps VGA movies, and face detection technology (this is meant to improve focus, not recognize celebs).

But what did I learn about the insides of Canon itself? This was not an occasion for detailed analysis, though I saw some shiny new cameras and videocams, and had an chat about video codecs which was interesting in the light of my discussions with Microsoft about Silverlight video last week.

Some quick stats though: Canon says it launches 100 products per year and takes out 2000 patents. Since 2004, EMEA (Europe, Middle East and Africa) has been its biggest single market, accounting for 32% of revenue (9 billion Euro in 2006). According to COO Tsenuji Uchida, Canon’s market capitalization is the 6th largest in Japan and 81st worldwide, at around 9 trillion yen. It is the world’s biggest supplier of digital cameras (by value), but only number 4 when it comes to video cameras.

A large part of Canon’s business is in paper handling – printing, scanning, copying – and I would have liked to ask how it expects to thrive in the paperless office era, when and if that ever arrives (I think it will), but didn’t get an opportunity. The answer I guess is in the new corporate logo “We speak image”. Another interesting question is about device convergence: I’ve heard that Nokia is now the biggest single supplier of cameras if you count the things built into mobile phones. When phone cameras become “good enough” for family snaps, will that hit Canon’s market? Possibly, though as photograhy (still and video) becomes cheaper and easier thanks to advancing technology, the hobbyist market is growing and will never be satisified with a phone cam.