| Flash not detected - Why Adobe should worry... |
[13 Jul 2009|08:14pm] |
If you visit riastats.com and check out the line graphs you'll see an interesting trend burried at the bottom of the page. Three months ago less than three percent of visitors didn't seem to have Flash installed at all. Today the number is about 4%. There may be any number of reasons for thiis very small 1% change. But one of them may be that Vista, unlike XP, ships without Flash and a small number of people are not getting around to installing it. Another reason may be more iPhone/iTouch visitors or browser add ins like "no script."
I doubt there's much Adobe can do about this other than hope their open screen project goes well. In the near term HTML 5 will have no real impact and if someone takes the trouble to install Silverlight they probably have even more reason to install Flash. if the trend continues for a long time the gap between Silverlight adoptions and Flash may not seem so big. At some point the "Flash is everywhere" argument may just go away.
|
|
| Microsoft to Adobe: We are going to Kill You! |
[11 Jul 2009|12:12pm] |
The competitive posturing coming out of Microsoft is as smooth and subtle as always:
http://news.cnet.com/8301-13860_3-10284362-56.html
Some nice quotes about Adobe's ability to offer a general-purpose Web platform via Flash/Flex:
"I don't believe they have the assets or the organizational structure," he said. "That's what we do for a living at Microsoft." Abu-Hadba said Adobe would be better off picking a specialty and sticking to it. "I don't think they will exist in 10 years in the form they are today," he said. It's a bold statement, he agreed, but added how unthinkable it would have been to predict in 2000 that Sun Microsystems would go away.
Microsoft evangelists talking about Sun going away reminds me of the Republicans boasting about defeating Communism when the Soviet Union fell apart. But I don't think Java has really gone away...
It's too bad there isn't an immediately obvious financial penalty for arrogance.
|
|
| What will Google bundle into their OS? |
[10 Jul 2009|07:50am] |
|
The iPod works with iTunes and the iPhone with the app store. That's part of how Apple makes money. Their devices ship ready to work with the services they need. So what will Google bundle into their Google Chrome OS. (Since they have no desktop monopoly they can legally bundle anything and everything they want.) My guess is that out-of-the-box you will get fully synchronized off-line/on-line work with all of Google's online applications including Wave. Google has a year to make their Chrome browser as good as having Microsoft office installed locally. Can they succeed at that? Maybe. The intersting part is that Adobe is a partner and Adobe has their own online services... Stay tuned...
|
|
| Adobe is Half Open or Half Closed? |
[17 Jun 2009|07:19am] |
Sometimes you just have to laugh:
http://www.contentinople.com/author.asp?section_id=450&doc_id=178083
Adobe has published and licensed for use their RTMP streaming audio, video, and data protocol. But reading Ryan Lawler's account you have to ask if they are really serious. First, the protocol was already reverse engineered. Second, features like adaptive bit-rate streaming are not published, and third, the secure and Peer-to-Peer versions of RTMP remain proprietary: rtmpe and rtmfp.
I'm sure the good folks at Adobe think they have accomplished something by publishing RTMP. But to anyone outside Adobe the specifications for streaming in/out of the Flash player are still mostly closed.
It's too bad for Adobe that their internal vision doesn't match reality outside the company.
|
|
| Microsoft to the Open Web: Use Silverlight! |
[19 May 2009|08:57pm] |
Tact and diplomacy are not strategies the marketing people at Microsoft deploy very often. Why should they? They work for the winning team and are paid to be competitive. So Microsoft's announcement that Silverlight is installed on more machines than Firefox, Safari, and Chrome combined was a nasty little slap upside the head for people committed to the open web. http://www.techcrunch.com/2009/05/18/microsoft-says-silverlight-installed-more-than-firefox-safari-and-chrome-combined/
Here's why it burns so badly: After defeating Netscape, Microsoft took its best people off their IE and CSS/scripting work and put them on WPF (Avalon) – the new presentation layer for Vista. Eventually that work led to the Silverlight browser plugin (or mini-WPF). Why did they move everyone to WPF? One reason is that they needed their best people for Avalon. And, Microsoft says HTML/CSS was not the way to build the next generation presentation layer for Windows. (It's hard to argue with that!) But that doesn't explain why IE 6 was left to rot for so many years. For that I think you have to look at what happened with VB script and ActiveX. First, VB script in the browser was not a success. Second, ActiveX backfired. Instead of tying the browser more closely to the Windows operating system, it just made the browser more capable. For example Microsoft added an Active X component to make XML over HTTP requests without reloading the Web page. They did that to support Outlook Web access in 2000. In 2002 Mozilla incorporated XMLHttpRequest() and later Google wrote code so that you could use the same XMLHttpRequest() API within IE. Not only did this lay the foundation for AJAX but it helped propel the belief among other browser developers that progress in providing new, standard, cross browser features could still be made despite Microsoft's lack of interest. (Another example is the Canvas tag and Google's ExplorerCanvas library that uses IE features so you can use the canvas API in IE.) For Microsoft it must have underlined that new features in the browser could be wrapped up in other people's APIs making any linkage with Windows irrelevant.
I think once Microsoft realized there was little value for Windows in extending the browser they completely lost interest in adding new Web application features. Today they say WPF provides a much better model of how to develop applications with rich user interfaces and that browsers should remain more document centric. For the other browser developers this is infuriating. Both Google and Mozilla have shown that JavaScript performance can be dramatically improved but are to some degree powerless to move new Web standards forward without Microsoft. Just how powerless can be seen by what happened around JavaScript 2 at ECMA.
So browser developers are in an almost no-win situation. They want to move forward without Microsoft but must have some cooperation from Microsoft to implement new standards. At some level that means they are forced to co-operate with Microsoft – regardless of how indifferent Microsoft is to new proposals. Look at Brendan Eich's language before the ECMA showdown and afterwards (Harmony) and you know what I mean.
Now along comes Microsoft with its mini WPF plugin that is literally a slice of .Net. It has its own .Net APIs and .Net display technology. Google, Safari, and Mozilla's ambitions of the browser as Web application platform just took an ice cold shower. Microsoft isn't going to move very fast with IE. But they are going to push hard to get their mini WPF plugin into every browser so .Net developers can extend .Net's proprietary reach into the Web.
I can only imagine two reactions from the people who are deeply committed to the open web. One is that they turn their ire on Microsoft. The other is that they try to maintain some sort of détente with Microsoft and turn their ire on Adobe. Of course Flash wouldn't be such a big part of the Web application space if Microsoft had kept moving forward in the browser market. But that won't stop the Flash bashing.
Either way Microsoft is saying it's .Net vision, and all the rich features it built into it rather than IE, are winning against the Open Web. It must really burn.
|
|
| Defeat and then what? |
[18 May 2009|10:15am] |
|
Reading about the Tamil surrender makes me wonder about two things. What lessons will the Israelis and Palestinians take away from the bloody defeat and what will happen to the Tamil population? Will Israel be emboldened by the Tamil surrender despite American pressure? Will the Palestinians reappraise the value of launching rockets into civilian areas after seeing the response to the worst Tamil tactics? If only surrender meant that the suffering was over. I'm not sure it does.
|
|
| riastats trends... |
[12 May 2009|07:05am] |
I thought it might be interesting to return to the riastats page. I visited it a little over a month ago. The numbers then:
- Silverlight Not Detected: 76%
- Java Not Detected: 26%
- Flash Not Detected: 3%
Today's numbers:
- Silverlight Not Detected: 71%
- Java Not Detected: 28%
- Flash Not Detected: 3%
The stats are only from 56 sites and Adobe and Microsoft claim higher numbers of installs, but the over-all trend is not surprising. Most new PCs come with Internet Explorer as the default browser without Java and there is little to drive users to install it. Will Oracle care about that enough to spend money on it? It's hard to imagine.
Of course Microsoft is willing to spend and spend and spend:
http://www.contentinople.com/author.asp?section_id=450&doc_id=170168
According to Ryan Lawler they will pay Dell to include Silverlight along with the Windows Live Messenger beta on new Dell PCs. I wonder what the EU will think of this. Is it simply bundling by other means? Is it OK to pay monopoly money to bundle software on new machines but not OK to bundle for free? It will be interesting to see if the EU cares about this deal or not.
On the other hand, I imagine Adobe is doing everything it can to help organizations like the BBC, NYTimes, MLB, and ABC to use their technology - so far with some success.
I guess if you like cock fights this must be fun to watch. For the rest of us, not so much.
|
|
| Where is Adobe's NanoJIT now? |
[03 May 2009|09:06am] |
Back in September 08 Brendan Eich wrote:
"TraceMonkey is only a few months old, excluding the Tamarin Tracing Nanojit contributed by Adobe (thanks again, Ed and co.!), which we've built on and enhanced with x86-64 support and other fixes."
So I guess that means that once harmony struck, Adobe and Mozilla's partnership was effectively eliminated and Adobe had to proceed on alone with their own JIT? If that is what happened you really have to give Microsoft and friends at ECMA credit for destroying an effective industry partnership between Mozilla, Adobe, and the academics working on tracing.
But it also begs another question: what is Adobe doing with their NanoJIT? Will tracing come to Flash? Of course I have no idea, but I think Adobe has an opportunity here. They can combine optional typing and the other unique features of AS 3 along with dynamic language optimizations like tracing in order to get close to Java or C# level performance. If they can do that, and it is a big IF, then their ActionScript 3 language might be able to offer something new and exciting: the advantages of both static and dynamic languages in a single language.
If they can do that - and I have to admit I've never thought of Adobe as a compiler company - then server-side AS3 is the next logical step. Imagine writing complete applications in Flex and you get the idea.
There is only one problem. Tamarin is written in C++. In the Java application server world ActionScript 3 should run in the JVM. Can Adobe do both? I doubt it somehow...
Adobe has been promising a better performing Flash player for mobile devices for a while. They are supposed to announce something this summer. Mobile devices don't have the memory and cycles available to easily make dynamic languages like JavaScript perform well. So I think the mobile announcements Adobe makes this summer will be a kind of acid test. If a NanoJIT for their mobile player appears and provides Java or C# like performance then we can start to think about Adobe as a successful compiler company. If they don't then we can think of them as being more like Macromedia when AS 2 turned out to be syntactic sugar on top of AS1. Macromedia's problem back then: lack of resources.
|
|
| What Oracle said... |
[26 Apr 2009|06:51pm] |
Why don't people believe what CEOs say? When Adobe acquired Macromedia they said they wanted Flash. So some people blogged that it was really all about Dreamweaver or something else and that Adobe should shut down Flash. Why? Years later it's obvious that aside from making the Creative Suite into a bigger revenue source, it was indeed about Flash. Now Oracle is going to buy Sun. What did Oracle say the aquisition was about?
"Oracle will be the only company that can engineer an integrated system, applications to disk, where all the pieces fit and work together so customers do not have to do it themselves."
And its true. That's largely what it is about. Where things get interesting is when you ask how Oracle is going to do that. What investments in Java, Sun's hardware line, even MySQL are they going to make in order to have that one seamless offering?
Now consider who the competition is for that one seamless offering. Sure IBM and SAP are competitors. But Microsoft is the elephant in the room and Microsoft's Windows server is winning. And Windows/.Net is winning on cost and depth of the software/.Net stack that you get for that low initial server OS cost. What is Oracle going to do to fend off that challenge? We'll see, but I think it's a lot more interesting to imagine what Oracle might be able to get right than to imagine they will jettison everything and then do everything wrong with what's left.
Don't believe me? Have a look at how their BPEL process manager and ODI came out. They work well and integrate brilliantly with Oracle's enterprise offerings.
|
|
| Silverlight not detected - Why Adobe should still worry |
[04 Apr 2009|08:13am] |
|
If you work at Adobe you might take some solace in this site: http://www.riastats.com/. When you first visit the site you see three pie charts that show the percentage of browsers that already had Flash, Silverlight, or Java installed when they visited some sites. If you look at the "not detected" slice of each pie today you'd see this: - Silverlight Not Detected: 76%
- Java Not Detected: 26%
- Flash Not Detected: 3%
If you work for Adobe this is good news, right? Unfortunately, it really isn't. If Silverlight was produced by any company other than Microsoft you could easily speculate that it could never catch up with Flash. Developers and designers who are aware of these numbers would decide to use Flash and that would be that. But Microsoft is a massive monopoly and so can spend enormous funds on subsidizing the adoption of Silverlight. To understand what that means all you have to do is press a button on the riastats page. Instead of looking at the pie charts which show an average over the last 30 days, click on the "Line" button at the top right of the page. As I write this it shows Silverlight was at 86% undetected three months ago and has steadily fallen at a very linear rate down to 73% undetected today. Of course there is no reason Silverlight has to show a 14% drop in the undetected catagory every three months. But with enough media outlets using Silverlight my guess is that the downward trend has more legs than anything Move Networks could have accomplished on their own. Microsft is unique in that it can lose money in a market place for years while slowly grinding down the competition. It is the only company with a software monopoly and that makes it unique.
When you look at the numbers in riastats you quickly realise that Silverlight may achieve the same level of adoption as Java or better by the end of 2009. That should truly frighten Adobe.
And, as for Java? Java's undetected rate is actually increasing. If the stats are accurate, fewer people have Java today than had it three months ago.
|
|
| Disturbing News |
[21 Mar 2009|11:06am] |
In science fiction movies and TV show episodes big things often happen light years away that can't be fully explained but are essential to telling the story quickly. So characters tell us there is "a disturbance in the force" or are shocked when an unseen starship full of Vulcans is destroyed off stage by a big amoeba.
For those of us who seem to still have secure employment, watching the news is a little like watching the opening scene in a bad science fiction episode. We know something big is happening but we really don't get to see what it means. And, unlike TV episodes that end in an hour, the sci-tech and economic news looks a lot more like a multi-year sci-fi epic. But it isn't. The writers of this story are politicians, CEOs, shareholders, and above all economics. And the story we are being told is one of ongoing concentration of power in the hardware and software industry. Cisco continues to succeed while other vendors like Nortel fail. Someone – perhaps IBM – needs to buy Sun before it falls in on itself. Microsoft continues to use monopoly revenue to erode most of the software markets it enters. Dreams of an open source rebellion are just silly. Microsoft's lock on big chunks of the enterprise server market is only increasing and even if IE falls to 50% of the market that means open Web technologies aren't going anywhere quickly.
Despite all the ideologically driven nonsense in the traditional advertising driven media and on the Web, the only thing that can reverse the current trend and preserve a reasonable level of competition in the market place is government interference in two areas: concentration and intellectual property. After so many years of advertising driven media and Republican governments in the U.S., I just don't see how it will happen.
The Web set big companies with expensive and proprietary ambitions back on their heels. Back Office died and it took Microsoft years to resuscitate it under different names. Everyone had to mostly conform to the simple standards of the Web. Today, the opposite is happening. The Web is being set back on its heels as proprietary technologies like .Net are aggressively pushed via plugins into the browser.
Before things get a lot worse we need a correction. We need it soon before a difficult economy forces more concentration and increasing market monopolization. But we aren't going to get it. More likely we are going to watch five years of increasing concentration of control and commercialization of the Web. I'm not looking forward to it.
|
|
| Anyone but Microsoft |
[05 Sep 2008|12:26pm] |
Is there any serious Web company out there that wants to work with Microsoft? Sure, if you’re a start up, then being purchased by Microsoft is a nice exit. Working for Microsoft or using some of their technology is one thing, but can anyone seriously entertain really working WITH them on the Web?
If you use Internet Explorer today to visit yahoo.com you’ll get this banner message: Yahoo! Recommends upgrading to the NEW safer, faster Firefox 3 - FREE. If you click on the banner you end up here:
http://downloads.yahoo.com/firefox/
Of course Yahoo is not too happy with Microsoft these days. If Microsoft ever buys Yahoo the use Firefox message will change. But today, Yahoo’s stock may be way down but their portal – courtesy of agreements with numerous ISPs – is the starting point for services like E-mail and instant messaging for millions of people. Microsoft’s hostile run at Yahoo hasn’t made Microsoft any friends outside of Wall Street.
Microsoft has worked with NBC for a long time. NBC used Microsoft’s Silverlight plugin as part of their streaming solution for the 2008 Olympics. So you’d think that when NBC was ready to start streaming regular sports events like the NFL’s Sunday Night Football, they would continue to work with their close partner. They chose to work with Adobe instead: Instead the NFL chose to work with Adobe:
http://www.alleyinsider.com/2008/9/nfl-nbc-tap-adobe-s-flash-for-live-football
Adobe has also been in Microsoft’s sights. Microsoft’s, Silverlight, their sort-of-WPF plugin, is designed to fend off Flash as a Web platform and promote using the .Net stack more widely on the Web. Microsoft is also taking a run at Adobe’s PDF with XPS.
Given Microsoft’s behavior, Adobe can never really work with them. Instead, Adobe is investing their Creative Suite profits in developing Flash as a platform and staying out ahead of Microsoft’s sort-of-WPF plugin. Adobe keeps adding value to Flash video and is getting traction with Flex and AIR. Meanwhile, most .Net developers haven’t had much reason to do anything with Microsoft’s plugin yet. Most .Net developers continue to focus on other ways to deliver applications using ASP, SharePoint’s APIs, and Visual Basic.
Microsoft sat on IE 6 for so many years that Web developers are pretty much fed up with them. So improvements in JavaScript performance from Mozilla and Google are causing a lot of excitement and IE 8 is causing a lot of yawns. Firefox 3.1 will have something new – a tracing JIT. You can read about how that will make JavaScript run two or three times faster in Firefox here:
http://andreasgal.com/
For some tasks JavaScript may run more than ten times faster than in IE 7 and five times faster than IE 8.
Then along comes Google with a beta of their new browser. Its JavaScript engine is also much faster than IE 8. It’s designed from the ground up to run Web applications – both online, offline, and safely.
Microsoft has always tried to work closely with hardware vendors. Today, when I start up IE on my new Dell I go to the Dell/Google home page. Government action - especially in the EU - has made it impossible for Microsoft to do anything about it. But at least it is still IE, right?
When Google Chrome gets out of beta, I’d guess Google will ask Dell to make Chrome, and Google’s services, the default browser and default home page on every Dell that ships. With total annual revenue above last year’s 16 Billion, Google can afford it. Imagine every new user of a Dell computer never seeing Internet Explorer. After all these years we’ve grown accustomed to the idea that what Microsoft wants to ship with their OS is what everyone will use. But it may not be true much longer. Microsoft cannot ship Silverlight with their OS without inviting a nasty and very expensive court battle and they can’t do anything about Google paying Dell to ship Google Chrome. Things have started to change.
At some point, no matter how big you are, if you make it impossible for anyone to work with you, then they will find a way to hurt you. It may take years. But Microsoft, a company convicted of illegal business practices on two continents and under a lot of careful scrutiny, is about to face competition of a type they haven’t seen in a long time. Too bad for them there are so many people out there who hope that competition bites them hard.
|
|
| Unsound for the Web? |
[22 Aug 2008|06:52am] |
|
Brendan Eich writes> "Some ES4 proposals have been deemed unsound for the Web, and are off the table for good: packages, namespaces and early binding. This conclusion is key to Harmony."
See:
https://mail.mozilla.org/pipermail/es-discuss/2008-August/003400.html
Unsound for the Web? Off the table for good??
I understand there are real challenges with intrinsic and with dynamic code loading, but "unsound for the web"? That's a big leap. Maybe he could have written "too difficult for the next generation of browsers" and "off the table for now and possibly forever" or something more modest. But "unsound" for the web? I don't think so. Something good happened in Oslo. There was agreement to make modest progress. But, something bad seems to have happened as well. It's as though all the edge cases are all that matters. For example, pointing to the fact that optional typing in ActionScript 3 does not always produce enhanced performance is true but not that significant. In practice typing provides dramatically improved performance. Along with other improvements in Flash player 9, performance is way ahead of Flash Player 8 often by a factor of 4 or 5.
Not worried about improving performance with optional typing because of some tracing jit benchmarks? That's a big bet to make.
It is possible for committees to come to a consensus without losing the independence of the voices on it. When I read "unsound for the Web" I started to wonder if maybe the committee agreed to speak with one voice as well and, I must say, I'm not enjoying the sound it is making.
Update: Some comments by Brendan Eich:
http://www.itwriting.com/blog/833-ecmascript-4-deemed-unsound-for-the-web.html#comment-105527
Update: Here's a post describing trace-based JIT compilers:
http://andreasgal.com/2008/08/22/tracing-the-web/
|
|
| Time to Retire ColdFusion |
[11 Jun 2008|05:19pm] |
I've been reading a few blogs recently about what new features should be included in ColdFusion. Stepping back from the small details and looking at Adobe's offerings it seems like now would be a good time to retire ColdFusion entirely.
Well, at least the name.
At best cold fusion (two words) is a theory that low-energy nuclear reactions are possible. For two decades of funded research no one has done anything useful with the theory. At worst cold fusion is bunk. For two decades now ColdFusion has been a single word name for an application server from Allaire, Macromedia, and now Adobe. For me, its name conjures up the excitement and ultimate disappointment with the idea of low-energy nuclear reactions. ColdFusion and Adobe's application server strategy need a makeover.
What should Adobe call the next release of ColdFusion? That's Adobe's problem. For now I'll just call it RealFusion. What should RealFusion look like? It should include a fully licensed version of LiveCycle Data Services, an XMPP server, Flash Media Interactive Server, a fully integrated ECMAScript 4 engine, and whatever else Adobe wants to dump in there. RealFusion should not only fuse Adobe's servers together in one cost-effective package it should also make it possible to build and scale powerful http/rtmp applications using one application server. It means Adobe evangelists can post articles with demonstration applications that don't have to be created to run across multiple servers with multiple languages. It would represent a real bit of innovation around JavaScript and it would protect JavaScript and ActionScript from what Microsoft is about to do to the Web via WPF/E.
Adobe, its time for some real fusion.
|
|
| Adobe AIR: Turning Web Applications Inside Out |
[25 Feb 2008|10:13pm] |
All the enterprise applications I use today are Web applications. Their interface is usually mediocre but they work. Part of the problem is that Web applications have to work inside Web browsers. But what if all the developers of those applications could turn things inside out? Instead of their Web applications living inside the browser, what if the browser was just part of what lived inside their Web-delivered applications?
Desktop applications have had some of this capacity for a long time. For example, you can write a Windows application that controls Internet Explorer, accesses local files, and can connect to the Internet. But then you are still writing a desktop application for one operating system. You're still stuck with all the extra development work, distribution problems, and maintenance issues that writing native applications for a single OS requires.
Adobe's recently announced Integrated Runtime (AIR) inverts the normal structure of Web applications. It puts the developer in control by providing a desktop runtime that runs in Windows, on the Mac, and in Linux. Developers can write their applications using Web technologies like JavaScript, DHTML, Flex, and Flash. AIR lets your code control AIR's Web browser, PDF player, and Flash player as elements inside your application. At the same time your AIR application can access local files, a local SQL database, high-level network protocols or binary sockets if you need them.
Now, let's check off some problems this solves for developers:
- The user interface is not locked inside a browser window. You can have multiple collaborating OS style windows if you need them. There is no browser chrome, unless you add it yourself.
- Browser compatibility issues are gone because AIR contains its own Web browser based on the open source WebKit browser, so DHTML is easier.
- Cross Scripting works. With your application in control you can load a local or remote PDF into a PDF control inside your application. Then use the PDF API to control what the user sees. For example your users can synchronously view exactly the same PDF, and with a synchronized server connection, even control what each other sees. The same goes for Flash Movies and Web pages.
- Desktop integration works. A user can drag and drop files and data just like they do in ordinary desktop applications. For example imagine providing a richer interface into your Wiki or blog software. The user drops an image on a page and the image is automatically laid out in place. Say goodbye to making your users browse, upload, and link before they can put a picture on a Wiki page.
- Users don't have to be connected to work with an AIR application. You can save their work to local files and to a local SQL database.
- Local storage also means your application doesn't have to constantly send updates back and forth to your server even when it is connected. You choose when the right time to synchronize data occurs.
- The auto-update API means you can post your new, digitally signed, versions of your application and users can automatically pick them up.
Adobe's competitors will complain that AIR applications are less secure than AJAX, Flash, and Flex applications executing inside a browser. And of course they are right. AIR applications are running on AIR which provides controlled access to the local machine. So loading an AIR application means the user is granting more trust to the developer than they do to browser-based applications. But AIR applications are not like downloading and installing native applications, so Microsoft's statements about AIR security are hard to take seriously. At first blush Adobe looks like it has made some reasonable choices on security. One concern that has been raised is if anti-virus software will be able to scan AIR applications when they are being downloaded. Over time we'll find out. A stronger and better competitive comparison on security may be with Java's security model. We'll see how that goes.
In the mean time, developers working on enterprise applications should stop and take a good hard look at AIR. If you want powerful, effective user interfaces and you know JavaScript and DHTML or ActionScript/Flash/Flex, then you don't really need to learn a new language or a truck-full of new APIs to get there.
|
|
| The European Union, Microsoft, and the Free Market |
[24 Feb 2008|01:05pm] |
Just as there are no free lunches, there are no truly free markets. Any vendor who buys you lunch or participates in a market is always looking for an advantage. Competition makes markets the vibrant and innovative places we want them to be. But paradoxically, when competitors succeed in developing competitive advantages they start to threaten market openness and competition. In software markets, complex and evolving dependencies are regularly used to lock out competitors. Nothing could make this more obvious than Microsoft's recent release of Windows and Office APIs and protocols. Reading them is a withering experience. No developer seeing them for the first time walks away from them without a strong feeling of nausea. The sheer work required to integrate competing products with Windows and Office at a deep level is beyond what any single corporation can afford.
New software markets have been created or expanded by new companies bringing innovative products to market. Novell created the office server market. Their Novell Directory Server (NDS) was a remarkable achievement. Netscape made the Web browser a Web application platform and were the prime movers behind the Light Weight Directory Access Protocol (LDAP) which is now a standard. Real Networks introduced us to streaming media. All three companies are now defunct or marginalized to the point of irrelevance. Today, Microsoft owns the enterprise office server market. Their Active Directory product is everywhere. SharePoint, which takes Office/Server integration to a new level cost Microsoft over 2 billion to develop and is taking off. And every CIO I talk to is counting his or her pennies for the day when he/she can buy Microsoft's Unified Communications.
If no single company has the wherewithal to compete with Microsoft in its chosen markets maybe they could collaborate via Open Source initiatives? But while Linux has had modest success there is little evidence of this happening on a scale that really competes with Microsoft.
So if there are no free lunches and no free markets, so what? Maybe the idea of free markets should be abandoned? Maybe we should just buy Microsoft stock and relax. But, I don't think so. There is always a balance to be struck between freedom and necessity. And elected governments have a role to play in guaranteeing a certain level of market freedom. In fact the difference between organized crime and an open market place is that an open market place mostly follows the rules set down by elected regulators.
In the United States this simple truth has been mostly lost. The American government - especially under the influence of the Republicans - sees its role as promoting the success of very large American-based corporations around the world. What is good for Haliburton, General Electric, and Microsoft is good for America and the world.
So, it was left to the Europeans to try to reign in Microsoft in order to protect software markets where there is still some freedom and openness. It has taken them since 1998 when Sun Microsystems first filed a complaint until now (2008) to force Microsoft to start behaving a little differently. That's a decade, and a decade is a long time to wait! Many people believe the EU has failed. By delaying as long as it could, Microsoft has won. Their platform is so well entrenched in so many markets, their cash flow gives them enough leverage to move into any market they want to, and their open APIs just mean that their platform will be used more widely in newer markets. In the short term there's no question that all this is true. Microsoft has won decisive victories over the last ten years and is well positioned to win many more.
There is however some hope that the regulators in the EU, with help from a small group of people (Samba, FSFE), have really done something important. They have set real limits on Microsoft's behavior and opened the door to new actions which will further limit Microsoft's advantages. The first corporate beneficiary of this is probably Adobe. Microsoft has not been able to bundle Silverlight and its own PDF converter in Windows and Office. Instead Microsoft has to pay media companies to use Silverlight. But that still makes it more difficult for Microsoft and leaves more room for Adobe to compete. Anti-trust action has also protected Google - for example making it possible for them to work with Dell.
The EU has not only set limits on Microsoft's behavior its actions also forced Microsoft to pay off all the original corporate complainants. Microsoft paid out something like 3.8 billion to Novell, Sun, and Real. Sun Microsystems got something like 2 billion dollars. What is Sun doing with all that money? One thing is that they are trying to take open source initiatives like MySQL and Dojo to new levels. Look for Sun to re-enter the enterprise office/communications server market in a more serious way over the next few years - paid for in part by Microsoft under pressure from the EU. And watch as Microsoft is mostly powerless to make protocol or API changes to marginalize or destroy them.
Fortunately, the EU and American and European courts are not finished with Microsoft. In the United States the "Vista Capable" lawsuit is now a class-action. In Europe the EC is investigating Microsoft's attempt to bulldoze OOXML through standards bodies, Opera's complaints about IE, as well as ongoing complaints around Office.
In fact there is discussion in Europe of bringing in legislation to allow anti-trust class-action lawsuits. The long term financial damage to Microsoft of European class-action suites is hard to predict.
After ten years, we've only seen the beginning. Human creativity is hard to repress. In markets it needs some level of freedom and openness to flourish and that can only be provided by elected government regulation. Maybe now that there is a little more of that, we can hope for more sustainable, vibrant, and innovative software markets.
|
|
| The Yahoos at Microsoft |
[10 Feb 2008|04:41pm] |
In the consumer communications market - E-mail and Instant Messaging - Microsoft and Yahoo are the leaders. Many ISPs have given up on providing their own communications services. They've partnered with Microsoft or Yahoo. The cost of switching communications providers is high. If Microsoft buys Yahoo, many ISPs value-added consumer services will be owned by one company.
Yahoo has the most popular collections of places to visit on the Web. According to Bill Tancer Yahoo gets twice as many visits as Google. Microsoft is almost off the radar with one fifth of what Yahoo has and one third of what Google gets. In search Google is the clear winner with over 65% of searches while Yahoo and Microsoft together only account for over 27%.
Google's dominance in search and the way it has chosen to provide relevant and relatively unobtrusive ads within its services is driving ad revenue to Google. As long as Google continues to increase its ad revenue Microsoft will see it as a dangerous competitor. Google will continue to attract talent away from Microsoft and will continue to pose challenges to Microsoft in numerous markets such as the mobile, entertainment, and software as a service spaces. Thanks to convictions in the US and Europe, some of Microsoft's natural responses are prohibited. They are powerless to stop Google from doing deals with companies like Dell that drive even more traffic to Google.
So Microsoft's recent offer to buy Yahoo isn't anything like Adobe's acquisition of Macromedia. In that case two companies came together to make a technologically larger and more effective whole. And it isn't anything like Google's attempt to acquire DoubleClick to accelerate its advertising revenue.
Microsoft needs a way to counter how Google's dominance in search is driving increasing ad revenues. But, Microsoft's statements about getting more search engine rocket scientists are wishful thinking at best. It's really about making a disruptive move that plays one of two ways.
If Microsoft is not allowed to buy Yahoo then Microsoft must hope that the EU will not allow Google to buy DoubleClick. Yahoo will be damaged by the threat of an unfriendly takeover. If Yahoo initially accepts the deal it is stuck waiting for regulators. If it doesn't accept the deal it must face down a hostile take over. Either scenario creates fear, uncertainty, and doubt for everyone but Microsoft and provides more impetus for the best people to leave Yahoo. A pending takeover also creates uncertainty about what the communications services landscape will look like in the future. Would you ink a long-term partnership with Google in that environment?
If Microsoft is allowed to buy Yahoo it will assimilate Yahoo's users and services into Microsoft's online businesses. It will take time but eventually Yahoo will be forced to migrate everything it does to the .Net platform. The use of Flash will be replaced by Silverlight, C# will replace PHP, and a large contributor to open source projects will disappear. You can hear this in the words Kevin Johnson, President, Platform & Services Division chose in the Microsoft conference call:
"We have got clear line of sight to the synergies and the value creation we are going to unlock. We have got a clear set of principles and we are going to go through a thoughtful process with great talent from both Yahoo! and Microsoft to really make the specific decisions on how that lands."
Translation for the business speak impaired? The medium-term goal will be to move to one integrated service based on the "clear set of principles" set in Redmond. That means the platform will not be LAMP it will be .Net.
There are many reasons to acquire another company. One of them is to remove competing products from the market. That's really what Microsoft wants to do. They want Yahoo's users, Web properties, and advertising revenue in order to accelerate growth of their services.
In Gulliver's Travels Yahoos are described this way:
For if," said he, "you throw among five Yahoos as much food as would be sufficient for fifty, they will, instead of eating peaceably, fall together by the ears, each single one impatient to have all to itself."
I never could figure out why anyone would name a company Yahoo!, or for that matter why anyone would name a server ColdFusion. But the description from Gulliver's Travels seems to fit Microsoft's management. And its not surprising. With 50 Billion a year in monopoly-driven revenue they have to do something new to expand into every new software market and crush every serious competitor.
|
|
| Runtime Religion |
[19 Jan 2008|12:12pm] |
Developing useful software that works reliably is difficult. How should the user interface work? Will anyone figure out how to use it? How can I build something that always performs well? Will the choices I make today doom the product in the future? When you start a project it’s a little like looking into the abyss. The future is always uncertain.
Whenever a developer first encounters a "bug" they also look into the abyss. What if they can't figure out what is wrong or can't figure out how to fix it without breaking other things? Some bugs are not fixed.
There are jokes about looking into the abyss such as Douglas Adams's Total Perspective Vortex. But chaos is what software developers spend their lives trying to avoid.
At a conscious level developers work with data structures, design patterns, architectures and so forth. But behind all the processes, best practices, and virtual edifices we build is the threat of failure and chaos. Failure and chaos are reminders of death.
Here's an interesting interview with the authors of the book In the Wake of 9/11: The Psychology of Terror:
http://www.esi-topics.com/terrorism/interviews/SheldonSolomon.html
Does "Terror Management" explain something about platform tribalism whenever runtimes are compared? Like most things in the "soft" sciences like Psychology, Terror Management will always be something of a theory that is endlessly debated. But it seems to explain my experience of watching people "rally round the flag," preaching the Bible, weighing down the smallest projects with heavy project management processes, and defending programming paradigms.
I wish it wasn't true, but every day I see things that remind me of what Solomon and Pyszczynski describe. It also reminds me of John Mayer's song Belief:
"Belief is a beautiful armor but makes for the heaviest sword… It's the chemical weapon for the war raging on inside." - John Mayer
One of my favorite books is - of all things - David Flanagan's JavaScript, The Definitive Guide. It is written for many different readers. It doesn't assume you should know too many things before you open it and it doesn't oversimplify things either. It's an honest attempt to share real experience and carefully acquired knowledge widely. And it starts with this little dedication:
"This book is dedicated to all people who teach peace and resist violence." - David Flanagan
Corporate "evangelists" please take note. We already have enough fear and uncertainty.
|
|
| navigation |
| [ |
viewing |
| |
most recent entries |
] |
| [ |
go |
| |
earlier |
] |
|
|
|
|
![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small}
notboss's journal