Monday, June 25, 2007

Misleading and uninforming

Recently, there was an article about Microsoft's war with Open Source in the Financial Express.

It was interesting, though quite misleading in a lot of facts.

Firstly: TCO (Total Cost of Ownership) studies with Linux are widely quoted, but usually fairly incorrect. The first study was carried out by the Yankee group, who were funded directly by Microsoft, and clearly, Microsoft came out the winner. Later studies are equally lopsided, and are debunked quite easily. A full report is available from Levanta's website. If you selectively set up the test environment, you can show anything you choose. So after all the criticism worldwide, if Microsoft sticks to its line of lower TCO, it is both misleading and uninformed. If there are new studies, then those should be presented. Otherwise, the Microsoft argument of lower TCO is plain marketing, with no real evidence. At this point, companies that are deploying thousands of servers should ask themselves what they know that Google doesn't. Google runs all its servers on Linux, and they have a lot of them, so clearly cost (and TCO) are very important to them. How important is cost: Google designs its own power supplies to get the most cost efficient machines.

Second: Microsoft's attempts to look like they're a part of the Open Source movement, with the launching of their Codeplex site is fairly meaningless. Free Software participation is measured by how much code is released as Free Software. Microsoft's contributions here are zero at best, and negative at worst. Their efforts at pushing competing standards (OpenXML to counter ODF), breaking standards (with the browsers), or forcing developers to reverse engineer their work (SMB version 2, because Samba was a better SMB server than Windows) is well known. Microsoft has never contributed any body of code to Open Source. Sun has released the Java language under GPL, IBM released many projects, including Jikes, and Eclipse, and are at the forefront of corporate Free Software development.

There is also the sticking point of interoperability. Microsoft's recent attempts are to pitch itself as a company that is keen to be interoperable with Linux. Nothing could be further from the truth. Two products at Microsoft form a bulk of their profits: Windows and Office. With ODF, and OpenOffice, Microsoft Office's user base is being steadily eroded. OpenOffice is very capable, and does nearly all the functions that you would expect (Word Processing, Presentation Software, Spreadsheet, ...). It runs on Windows, Linux, and Mac OS and is free of cost. You can download it and start using it right away. It might not have all the millions of features that Microsoft Office has, but it has all the ones that most people use. It can even open files written in MS Office format, which has traditionally been the lock-in for MS Office. To restore the lock-in, Microsoft is pushing its new document format called OpenXML, which, despite its naming, is neither Open, nor completely XML. If you have all the time, you can read why OpenXML is not really Open, or XML. Only Microsoft can implement it completely, as it depends on proprietary information, and breaks many existing standards.

The competition to OpenXML is Open Document Format (ODF), which is a published spec, that anyone can implement. It was created by the Oasis Group, which is the same team as the OpenOffice project, and has been successfully implemented by other independent groups. It is in Microsoft's interest to kill ODF, before its adoption is widespread. If Microsoft were to really support interoperability, they would implement ODF in Microsoft Office. This would be great for interoperability, but terrible for Microsoft's shareholders, since the Office cash cow would essentially vanish overnight.

The other place where Microsoft's Interoperability claims fall flat is its adoption of Web Standards. Microsoft's Internet Explorer is among the least standards compliant web browsers. This causes endless amounts of grief to web developers who write correct HTML, only to find that Internet Explorer renders it incorrectly. So they have to struggle to make their page work on IE, when every other browser (Firefox, Opera, Safari) does it just fine. Internet Explorer 6 was absolutely terrible at following standards, and Internet Explorer 7 is also bad with standards, and terrible at correctly implementing the standards it does follow. The previous links aren't about security bugs, but just bugs in correctly showing the web pages. And I should remind you that Safari (by Apple), Konquerer, Firefox (Free Software) and Opera (commercial) all are excellent at web standards. Either Microsoft lacks the incentive to write a good browser, or it lacks the people.


In the article, CESC Ltd's chief information officer Subroto Das makes some fairly uninformed claims. The first claim is that using Linux would mean that you have to release your code for free. This is completely false. Using Linux does not mean that you have to release all your source code under an Open License. Many companies use Linux desktops and servers, and maintain giant bodies of proprietary code: Google is a prime example. All their search code is proprietary, and they don't have to release it.
He claims that companies would have put their databases under third party control by using Free Software. This is also absolutely false, and a remarkably callous statement.
Lastly, he claims that "there is no such thing as free software". You have to pay Red Hat or Novell, or someone. Hey Mr. Das, why not download Red Hat's software free of charge, or if you can download 1.6Gb of software, here's SuSE, from Novell, or Open SuSE from Novell. It is absolutely free, in that you pay nobody to download it, or use it. So there is some such thing as free software.

Finally there is a gem of a quote from Microsoft India's Radhesh Balakrishnan,
Microsoft clearly stands out with a clear advantage of being able to give a short term and long term vision and roadmap.
Well, if Microsoft has such great vision, why is Vista floundering so badly? Vista sales are absolutely terrible, and corporates are not going to move to Vista anytime soon. Home users are either sticking with their old machines, or reinstalling Windows XP on their Vista computers. Why is Microsoft's online presence so shoddy, their web services are poor at best, with Maps Live being the only saving grace. Why is Linux on so many machines, starting from desktops, to the largest supercomputers, and everything in between? Developers are making web applications, or using Java, or other free tools. The amount of developer interest in .NET technology is very low. And why is Microsoft's share price so flat with little or no prospect of growth in the near future?

Microsoft has a clear vision for all its customers, and that vision involves a proprietary lock in for an unlimited amount of time. Indian companies, much like companies elsewhere are looking for a cheaper, safer, and competitive solution that ensures that they're not at the mercy of a single supplier. There is plenty of Free Software development and use in India. No matter how hard Microsoft's spin machine tries to change the facts, it is established that Free Software is a very viable solution for many tasks previous handled by proprietary software.

Dynamic companies are evaluating Linux to take advantage of cheaper, higher quality software.