Dell’s amicus brief in support of Microsoft totally unconvincing

Seattle PI’s Nick Eaton got us the PDF for Dell’s amicus brief in support of Microsoft’s appeal to delay the injunction banning Word. The PDF link is available via GrokLaw. It is extremely heavily redacted, and I do hope the redacted part is more convincing than the non-redacted part on the hardship felt by Dell if the appeal court does not delay the injunction.

The non-redacted part is simply not convincing enough. First of all, Dell is doing hard disk image all the time. New computer in the line up usually means a new image. With new computer, I do not mean a new line only, but include existing lines that have things changed, e.g., a new graphic card or a new piece of hardware etc. I am sure Dell got the imaging process and the testing process down to a tee. That should not create much hardship. Moreover, if imaging and testing had been so difficult, why do Dell preinstall all those “crapware”? Dell have hard disk images without MS Office. How do I know it? I usually buy computers at work without MS Office and Dell is happy to sell me one.

The second part is about testing. Why? The big part in imaging is actually the testing. Testing is only difficult if you change hardware. For software, it is easier, particularly if it is about removing non-critical software. Microsoft would had been responsible to make sure that removal of MS Word will not affect other part of the Office family. Therefore Dell’s testing requirement is about informing Microsoft of any testing problems. That is easy. All it has to do is to run the program through its existing test suite and report this results to Microsoft. The other part is about collecting customers comments and send a report to Microsoft. Whatever Dell cares to say about how rigorious its testing process and how it will affects its customers is rubbish. If it finds any problem, it has to wait for Microsoft to supply a fix. The last time I check, Microsoft do not issue Dell-specific fixes outside its normaly update process. The update I got directly from Microsoft is no different from the one Dell’s get. Also, the bottom line: Dell will happily sell you a malfunctioning Microsoft Office even if it finds fault. This is partly because it cannot do anything about it.

So it boils down to will removing MS Word make a dent in Dell’s profit? Yes. A big one. But that’s life. Dell will not go under if it does not sell MS Office. Particularly if its competition cannot do the same either.

Sad but true: SCO gets second bite at the cherry

Groklaw is reporting that the Appeal Court had reverse some of Judge Kimball’s judgement. I would call it a disappointing decision as Novell has to work harder to get what is rightfully theirs. While I think that Judge Kimball’s decision to recuse himself means there is less scope for appeal later, I cannot help but think the SCO is given a second bite at the cherry. This means a fresh start for the trial, and we know that SCO will be playing its usual tricks again with the new judges.

Do I think a jury trial will reverse the results? Not very likely. Still, the “certainty” that Judge Kimball gave us now evaporated.

Unfortunately for SCO, the Appeal Court also says that it has to pay Novell. That is not going to look good in its bankruptcy case. It makes it even more likely it is going for liquidation (Chapter 7).

Liquidation or not, fight IBM and Novell or not is now the decision of the yet-to-be-appointed Chapter 11 Trustee. It is possible that the trustee put the litigation up for auctions and someone (or a proxy) buys it just to continue the trouble for Novell/IBM and Linux.

The more likely scenario is the Trustee decided to drop the litigation, or the successor-in-interest to the Unix business, assuming SCO is broken up or went into liquidation, will drop the litigation. While common sense will say SCO has no case, critics will continue to say that there is a “dark shadow” over Linux.

I think I can sense that this whole thing is heading towards an unsatisfactory conclusion.

Microsoft proposal to resolve IE issue in Europe gets the necessary attention

Remember a while back Microsoft blogged about its proposal to settle the IE antitrust issue with EC? Back then, Andy Updegrove has an excellent analysis of what it means? We had Opera’s immediate responses back then. Most of the points that Opera raises about the settlement proposal are not discussed in any detail and I am a bit lose on why they are not good enough. Moreover, at times it looks like Opera are complaining for the sake of complaining or  trying to take advantage of the situation by demanding more than I would like to give them as remedy to their injury.

Now, Mozilla Foundation had completed its analysis of the proposed settlement and decided to launch a public discussion, or as Johnston aptly call it, blogging campaign. By putting more meat on the arguments and spending more time polishing them for public consumption, I find that where Mozilla shares the same concerns as Opera, I now understand the reasoning clearer.

These are healthy development in the antitrust issue. A regulator, no matter however competent, like the EC, cannot foresee a lot of potential problems with the proposed settlement. The purpose of putting the proposed settlement out is precisely to allow other players in the industry, including the competition, to go through the remedies and suggests clarifications and changes. This is an important part of the remedy process. To tell the truth, if you compare US courtroom-based approach where a judge OK a broad range of measure then take periodic review and make minor adjustment, to EC’s approach of outlining the broad principles, then fine-tune it later as in the case of protocol access by SamBA, I think the EC’s approach is better. It is less confrontational and allows decisions to be made more timely and in small increment. It has more potential to achieve its aim then the periodic review process which has the problem of potentially abrupt change of course and lengthy consideration time before the judge decides on what to do.

The damages i4i asked for are, compared to other lawsuit, mild

The Industry Standard carries a post about what the judge thinks about Microsoft behaviour.

As for Microsoft, I can see it is using its good old  trick that “You asked me to fix something that is not easily fixed”. This time, the timing is bad, really bad. They touted a feature in the next office release that will allow customers to disable the custom XML capability, which the judge seized on and say the fix is not that hard. I am sure merely disabling custom XML will fall short of the complete removal required by the courts, but it does means Microsoft can disable it quickly and that disabling it will not affect Word significantly. And of course it helps to get the judge to agree that it is an good enough interim fix.

Except it appears that the judge runs out of patient with Microsoft. Finding it “wilfully infringing” i4i patents [Note: read “wilful” as the less strong word “knowingly” in everyday speak], he do not want to wait five months for a fix. Rather, he wants to light a fire underneath and did that by giving Microsoft a short time period to correctly. Perhaps he view it as part of the punishment since you ask a techie like me, five months is not an  unreasonable time to ask for.

This detour brings me back to the subject. i4i’s requested remedy is mild. They did not ask for retrospective disablement of custom XML, which they are FULLY entitled to. If they had done a lot of Microsoft customers would had been inconvenienced. It is surprising to see that a prevailing parties thinks that a monetary damage from the other party is good enough to cure past infringement when it comes to computing technology.

Why didn’t Microsoft tell us about i4i’s patent litigation during the OOXML standards push?

PJ asked “Why didn’t Microsoft tell us about i4i’s patent litigation during the OOXML standards push?” Fair question. Let me try to answer the questions. As you will see, most reasons are not cynical at all.

First, the left hand does not know what the right hand is doing. Microsoft legal and office team does not know that the standard team is doing. It is possible. This happens more frequently than people cares to admit. Don’t believe me, look at the two biggest contributors from Microsoft on the OOXML vs ODF debate, Jason Matusow and Brian Jones, do they struck you as someone in the trench or someone in high management way away in some stately home directing the battle?

Second, Microsoft did not think much about the lawsuit. After all, lets not forget Microsoft gets lawsuit all the time. Patent lawsuits are notoriously difficult to predict. If they expect this is a nuisance suit, why disclose? [I had seen this troubling post about the what the judge said, but it is another blog post]

Third, they did not want to add flame into the already heated OOXML vs ODF debate. If they reported this, rest assure a lot of anti-OOXML people will list it as another reason not to adopt OOXML, regardless of whether the lawsuit has merits or not.

Fourth, they did not want to disclose it to ensure the OOXML path is not so bumpy. This is what PJ is hinting. There are, of course, some truth to it. Microsoft is treading a very thin line here. It of course thinks it is always on the legal side. However, when it comes to standardization, we have ECMA and ISO view to think about as well. Do they think what Microsoft did is legal? I don’t know. But is there going to be any sanctions? Unlikely.

Fifth, this patent litigation might be totally unrelated to the OOXML push. Cyber Cynic is not a MS fan. His brief review of what the patent is suggests that it  “covers a fast way of saving XML (eXtended Markup Language) documents”. If this is true, it has nothing to do with OOXML standardization. Therefore, no need for disclosure. Given the heat of the debate at the time, it is definitely unwise to disclose it.

Sixth, they actually think they will win the lawsuit. Anyone, except SCO perhaps, always thinks that they will win the lawsuit until the result says otherwise.

Finally, the most sinister one, regardless of whether they win or lose, they think they can buy their way out.

How to write a useful (and useable) tutorial

By way of Rob Weir blog post, I found an excellent beginner’s text for ODF by J. David Eisenberg.

Why do I call it an excellent tutorial? First of all, it start from the very beginning: The first tutorial show you how to create a really simple and basic ODF document with one paragraph (and one image). That is a extremely important starting point for beginners. Let’s face it, we all have to start there. One advantage of starting from there is, if you manage to compile and create the document, you know you can always start from here for your document (or go back to here and start again). Moreover, even season programmers do not necessary will be able to compile the program correctly on the first run. By presenting a bare-bone program, all you and I has to do is to concentrate on getting the program compile. We are not distracted by syntax or other issue.

Second, the tutorial does not advocate any technology.  With this, I mean one do not have to learn a lot of surrounding technology to get oneself started. The first few tutorial only requires you to know DOM. To me that is something most people with XML experience will know how to use. Without it, you cannot manipulate any complex XML beyond the simple one you created yourself.  Even if you do not know how to use DOM, the DOM syntax is written in the way developers are trained in, so it will still be easy to follow.

What are the alternative approaches? One way is to lean very heavily on what the toolkit offer, i.e., go away from the bare metal feel and start relying on toolkit. This is the way the OOXML camp choose to introducte OOXML. Compare tutorial one, two, three and four here with Rajabi’s OOXML tutorial number one, two, four and three (I swapped four and three to keep the comparison valid) respectively on Brian Jone’s blog, remember to strip away all programming language specific fluff and concentrate on the content only, i.e., how to create documents.  Now tell me, which one tell you more about the document technology itself, and which one tell you more about the toolkit and has the effect of hiding the technology from you? Which give you more confidence to proceed to experiment about the toolkit by messing around with  the tutorial materials. Also tell me which one is easier to follow and understand.

That is not to  say Rajabi took the wrong approach. The approach he took requires you to put all your faith in the toolkit but allow you to scale up really quickly to do something interesting. Nobody in their right mind will write ODF as raw XML, and sooner or later Eisenberg will have to move away from bare metal XML and start showing us how the tookit simplify creating and manipulating XML the way Rajabi introduces OOXML. The way one eventually use ODF will be through the equivalent fluffy functions in ODF toolkit.

I prefer building a strong foundation on the document technology and Eisenberg does it. I believe that Rajabi approach has a fatal flaw: The speed of scaling up comes at a very high price: You have not learned about the document technology so  if something goes wrong, you cannot tell what is wrong. It’s like building a house without a solid foundation. Without the foundation, you are going to spend more time figuring out what went wrong later, and will find it difficult to switch toolkit because you are afraid that you have to relearn everything from scratch again. For these reason, I believe this approach is not for beginners.

In defense of Rajabi, if you look at the books published by Microsoft Press, you will find that Rajabi’s approach is the standard Microsoft approach to teaching technology.

TheRegister got it right: Reality Check for Linux vs Windows on Netbook

TheRegister’s first line on Linux vs Windows return rate on Netbook is a spot on:

“Dell has delivered a dose of reality for both Microsoft and the Linux community on the subject of netbooks.”

According to the report, it is true that there are Linux-based netbooks that are being return because the purchaser was attracted by the lower pricing but has expected Windows. As an Window environment is where the user had came from one do expect this type of problems to occur.

The most important take home message is there are enough satisfied customers on Linux as far as Dell is concerned to continue the Linux netbook product line. What I will be interested to know is whether those were surprised that they did not get Windows but did not return it do so because they accept that lower price means no Windows, meaning price is their main consideration, or find that the Linux-based netbook is good enough for the job.

While people jump on the fact that Dell refuted Microsoft by saying the return rate for both OS is the same, this is less important to me. I always see any vendor’s effort to bad mouth the competition with suspicion. I cannot see Dell, HP or other big sellers of Netbook disclosing to Microsoft about the return rate for Linux-based netbook as I am sure they regard that as important company competitive secret. Microsoft, just like any vendors, is prone to taking artistics license when it comes to blowing the trumpet for their product.

Finally, to tell the truth, I do not think Microsoft’s claim that Linux return rate is higher is actually aim at the retailer, may it WalMart, Dell, HP or someone else. Those people know their market well and is unlikely to be sway by Microsoft’s claim. They have the actual market data (or rather, data about what their customers want and do not want) and as long as the Linux netbook line is worth keeping, they will sell Linux-based netbook, whatever Microsoft may choose to say.

I think that Microsoft marketing claim is aim at you and me, i.e., joe consumers. We can be sway by these type of claims. The purpose is to plant the seed of doubt into our mind. The next time you see a non-Windows netbook and cannot decide whether the price difference is worth trying non-Windows netbook, this piece of information planted into you head by Microsoft, whether it is true or pure lie, will bias you if not sway you away from them. That, as far as Microsoft (or indeed any other vendors) is concern, is a win.

Microsoft is barred by the courts to sell Word in United State

News on the business front is Microsoft is stopped from selling Word in the United State, after a successful patent lawsuit brought by i4i.

Frankly, it Microsoft was just awarded a fine, however big that is, or announce before hand how the company has plan to disable the feature if the lawsuit was successful, the impact would not had been that large.

Hopefully this is not lost in the avalanche of news surrounding this issue: Ina Fried of ZDNet has to report the obvious:  i4i says it has no plans to destroy Word. First and most important point, even if i4i wants to destroy Word, this lawsuit cannot have this effect from day one. What i4i is entitled to is a cure for the patent violation, a cure in this case is never  the wholesale destruction of Microsoft Word, since the patent infringement is only a small part of the full package. Anyone thinking the contrary is living in la la land. Second, if you check out i4i’s website, they appear to have a feature product working with Word. To destroy Word empire means they will suffer collateral damage. Third, what they really want is Microsoft licensing their patent, because this will bring big revenue stream for time to come. Failing  that, which looks more and more likely each day as the lawsuit drags out, a huge damage award from the court is acceptable. Fourth, while it might take time for Microsoft to disable the feature, it can be done. Finally, I can see the Appeal Court strike down this injunction as disproportionately harsh on Microsoft. The correct outcome, as far as I am concern and I think the Appeal Court will agree, is to give Microsoft reasonable time to cure this infringement, i.e., to work the infringement out of the program should Microsoft choose this route, but will require Microsoft to pay compensation and some monetary penalty to i4i until the cure is complete.

Want the whole background to the lawsuit? Go to Groklaw as usual.

As for the patent infringement in question? It stumped me. Putting a not-predefined set of XML inside a defined XML set is so obvious from day 3 of XML. I haven’t read the patent involved, and cannot do so without it significantly impact my professional life. Following PJ’s article, the Doug Mahugh’s post on the issue, known as Custom XML,  offer some clues, however. Stripping Mahugh’s post down to the technical parts shows there are two parts: XML and how Word handle it. While I cannot see how embedding another XML inside XML, even if you get your XML parser to read/parse a different schema for the embedded XML, but I can see parts like the actual mechanism of recognizing the Custom XML part, get Words to interoporate with it then faithfully write  it out being patentable however much I don’t think that the idea should be patentable.

As anyone in the OOXML vs ODF debate will remember, Custom XML was a big fighting point. I would like to see Custom XML goes down, but definitely not this way.

Is there value in Custom XML? Mahugh was right on the money when he says that it means developers can embedded anything XML they like into Word’s XML, without having to write any conversion code. I actually think anyone using this to avoid writing conversion code is a lazy developer and in practice, the conversion routine need only to be written once only, but I can see the appeal. In fact, this support of laziness is an definitely an competitive advantage of OOXML vs ODF. Of course, in my opinion,  another reason it should not be allowed is, this approach has huge potential to turn a document written in OOXML into a proprietary format. The RDF framework, as supported by ODF as described by Schubert is not really a cure for this problem, because the RDF descriptors can be represented in a variety of ways and it is difficult to write a schema that can support all different ways of representing your data inside the RDF descriptors. I know, I used the Jena Framework to parse my RDF descriptors. It works great, and definitely better than anything I can come out with. I, however, findit difficult to write out a XML Schema that satisfy all possible permutation of RDF descriptors format for my XML.

People moving away from GPL??

Yahoo carries a news item about a company moving away from GPL to Apache License. It is full of quotes about how GPL is not business-friendly, including a quote from Mike Milinkovich, executive director of the Eclipse Foundation. That quote surprise me because I did not think Milinkovich will speak in such harsh term, even after considering he has an axe, in the form of Eclipse Common License, to grind.

Initially, my reaction was “yet another open source bashing story in general, GPL-bashing story in particular”. After reading the first paragraph, in particular, the words “…move the license to Apache”. I do not think it is a anti-open source story. However, after reading the full story, the anti-GPL undertone is unmistakable.

It is not an anti-open source story because the company in question moved from what I see is a more restrictive GPL license to a more relax Apache License. The CEO stress that this is purely a “business decision” and I can see why. However, I think the decision is even more difficult to make than the decision to go GPL.

SCO in Chapter 11 Trustee hand. No surprise here

Groklaw carries the news that SCO is to be placed in a Chapter 11 Trustee hand. I do not think anyone, PJ _and_ SCO included,  is surprised by the decision.

We know SCO is hemorrhaging money and no one is coming forward to rescue it. The best deals it can gets appear to be from vultures aiming at tearing the company apart. One cannot say it did not try. One cannot say Darl McBride did not try his best as he put forward some of his own money. The effort is quite simply not good enough unfortunately.

I think IBM and Novell will be pleased. Yes, they asked for Chapter 7 Trustee. Realistically, I think they know the best they can get is Chapter 11. SCO has been spinning the “possible” beneficial effect if it wins the lawsuit. While you and I know it is not possible, the judge probably don’t and don’t care. I mean don’t care in the kindest possible way to the judge. His has the duty to try to keep the company running if possible. He has to be fair to everyone, SCO included. Chapter 11 bankruptcy as I see it is about trying to preserve the business, so the judge has to err on the side of caution. He has to evaluate SCO’s claim of the non-existing, impossible and improbable win. He did the right thing by asking a trustee to do it on his behalf. The trustee has more time than him to evaluate ths lawsuit.

It appears that the trustee has the power to terminate the lawsuit, should he choose to. My bet is he will. That is good for Linux. Although I would had preferred the court to rule on it. For the sake of effort and money, its better for IBM and Novell if the trustee pull the plug. They already fought a brilliant war and it is already shown that SCO’s claim is groundless. That is good enough to call it a day.