At least one good article against Longhorn

November 20, 2003 Edit longhorn

As some of my beloved readers thinks I can’t give credit to someone slapping Longhorn, here is an article from eweek:

http://www.eweek.com/article2/0,4149,1388616,00.asp

This guy at least knows what Longhorn is and intend on being. A few comments though: (that’s why I have a blog you know, giving my opinion without anyone asking for it :-) )

Developers don't hunt for brilliant technology -- their instincts tell them to go to where the volume is, because that's where they can monetize part of that $80 billion.

I can’t disagree more strongly to that. Developers are just like flies. They go where the nice colored light bulb is, because they like it k00l, because they are g33ks. (I think Slashdot writes it like that).

Longhorn provides more than just APIs. It’s more than a platform. It’s an entry in the world where being cool is as important as being technically challenging. Why? Because even people typing all day on their computer like nifty funny user interfaces. It’s not only about technical excellence, it is about bringing back some cool experience on the windows platform.

If you think about it, Microsoft is pretty clever. They know they would have a hard time convincing terminal mode guys and linux aficionados from going full way into an OS. They know Java and Linux can provide data entry user interfaces. To keep their advantage on the desktop, they have to go somewhere else, somewhere where people enjoy the cool factor. This is your mother, you, your children, maybe even your dog. Anyone who just likes it cool. Apple survived because of that coolness factor and are building their whole product line on it. Microsoft started doing it with windows xp, as a form of “will it work?”. Now that they know it does, they are going there full time.

Another quote:

Curtis Sasaki [vice president of engineering, desktop solutions] is going to drive those priorities based on what customers tell us, not because I come in and say I have a vision of what is interesting that could be delivered. We don't run the company that way. I run the company based on what it's going to take to close [the sale with] the Indian Ministry of IT.

Now this is also where Microsoft and Sun differ a lot. Schzartz listen to consumers, and try to bring them exactly what they asked for, not more. Microsoft listen to their customer, but rather than trying to deliver features A and B, big teams are organized, design is done for the next 2 or 3 years, and A, B, C and D become part of a big plan, a big building block. This is what Indigo is, this is what Avalon is. This is looking into the future, not into what people want today.

Schwartz is absolutely right by saying that being an end node on the network is not enough. But now again, peer 2 peer technologies are there exactly to transform end nodes into central communicating nodes. And this is going to be integrated in windows. Just like JXTA is integrated in Java. (Stay tuned for my rants on the JXTA technology).

Yes, the PC is only one of the devices we’re using. But it’s the most powerful, the most connected, the one we’re the most used to. If you get it even more compelling by providing rich interactive fun clients to deal with data, and let it integrate with any device anywhere at any time (WinFS Sync), then you’ve got the killer app. People won’t abandon their PC to switch to Java phones. Ever.

More on Longhorn FUD

November 19, 2003 Edit longhorn

It seems my little entry generated quite a lot of feedback. Mostly I've been accused of just trusting Bill Gates word on groundless discussion. So here is a second post, facts, just facts mam.

Myth #1:

Fact: People in an office use more than word, whoever they are. Where can you see it? In any office, including mine. You want some pictures?

Fact: Microsoft is remodeling their licenses to try to give more choice. But they are not always successful, can do some mistakes. Where:

http://www.computerworld.com/managementtopics/management/itspending/story/0,10801,72214,00.html

Myth #2:

Fact: Longhorn is a client side platform, not an API. WinFX is an API for the client side. Indigo is for both the client and the server side. You can use Longhorn to access an LDAP directory and a Web Service GXA stack sitting on a server running another platform.

Fact: Longhorn is windows, so you develop software for windows. And people like it. No one really cares about having the 3d accelerated Desktop Composition Engine running on Linux.

Fact: Java can't do that.

Fact: Middleware doesn't have anything to do with Longhorn.

Fact: Middleware can be inter-operable through GXA and Web Services, in an inter-operable way.

Myth #3:

Fact: The Java Community Process is plainly slow at defining new standards to improve the language.

http://www.eweek.com/print_article/0,3048,a=108119,00.asp

Fact: Microsoft and Java are both innovating, but they are on different release cycles. Oh, that, and Longhorn is not .net and Longhorn can still run Java. Doesn't have anything to do with the underlying Longhorn API (WinFX).

Myth #4:

Fact: Visual studio versions road map.

http://msdn.microsoft.com/vstudio/productinfo/roadmap.aspx

Fact: Side by side execution works.

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/dnnetdep/html/sidexsidenet.asp

Fact: the framework provide compatibility.

http://msdn.microsoft.com/library/default.asp?url=/library/en-us/cpguide/html/cpovrversioncompatibility.asp
"The .NET Framework supports both backward and forward compatibility"

Fact: Developers must learn new things all the time, this is only a question of staying competitive. This has nothing to do with how much you charge your clients. And it's true for windows, for java, for Linux, for any single piece of software on this planet.

Myth #5:

Fact: Linux cannot run out of the box on any platform. The kernel can be modified to support it.

http://www.handhelds.org/geeklog/index.php

Fact: Windows doesn't run everywhere. The kernel can be modified to be re-targeted. XBox, Windows on IA-32, IA-64, AMD64, and in the past PowerPC and MIPS. Looking forward, XBox2 on the G5 processor.

Fact: The WS-* specifications on top of which Indigo is built are done by Microsoft, IBM, BEA, TIBCO, etc... For example, WS-ReliableMessaging states the following:

WS-ReliableMessaging was published as a public specification on 13 March
2003. This is the first joint BEA/IBM/Microsoft/TIBCO Software publication
of the specification.'

Fact: All of these specifications are provided to the community royalty free.

Fact: Some are standardized by other groups. WS-Security by Oasis, for example.

I think I brought enough facts on the table. Now if you really want to debunk the facts, please do it. (I'm starting to getting impatient at finally seeing a correct argument).

Longhorn FUD: Tony McCune Is Clueless

November 17, 2003 Edit longhorn

This article is one of the THE most stupid article I’ve ever read.

Myth #1: Never heard of Open licenses? Microsoft is doing a lot of work to provide for suitable packages. Oh, and saying people only use Word… Maybe you’ve been independent for too long and don’t remember how real people work in an office.

Myth #2: This guy is absolutely insane. He goes on and on saying how Microsoft is locking you on the server side with Longhorn. Hello Mr McCune, if you attended the PDC, or even took some information on it before writing this, you would know Longhorn (and you ARE talking about Longhorn there aren’t you) is a client platform. Talking about “You’ll have the freedom to use your choice of middleware and hardware, including Windows” to my mother might be interesting and a compelling reason to choose Linux. If he was talking about writing software for real people out there, well, I’m afraid you’ll have to do a reality check Mr. McCune, doing client side software in Java is not compelling for anyone at the moment, unless you want to sell it to people who share an ideology with you. Yeah, try to write 3d ui, video, real time compositing UIs using Java (or Win32 for that matter) and come back after your reality check.

Myth #3: See above. “Java has made great strides in usability”, I beg you pardon? Are you talking about the so called Community Process? Now this is interesting. As for Java not being tied to a platform… Do you have any, and I say ANY, non trivial J2EE application that can be moved between two application servers? Reality check again. I wouldn’t publish on ZDNet saying that I can swap my Microsoft .net components on Mono without carefully planning to do so. And I wouldn’t do it for an already written server side application. You know why? Cause I wouldn’t even want to see the light of day again, by fear of being looked at the stupid Java Zealot bragging about a product without even knowing what it’s all about.

Myth #4: Alright, a few things here. First of all, learn how to read. That would be handy. Not three, but two versions. First, Whidbey (and not Whidby) will be released next year with a version 2.0 of the framework, available on all supported .net platforms. Then, a Longhorn Orcas version (Not Oracas) that will feature Longhorn specific tools for the Longhorn specific API. Can you develop using an old version of the language on the new version of the framework? Yes you can run the programs, compile the programs, use the programs, and publish the programs. You can also run each of these versions at the same time. Or just use all of these applications with a single framework version. Best thing is, you say “Developers must continually update their skills which costs customers time and money”. Interesting, how you as a consultant would see Developers having to learn new things as a bad thing? Or are you out of job because you’re still thinking Java will change the world? Yeah, more likely.

Myth #5: Excuse me, IBM wrote an OS running on anything from a hand held to a mainframe? This is really interesting. Oh, you meant Linux? Like you can use GIMP out of the box from an iPaq? Oh, prove it; I’d love to see that. And if you did go to the PDC (Please show me the ticket, I’d love to know where you were hidden for your absolute lack of knowledge makes me believe you could have been locked in a toilet with this guy), you would know for sure that the whole SOA and Web Services stack (Indigo, which I’m sure you’ll love to write Indigeno) is done by IBM and Microsoft using common protocols that are published standards. But it cannot be otherwise, as 50% of it comes from IBM which “are agnostic and run on virtually every operating system and device”… Wait wait, you mean that each service will be able to talk to a Mainframe, an iPaq and windows? Welcome to the real world.

In the name of the sane Development community, I would advise anyone who survived the reading of that piece of flame to discard it, make a lot of fun about the clueless author, and blame ZDNet for letting that kind of articles from getting through. Even Slashdot wouldn’t have published something that clueless.