|
Welcome, XXX.com readers, but I bet you've surfed to the wrong place: Despite the racy headline, we're talking Monad (with an "m," not a "g").
So what's with the spanking reference? Some Microsoft advocates over on the Channel 9 Web site seem to think that's what I deserve for telling the truth about what's happening with Longhorn. (Note to Mr. ManicKernel on Channel 9: You are the one who needs a spanking. And then to be put to bed with some warm milk and a pacifier. All IMHO, of course.)
I won't even broach the issue of whether anyone would suggest a male columnist be "spanked" for daring to rage against the Microsoft marketing machine. That's the subject for another column. Or 20. Or 50.
I am talking about much more mundane and family-friendly topics in this column. Specifically, the .Net Framework 2.0 and "Monad," the scripting shell that was supposed to debut with Longhorn, the next version of Windows.
I recently had the audacity to point out that Microsoft is talking about Longhorn very differently from how it tackled the subject at the Professional Developers Conference (PDC) in 2003.
Read: The Dirty Little Secret About Longhorn
And: 'Monad' Scripting Shell Unlikely to Debut in Longhorn
A number of folks — including (but not exclusively) yours truly — believed until recently that Microsoft was proposing to make WinFX the programming model in Longhorn. WinFX ultimately will be the successor to Win32. It turns out that at the PDC 2003, Microsoft omitted a key word ("ultimately") in describing WinFX.
In a similar vein, Microsoft also led a number of PDC 2003 attendees to believe that the .Net Framework 2.0 would be the plumbing upon which all of Longhorn would reside.
Now, some Softies are admitting that they might have not explained things all that well. Sure, the .Net Framework 2.0 will be part of Longhorn. But that doesn't mean by any stretch of the imagination that Longhorn will be based on it or that Longhorn will consist completely of managed code.
A number of Microsoft customers, developers and pundits are just figuring this out now. Microsoft says it tried to do damage control in the post-PDC timeframe on this misunderstanding. But it seems its education campaign didn't trickle down to the masses.
"Longhorn is largely written in non-managed code," John Montgomery, directory of product management with Microsoft's developer division, told us in no uncertain terms last week at the Tech Ed conference in Orlando.
That doesn't mean Microsoft is abandoning managed code. Nor is the company spending huge cycles rewriting Windows components that were written in managed code so that they will debut as unmanaged code, he said.
"There are 2,000 new features in Longhorn," Montgomery explained. "Our goal is to allow developers to access them through Win32 or managed code," Montgomery continued. "Today, you can access all the Win32 features through managed code, and vice versa."
In fact, there are quite a few Longhorn features that will consist of half-managed and half-managed code, Montgomery added. But he said this was Microsoft's plan from the start.
It all sounds (relatively) clear now. But this managed/unmanaged quagmire has been a murky one. And now a similar situation is evolving on the Monad front.
Until quite recently, Microsoft execs described Monad as the Longhorn shell. But within the past couple of weeks, for reasons that remain largely a mystery (we're guessing failure to pass certain quality gates), Microsoft has cut the scripting shell from the planned Longhorn feature set. While Monad is supposedly still part of WinFX, it is not part of Longhorn. (For more on that distinction, check out Montgomery's blog posting, regarding what being "in Longhorn" means.) The new plan: Monad will debut in Exchange 12 next year and likely be incorporated into Windows as of Longhorn Server R2 (2009?).
Posters around the Web are wondering aloud whether these "resets" should convince Microsoft watchers that Longhorn will be little more than Windows XP Service Pack 3 by the time it ships next year.
Microsoft maintains that users will get a lot of brand-new value-add with Longhorn. For our part, we think Microsoft could do a lot worse than simply make sure that it simply gets the base-level Windows fundamentals right with Longhorn. But that's not what the Redmondians want their Longhorn marketing message to be. They want to be able to proclaim that Longhorn is a major Windows release. They want to be able to demonstrate that there are enough new features to merit a Windows upgrade, even for those recalcitrant Windows users who've clung to Windows 2000 and other older releases.
Our takeaway from all of this? Microsoft and its fan boys/girls should stop confusing the message with the messenger. Instead of spanking the folks who are setting the record straight about Longhorn, why not cast blame on those who confused the messaging in the first place?
What are your two cents? (Disclaimer: Non-PG13-rated spanking comments are highly unlikely to make it past our e-mail filters and forum moderators.) Is Longhorn falling short of what you expected? What else does Microsoft need to come clean about, regarding Longhorn, so that there are no negative surprises by the time PDC 2005 rolls around?
Talk back below, or write me at mswatch@ziffdavis.com, and
let me know what you think.
|
Comments (5)
no one cares that much.
Posted by rad19 | June 14, 2005 12:29 PM
Mary Jo, I read a few of your columns, and I must say, the topics that you seem to focus on regarding Longhorn, and other Microsoft related issues, don't bother the people who care about them the most..namely software architects and developers. Believe me, most IT folks, and defintely business people could care less about Longhorn right now. Microsoft has, and continues to cater to the developer ... and for good reason.
I went to the PDC 2003, and the one thing I AM disappointed about is the fact they are pulling WinFS from longhorn. Does that really piss me off? Not really. Why? Because I'm a software architect, and I use the tools and technologies available to build solutions that solve business problems. The articles that I've read from you, and basically every other analyst's columns I've read, miss the big picture of what's going on right now with Microsoft. Longhorn is simply a small piece of the puzzle..allbeit an important piece.
You all analzye a given application in isolation. You ask how does Longhorn compare to Tiger or a Linux Distro? You ask about feature XYZ or write a column on how Longhorn isn't including some new scripting shell. And then the comparison of Sql 2005 versus Oracle or DB2. Or Exchange versus Notes. Or .NET framework verus JAVA. Or Sharepoint versus other collaboration software, or Biztalk versus WebMethods...and on and on.
Meanwhile, Microsoft, ( with .NET 2.0 getting enabled everywhere, web-services technology getting deployed everywhere, Indigo, Avalon, XAML, SQL 2005, Biztalk and new versions of ALL the Office software ) is providing a foundation that cannot be paralled in terms of ease of use, integration and sheer ability to DESIGN and BUILD software applications that solve business solutions.
Does Microsoft mess up? Sure! Could they improve things? Absolutely! Could Longhorn have a slow adoption rate initially? Very possibly.
But, Do most developers care or even know that MONAD was gonna be dropped from LongHorn or that LongHorn isn't written on the .NET 2.0 Framework? No. We don't care.
What I DO care about as a software architect is what is available and what is happening up and down and across the technology stack. And what's happening there is changing the technology landscape and how software gets designed and developed.
Software Architects look at ALL the pieces and decide which set of technologies will best solve the business problem. In terms of where Microsoft is taking .NET, Smart Client technology, and their entire technology stack, LongHorn looks like yet another great piece of the entire puzzle. And there's plenty to do between now and 2006/2007 to prepare for that change.
Posted by MikeNeutral | June 14, 2005 11:03 PM
However, I think you deserve to be criticized for making an ad hominem and demonstrating that you didn't understand what you were talking about while at the same time attempting to make yourself look like you were revealing some big scoop. You made yourself look foolish. As I pointed out in my forum post to your last article, I think you've misunderstood what's been going on. Surely you were informed, or saw when Jim Allchin did the DOS Lotus 1-2-3 demo on an alpha version of Longhorn to show that it would be backward compatible with very old legacy applications. Did you seriously believe that Lotus was running on top of .Net 2.0?? Think about it. Microsoft had just released .Net 1.1 that year! 2.0 was probably in early development at that point, surely not something to base a critical technology on at that point. In your article here I think you made another gaffe. You said that Monad would be part of WinFX, though it would not come with Longhorn. This development with Monad is probably the last of the previously announced features of Longhorn (though little known--I'm surprised you picked up on it) that's going to be stripped out. I doubt there are any more "secrets" to be revealed. I agree with you: what's special about Longhorn now?? But to my point. WinFX is a forthcoming new .Net API. Monad is a command-line interface, based on object-oriented principles (I think written in .Net) currently in development. It's a tool, like the DOS Window, not a part of the WinFX API, which is used by developers to support their applications. A piece of advice: Stick to what you know. This piece was a bit better, but reading your last one was an embarrassement. Up to "A Dirty Little Secret", your work has been articulate and interesting, and has helped keep me informed on what's going on at Microsoft. Please continue in that tradition. Maybe just forget about writing articles that involve the .Net Framework/WinFX. I think you misunderstand what it is.
Posted by Makeshift | June 16, 2005 3:57 AM
After following MJF's links to Channel 9 discussions, I can see where she's gotten some of her impressions. Robert Scoble participated in some discussion threads about what would and would not be in Longhorn, and why the "reset" occurred. He said that those managing Longhorn development realized that all the stuff that had been promised for it, being developed in managed code, had such a long time horizon that they wouldn't make it into Longhorn as they were. So focus was shifted to taking the same functionality and implementing it in unmanaged C++. He explained that most of the developers at Microsoft had been developing in C/C++ for about 20 years and so it's logical to conclude that they would be the most productive in that. I believe he also explained that even if something was implemented in unmanaged C++ that it would be completely accessible through managed wrappers, so .Net coders would not be left at a disadvantage. He was unclear though about how much of the new stuff--Avalon, and Aero in particular--was implemented in managed or native code. He said it was impossible for him to know because there are literally thousands of developers working on this stuff. The kernel was most definitely 100% native code, and I think he indicated it couldn't have been any other way, anyhow. He said that some technologies were being implemented in part managed/part unmanaged code. In the end I was feeling like, "Okay. So it doesn't matter." No matter if something is implemented in managed code or not, to a .Net developer it's going to all look managed. And likewise, the managed pieces would have native interfaces, so that C++ (or even VB6) developers could still access them. Probably the only area where people would feel it matters is in security. Native code has a reputation now of feeling insecure, since past worm attacks have utilized weaknesses in native code in the past (whereas .Net has buffer overflow checking at runtime). Just to show that a fully managed-code OS is a far off dream, Channel 9 does have some video of a research team working on an operating system "based on managed code", called "Singularity". It's in an extremely primitive stage, and when you drill down deeper, looking at the discussion, you realize that while they've made as much of their prototype OS managed code as they could, they still have to depend on some assembly code to make it all work. Makes sense to me. Also, there was some mention on one of the Channel 9 discussions of MJF's assertion that Monad would not be in Longhorn but would be part of WinFX. People speculated that she saw this stated this way in the transcript of an interview with a chief Microsoft developer and misunderstood what he meant. The sense I got from the discussion was that Monad will *be distributed* with the WinFX *distribution*. In other words, when you install WinFX, Monad will be in the install package as well. It seems logical then that Monad will be in Longhorn, since WinFX will come with it, though it's hard to say. I didn't see anyone say this definitively. In any case, it will surely be available for download. There were several people who asked, "So what's new that's going to be in Longhorn?" Scoble indicated there are lots of new things, but he could not reveal them now. Some or all of the as-yet unrevealed features would be shown/discussed at this year's PDC.
Posted by Makeshift | June 16, 2005 11:35 PM
While Microsoft is playing "Which feature do we cut today" with Longhorn we find out as some of us knew that not only has Apple be releasing new OS's on schedule but for both Macs and PC as well. Now with Apple moving to intel chips next year one has to wonder if Microsoft will someday have another competitor.
Apple says the new Macs will only run on their built PC's and not one for Dell but one has to wonder what if they did open it up to any PC? Apple has been doing very well in cross-platform ease with included programs to transfer the contents of a PC to the Mac, so your documents come along with you if you switch. Tiger now uses ACL's and its ability to bind and be managed under active directory just keeps getting better.
If Longhorn is nothing more than XP SP3 and does nothing to address the security issues plaguing the current XP and further imposes its restrictive DRM. One has to wonder if Apple will give an option to all those Windows iPod users who have had a taste and give them the choice of using Mac OS as simply as updating their old Windows PC with MacOS?
Posted by Randy Smith | June 18, 2005 12:12 PM