Whats going on with Longhorn anyway?
First, WinFS was dropped. Then, it came out that Longhorn wasnt going to be based on .Net Framework after all.
And now, Microsoft tells us that Monad, Microsofts super-duper combination shell and script language, isnt going to make it into Longhorn either.
Now, while some people were excited about WinFS, I really didnt care for it. Does the world really need a processor-intensive file system? I dont think so.
I agree with Hans Reiser, creator of the ReiserFS (Reiser File System), that if you have to build a layer on top of your file system, it means you need to fix the file system, not add another layer of complexity and abstraction on top of it.
I was also unmoved by.Net Framework removal from Longhorns heart. To me, it just showed, as I had thought all along, that .Net Framework wasnt suitable for building something as complex as an operating system.
Lest you accuse me of anti-Microsoft bigotry, thats not the case here.
I dont think any high-level language, such as Java, is suitable for building a production operating system.
Even as our processors zoom past 3GHz, operating systems need speed, speed and more speed.
For that, you need languages that produce fast machine code easily, and that still pretty much means you need to use the C language family.
But Monad, I thought, had promise.
Monad was going to be far more than just, as some have called it, the next generation CLI (command line interface) for Windows.
It was going to be a scripting language that tried to one-up Python, Ruby, PHP, Perl and all the other advanced script languages that make quick and dirty programming so darn useful for system administrators and power-users.
It was going to put the long-in-the-tooth VBS (Visual Basic Scripting) in the trashcan where it belonged.
It was, for as far as Im concerned, the most potentially useful part of Longhorn.
Advanced Script Languages Benefits
Advanced script languages are good for far more than just getting your laundry lists of automatic network jobs working, though.
You can use them to build complex applications. For example, the best anti-spam, add-on program around, POPFile, is built on Perl.
What Monad brought to the table, which I thought gave it a real shot at being an outstanding shell programming language, was that instead of simply passing structured text from one pipe to another to a result, you could also pass .Net objects.
This meant you could work directly with ADO (ActiveX Data Objects) from the command line or, far more likely, a script program.
This in turn meant that you could easily access, and act on, a really wide variety of data stores.
For example, with the right connectors, you could quickly build programs that could easily update a MySQL staff and equipment database with information hidden with AD (Active Directory). Neat!
Now, power alone is darn dangerous. As the saying goes, anyone can screw up, but to really screw up you need a computer.
Microsoft has long been guilty of creating program and operating system IPCs (interprocess communications) mechanisms, like DDE (Dynamic Data Exchange) and OLE (Object Linking and Embedding), that are inherently insecure in a network environment.
Indeed, as far as Im concerned, its these fundamental security mistakes that have made Windows security a bad joke.
Monad, which would have enabled users to remotely execute commands, certainly had major potential for abuse.
Eric Chien, a Symantec researcher, has said that he feared Monad might lead to a new wave of “script viruses,” like 1999s Melissa virus.
Chiens right, of course. Given Microsofts dismal security track record, youd be a fool to trust Monad security.
Still, perhaps Im being irrationally optimistic, but I really thought that Microsoft was finally learning some security lessons and that Monad might, just might, be able to balance security and usefulness.
Well, its all moot now. Whether its because Microsoft couldnt solve Monads security concerns or because handling objects in a scripting language as easily as PHP handles strings proved beyond Microsofts developers, were not going to see Monad anytime soon.
This leads to one final question: What is going to be left in Longhorn that will make anyone want to “upgrade” to it, anyway?
eWEEK.com Senior Editor Steven J. Vaughan-Nichols has been using and writing about operating systems since the late 80s and thinks he may just have learned something about them along the way. He can be reached at firstname.lastname@example.org.
Check out eWEEK.coms for Microsoft and Windows news, views and analysis.