March 2004 - Posts
Came across this post today, and couldn't stop laughing for some good minutes.
Douglas Adams would have loved it. Come back, Arthur Dent!
BTW - note that the picture of books on the upper right corner of the page contains the Hebrew version of “Mostly Harmless” - the 5th installment of this amazimg series.
Google has tweaked its UI. Nice touches.
A
report from Netcraft shows ASP.NET deployment overtakes JSP / Servlets deployment.
There was a reply in one of my recent posts, which states that, for programming languages, “if it's easy then it's flawed”, because newbies tend to do much more mistakes if only they can.
I must say the I don't agree with that. The ASP.NET page model is really easy compared to the traditional page model (ASP / JSP), especially when combined with the VS.NET IDE, and it's true that it is really tempting to push all the logic in the code behind file, and then state that we achieved a seperation between logic and presentation (they reside in different files, don't they?). However, it won't be smart to block this option. The .NET is wide enough to allow all kinds of apps, from N-tier clustered applications to stand alone 2 pages sites. If the framework will force to write according to specific standards, its scope will be shrinked to those apps which their requirements are fully compatible with those standards, which can change tommorow.
What do you think about this issue?
Well, one of the reasons lies undoubtly in this. If a company can let itself put so much effort in a mere Icon, no wonder almost everyone likes the whole bundle.
BTW, did Sun use a usability lab for their JDS?
An interesting article (blog, actually) posted by one of new Sun employees.
Althoug there are some interesting points in this piece, I think his .NET UI rants are really funny, when looking at what Java has to offer in this field (Swing).
We've got used to the Java guys telling us miracles about the Java “Multi...” features: Multi-Platform, Multi-Vendor, Multi-Contributors, and so on.
However, when looking back to the recent years, it looks like actually Microsoft is the company to support the Multi-something.
In the early years of the JavaScript, MS has come with the DHTML idea. DHTML is not a language. It's a DOM - Document Object Model (although I've heard some developers argue heartily that it is a language, and I still see resumes that under the Languages section appear, side by side, VBScript, JavaScript and DHTML). You can access this model in any way you want - JavaScript, VBScript, COM. The NS guys, from the other side, support only JavaScript.
Then MS came with the COM. This was a big hit.One could write a C component, throw it somewhere on the web, and other developer could grab it, plug it into its VB or ASP app, and voila! there it works like a charm. And where is the Java in this story? Out of scope, actually. Java has its own ways of connecting to COM, but it requires 3rd party products. It is not inherent in the platform.
And finally, then came the .NET, which made the real seperation between the Platform and the Language. Here the Java has really nothing to say.
Now, why did I came to this subject in the first place? The first reason is an Email sent to me by the Java team leader in our company, titled: “Will Java become like .NET?” Inside it I found this link.
The other reason, and maybe the more important one, is that currently we are evaluating some EAI products for one of our projects. The environment that should be integrated includes AS/400, VB / Access, and .NET / Oracle. All the vendors, including IBM, BEA and, of course, MS, told us that the integration with the VB & .NET apps will be a snap, and the main focus should be on the AS/400. Now, if IBM tells me that they have no problem at all connecting to VB and .NET, this is, in my opinion, the real proof that MS did a good job with the “Multi...” issue.
Yes, I joined the happy family.
I've installed the Whidbey today (right after I removed the VS.NET 2002, to make more HD space for it. How symbolic...). The setup was flawless, if a bit long...
I tried to create a small site right after the installation, and got confused by the lack of normal web project. How can I start a web project with the good old web.config but without all the starter kit fuss? I guess it will take some time to find it out.
I looked into one of the PDC's hands on lab, the Data Controls one. Quite impressive, although I do hope no one is going to drop a data source right on the ASP page, with the connection string exposed with all its glory.
I'm especially curious about the new server control features - UI Adapter, web resource file (my 2nd favorite), script callback (my 1st favorite. Oh, I missed you, ya old naughty remote scripting), and more.
Yep, there's a lot to learn. Good to know MS has pushed the release date to 1st half of 2005. This way we'll have a lot of time to dig into it. :)
A few months ago I delved into the new Office 2003. I've created some SmartDocuments, and had some fun with the new interfaces. What I liked the most was that all the work was done in the familiar VS.NET IDE.
Today I finished a large PlaceHolder for the MCMS 2002. I liked the placeholders framework strcuture, and the development was quite smooth, and with some help from Steffan, the undisputable MCMS king, everything now works fine.Again, what I liked the most was that the entire work is pure .NET dev, in the beloved VS.NET.
When I remember the not-so-good old days, when we had a seperate IDE for each environment, I get shivers. After going through Office Developer (for SPS!!), VBA (for Office), VB6 (for...well...VB6), Edit Plus (my favorite ASP text editor), and then some, I can't but praise MS for the great unity they put into the different environments. Undoubtly, the learning curve for the new Office system, SPS2003, MCMS 2002 and all the future products of MS will be much smoother.
We had today a user group meeting, where we had a presentation about the new features of Whidbey. Already familiar with most of the new features I saw, I was still amazed by the wealth of new features over there, including especially Master Pages, new Server Controls, Personalization etc.
However, this made me think. Looks like most of the features added to the new VS.NET are actually MS-in-house implementations of products that are available today as a 3rd party product. I can mention, as an examples, the GridView control, the TreeView control, the Master Pages and, of course, the ObjectSpaces. All these components are available today as a free components or full blown 3rd party products. Is that what we really want the .NET to become? I think there are some problems with this concept:
1. It will kill small companies with expertise in various areas of the .NET framework. No company which offers a great / flexible / beautiful DataGrid based component will survive against the not-so-great-and-quite-average MS born GridView.
2. The market will unify around average, sometimes limited, components. Instead of using advanced O/R mappers, for example, everyone will use the ObjectSpaces. Alas - you can't use it with Oracle! But will software vendors continue create O/R mapping tools just because of that? I doubt.
3. Maybe the worst of them - the .NET movement may come to an end. Yes, I know that may sound a bit radical, but think for a moment: The reason that software movements, such in the Java or .NET community, are created, is that their developers just LOVE what they do. They feel challenged by the platform. The .NET or Java geek wakes up in the morning, and the first thing he thinks about is: “How can I improve what I devloped yesterday until 3:00AM?”. He knows he can improve the platform, add new features, create new projects and workspaces, and make the platform alive. But if he will know that all what he wants to do already exists, though in a limited form, he just won't bother. Instead of creating his own DataGrid, he will compromise on the GridView. And compromising is the worst enemy of the geek. Compromised geek is not a geek - he is a plain normal 9:00-17:00 developer. And movements are not created by such a developers. One of the reasons of the huge success of Java is what it lacks - it does not have, until recently, something similar to the WebForms architecture. There is nothing like DataGrid, Repeater, DropDownList and so on. There is EJB, but there is such an argument about it that there are some really good alternatives for it. The geeks looked at Java and said - “Hey, there's a really good platform, let's improve it!” and a new movement was born. The geeks (not the same one, unfortunately...) said the same about the .NET, but the moment they will start saying something like “Hey, there's a really good platform, let's use whatever it gives us, because it already contains everything we need.”, it will mark the end of the platform.
So, please, Microsoft. Continue give us such a great core platform features such as Generics, Async ADO.NET, Design Time Improvements and so on, but let us, the geeks, create our own GridView, TreeView and all the other fancy staff that makes us swear for you forever.
More Posts
Next page »