Opinions regarding the Google Web Toolkit: "Is it time to move to JAVA?"

An interesting article. Bottom line -- if you're going to write JavaScript, write JavaScript. A limited Java to JavaScript translator is sort of a silly way to go about it. Besides, what's so bad about HTML and JavaScript?

Microsoft is definitely behind the technology wave a little, but it's far from lost. Atlas is just on the horizon and with seamless integration between C#, VB.Net, ASP.Net, and Visual Studio that's going to be a winning combination for Web 2.0 applications running on windows servers. Meanwhile, while Google's tool kit is greatly hyped and Atlas has yet to go into production, it's things like prototype, script.aculo.us, and Ruby on Rails that are going to win out in the end. They are accessible, quickly and correctly respond to the needs of the community, and are available now (and have been for sme time). So if anything, both Microsoft and Google have lost the war to open source (really, 37Signals)...well, at least the web dev battle (perhaps it is only a moral victory, not an actual crushing defeat for Microsoft or Google).

The web application war is far from over. Google vs Windows live, IE7 vs Firefox -- That's when the war gets bloody. And out of the ashes, comes raises open source and the "small" guys.
 
An important thing to understand is that Microsoft is already preparing for web development competition. They've released their Express edition software for free (as far as I know it was supposed to have become a pay-for product by now, but it's still there for all to download). They've also distributed free compilers for VB, C#, and C++, as well as an open-source .Net implementation, and there are plenty of open-source IDEs for all three languages (and more). If you can't beat 'em, join 'em, I guess. If they fight the concept of open-source, they'll lose, so they're doing the best they can with it, and they aren't doing half bad.

Microsoft's biggest advantage is their pre-existing clout. People will continue to use Microsoft because so much of the world runs on Microsoft, and Microsoft tends to make a point of maintaining incompatibility with their technology and the rest of the world's to make the switch from Microsoft to open-source or another software developer increasingly difficult. And yes, they fell behind a little bit, but they've been stepping up and delivering better and better lately and I think they will catch up.

As far as open-sourcers stepping out of the ashes, yes, they will always be there and they will always have their place, but so will commercial software. As much as I would like to see a thriving "communist" (or open-source if the term communist offends you) world of computing, the fact is that the world of computing is primarily commercial and will stay that way for as long as I can see.
 
marble_eater said:
...as well as an open-source .Net implementation...
Mono? That's sponsored by Novell.


marble_eater said:
Microsoft's biggest advantage is their pre-existing clout.
Meanwhile, Apache has what, an 80% market share for all web servers? Most of these are on Linux and as hard as Novell may try, supporting .Net on Linux is still just plain hard. So while it may be true that so much of the desktop world runs on Microsoft, the world wide web could not survive without Linux and all the open source tools and languages that make it go. The web is basically the exact opposite of the desktop world -- linux is king, text rules over graphics, and all of Microsoft’s incompatibilities are preventing new users from switching, not locking users in.
 
mskeel said:
Meanwhile, Apache has what, an 80% market share for all web servers?
I guess I should have been clearer. I was speaking of general computing, i.e. destop, not web hosting/development. (I guess I was also a bit off topic then.)

And, to clarify (and correct myself a little I suppose), I was not speaking of Mono, but rather this CLI implementation which is not really the same as a full .Net implementation, and which should be more appropriately referred to as "public source" or "shared source."
 
Last edited:
marble_eater said:
As far as open-sourcers stepping out of the ashes, yes, they will always be there and they will always have their place, but so will commercial software. As much as I would like to see a thriving "communist" (or open-source if the term communist offends you) world of computing, the fact is that the world of computing is primarily commercial and will stay that way for as long as I can see.

I tend to agree that both commercial and open source will have their places. As a for instance example, I acquired my own MSDN Professional license which gives me the VS.NET 2005 Professional edition. What is missing and available only in the upper tier packages of the MSDN subscription is a source control package. So I turn to an open source alternative, subversion for the source control server, an open source client plugin that works in Windows explorer to make the client connection graphical, and another open source package that plugs into VS.NET itself to permit all the commands to be performed within VS without having to leave it.
 
I've been reading articles recently (java slant) that say "The War" has been over for years. Basically, their argument is that the .Net framework and C# have changed too much too fast since they were born to allow companies an opportunity to adopt the technologies. I'm not sure that I agree, but I don't really have any data to either back or oppose that argument. All I know is that it seems like a lot of people are using C# and .Net (and VB.Net too which the author didn't even take into account for some reason).

I seem to of lost the article now otherwise I'd link to it...if I find it I'll post it up.
 
As much as C# and VB change over each version, for the most part they remain backwards compatible and for anyone who is that opposed to change, they can just ignore the new features (easier said than done, but if you are stubborn enough, it is do-able). The fact that Java tends to be so slow to change can be a weakness.

Java's generics, for example, were implemented in a way as to work with the pre-existing runtime, and as a result they offer no gain in performance, some unusual restrictions, and no support for generics in reflection. All because they didn't want to change the byte code standards.

JavaBeans, Java's implementation of properties, another feature designed to work with pre-existing Java runtime, is a complete mess. The property name you see in a designer is not the same as it is in code. The property "Text" in the IDE might appear as a get/set combination in code, or the functions could be (and sometimes are) named completely differently. Again to avoid changing the standards.

When the world of computing is always moving ahead I would say it's suicide to avoid change.
 
The fact that Java is actually muli-platform (I wouldn't say platform indapendant) is about a million times more important/useful than a Java-> HTML/JS converter.

This is like a little gimmick. Seeing as JS is the biggist pain for compatability, I'd just love to see what autogenerated code does.

I'm Googles *****, and I can't wait for those AJAX Word Processessors and other tools. Just so much easier than having all this crap installed on your PC and dealing with compatability. Well if it's done well it would be.
 
Back
Top