September 2004 - Posts
I just saw this post about the C-Omega language and the authors comment that we should favor C# over VB.NET for this reason. Inspite of being a C# developer currently I DO NOT AGREE.
So C# is ECMA standardised which is something nice for MS and I use that often to hit my Java friends over the head with especially when they complain that C# attributes are really meta-tags when viewed through Java lenses and that sort of rubbish. Hey, if its ECMA standardised then if MS calls them attributes thats exactly what they are. However I can't blame them. James Gosling, the founder of Java recently claimed that C# didnt have attributes. He's either totally blinkered or he's been living in some cocoon that has no real world view. How can anyone realistically make such a claim?
But coming back to the point. In the Busting Development Myths session at TechEd Amsterdam we discussed the VB and C# issue. One thing which came up is that since C# is largely ECMA standardised it really cannot move forward very quickly cos you have all the 'standards body' head honchos pontificating over every proposal. However VB.NET, being purely proprietary can actually race ahead in productivity and features [even more than the My namespace] since MS doesnt need to answer to anyone.
So this C Omega thing is not a reason to move to C#. It will probably be just as easily used in VB.NET. And dont forget that we can always wrap the functionality in a DLL and call it from any language.(Which is why its fine if we dont have the My namespace in C# - I'm sure we can just as easily wrap it and use it). Generics are part of the 2.0 framework so they arent exclusive to any language.
I dont see any reason to dismiss the VB.NET developers when .NET makes everyone equal. A well designed class library will last a long time no matter what language it is written in now.
This XML focussed language thing reminds me of another post (dont remember exactly where) where the MS person said that XAML was the outcome of the X# excitement that went around a while ago. Wonder if C-Omega will end up being subsumed into some namespace dealing with XML !!!
I'm very disappointed and unhappy with MS and the TechEd 2004 Amsterdam organisers. Last year the PostConference DVD had webcasts for almost all the sessions. This year there were only 21. This is horrible. I was looking forward to viewing sessions I had missed and re-viewing key sessions on architecture and development and there was nothing in the DVD. Ok, they had the power points but those are not worth anything without the voice overs.
I'm going to try and contact someone in MS to complain. My company is just not going to send anybody next year if we have to come back virtually empty handed.
What did the rest of you attendees think? Is it worthwhile to get only 21 sessions? Can we get some kind of formal complaint lodged with MS? Do post back and tell me what you think.
If you ask me, I think it sucks!! Not having WinFS in the 2006 release is a cop-out. I have one question. Why would anyone want to move to Longhorn? You might say for Avalon and Indigo, but then Avalon and Indigo are going to be available on XP and 2003 too, so there's no reason to move to Longhorn.
Also if you consider security and stability, XP and 2003 (which would have had R2 by then) would be finally stabilising, at least in the eyes of security conscious enterprises. In contrast Longhorn would be a first release and very few people buy the first release of any software. Sure, a lot of these security features would be baked into the product , but then again there are bound to be a few hundred more 'potentially dangerous" critters out in the open at that time and the fact that only the client version is out wont stop them. Clients are more likely to be open to attack.
WinFS is the only thing that would make Longhorn really attractive for me. Having the RDBMS as the de-facto format for everything including Exchange databases is very cool from a deployment point of view.
From a .NET Developers point of view, ObjectSpaces moves further away. I saw a 3 part post on one of the ObjectSpaces developers blog about why ObjectSpaces was moved from Whidbey to fit into Longhorn WinFS, but where will it go now ? And when it does finally come out will we really care? NHibernate would probably be in version 10 by then.
Other posts on other sites have discussed how MS would be losing out to Apple in the database backed file system storage and all that so I wont repeat all that here but I wasnt aware that Apple had advanced so far !!
It seems to be just another money making exercise. Force everyone to upgrade to a new OS even if there isnt anything worth upgrading for. Hmm.. now where have we heard that before?