The preview day is always an expected event, especially when it's a first one I participated, because I don't remember Microsoft having other preview before.
First detail for the Microsoft team, we are talking about Dublin, so understand mental traffic, and total grid locks.
So please next time, can you try a start at Am rather than 8:Am. Not everybody live on Stillorgan road.
OK, now back to the main story, the conference. 5 parts, Longhorn overview (David Chappell), Avalon (Lester Madden), WinFS (Hans Verbeeck), Indigo(Clemens Vasters) and Whidbey (Nigel Watling).
I don't know if it was the time, but everybody was so silent. No loud 'Developers, developers, developers', no clapping, just a religious silence.
I can tell that Clemens Vasters had the most difficult part Indigo, with no demo, and just after lunch, the right time for a nap. By the way, very interesting piece of technology which will be made available before Longhorn, something for next year.
Just few thoughts on what I saw today. But before of course this is just my opinion and maybe I am wrong.
Avalon presentation give me this idea of deja vu, and after few demos I got it. It was seeing Logo meeting XML. No I am not joking. For the youngsters, Logo is this brilliant language created in the 70s/80s, using the concept of turtle to draw anything. XAML has this notion embedded in it, of course using more an XML approach, but the attributes, the coordinates, etc.. are stuff already done in the past.
The thing I don't get with Longhorn is that as developers, we can create a lot of nice things, but only for Longhorn users. for the other OS, we have to provide a downgraded version, so no fancy rotated animated button.
If I am right, Longhorn is not a server, so how we will deploy a navigator form for other platforms ? The answer this morning was not expected, create 2 versions of your application. Not sure I like that.
WinFS is surely a great idea at a high level semantic discussion, but it seems for me that something going wrong there. The demo today show a user creating a Word document, filling some properties like the Author name, a Title and save it.
This is where Microsoft miss something. Who knows a single user in their company who fill the properties in an Office document. Myself I don't know any single user doing this.
What I am saying is the 'average' user will never benefit from WinFS if we (developers) don't provide a better integrated system to enter the precious Metadata. The fact that NTFS is still there is not going to have the people moving to WinFS storage. What is the point to have a fantastic (no doubt about this) relational database feature if we have to spend our time answering questions my machine can do for me?
A basic example should be the Author. I would like to see WinFS storing this property for me. But I also saw a glimpse of what could surely be the future of WinFS. A prototype by MS Research of a face recognition software. That's exactly what we need. Multimedia stuff is extremely difficult to index and organize.
I remember this fabulous software for Macintosh in the 80s called Phrasea (Not sure it's still available). Indexing was excellent on this and they spent few years to have something really working. And there we are talking about 2 to 3 years for Longhorn ? I am not sure we will see the real storage revolution we need with high disks capacity.
To come back a second to this face recognition stuff, it was really cool. In a photograph, the software identified a zone around the head, and was able to link this to a particular name. This is what we need to see at Longhorn launch, some automatic and bluffing applications to show the advantages of spending time on cataloging our data.
I know NTFS is also there for backward compatibility, but it's surely true that a lot of users would still like to use the Folders architecture, and never see the benefits of WinFs.
The demo about a Law firm company with all their lawyers, cases clients was really cool, but I know myself a lot of lawyers who still use Word to print a letter and full stop. They don't use at all the current potential of their machine, so WinFs is for them a piece of science fiction.
Maybe Sharepoint is a more realistic approach for today.
For the other topics, nothing special to say. Indigo is great, but I am not a specialist in WSE and other things, but Clemens you did a great job. It opens for me some horizons.
Whidbey presentation was a bit messy, jumping from C# to VB, IDE back to C#, and Nigel tried is best to impress the crowd, but I think everybody had enough at this stage. The good thing is the promising delivery date for .Net 2 ... 2004 !
So finally I say Longhorn has a lot of good ideas, nothing really new, but it's going to be a long road until the release ;-)