Potash Valley

Nice one! How I could've missed that?:)

This post is just what I needed to see. I'm looking forward to part 2

Tim

Félicitations!!!

Là tu ne peux pas être plus au cœur ;)

Always the best forever!
Bonne chance à toi!

Je te souhaite bcq de succes pour ce nouveau challenge.
Félicitations!!!

Fécilitations à toi, que du bon.
Bonne continuation dans ta nouvelle carrière. Le prochain moteur ASP va tout casser !
A bientôt.

Congratulations! And good luck for your new pro experience and life! Hope to meet you in Redmond some time.

Félicitation!

Welcome to the team, and have a safe trip.

(and in case some are wondering, this is no April fools joke, Sébastien is going to be part of my team)

Félicitations et bonne chance pour cette nouvelle aventure !

Hello,

The contrary is also true: Don't let designers working alone.
In this case you'll certainly produce a nice piece of engineering, but useless for business...

Communication should be continuous, and use a common language that both world understand (e.g. usecase don't need much background).

I fully agree with you and this is where UML modeling is stupid if used with MDD (e.g. model driven development). How business could alone decide about requirements if it doesn't see what would be the result. After a first iteration the architect or program manager shows to the customer what has been developped. After this stage they together decide to adapt requirements and what is really possible. After few iterations the project is getting better and better.
The problem with UML is that you can't add presentation layers and deploy the application without developers codding methods. If you code manually methods then the traditional MDD is impossible because only code from a static model is possible. This is why MDD is stupid !!
I mean that UML as a language is certainly the best way for documenting a project but using the model in a waterfall approach with MDD is just ridiculous !!

I have written an article on this waterfall versus agile approach available at: http://www.forum-omondo.com/documentation_eclipseuml_2008/waterfall_versus_incrementale_modeling_cycle.html

Vlad,

I will just say YES !

Merci pour cette précision.
Je n'avais effectivement pas vue que le projet ciblé entre autre .NET 2.0

J'approuve donc 100% (voir plus) ce projet.

Nice ;)

S'il n'y avait vraiment aucune différence, on n'aurait pas pris la peine de gérer les deux. La version string permet de garantir que les utilisateurs sous .NET 2.0 peuvent aussi utiliser la même syntaxe. Du coup il n'y a qu'un seul langage de requête à connaitre, et en plus il est légèrement standard sur .NET, LINQ.

Je ne suis pas (encore) trop le projet mais je vois déjà une différence de taille entre les deux approches.

La requête NLinq est écrite dans un string, on ne peut donc apparemment pas avoir accès à l'intelliscence.

Stop right there. :) EcmaScript 4 is DEAD. It will never be. It's ES 5th edition now.

I used to have similar problems with some others rss readers. I didn't use IE for reading my feeds (neither for browsing actually) so I don't have a clue about what's going on here.

But I can recommend you Google Reader. I use it for ages now, and it work great. No more feed issues, weird sync problems, etc. It supports offline reading, it performs well, and is synchronized by nature.

JScript.NET is basically old, deprecated, and has never been updated since .net 1.1. Doesn't really sounds like a good choice.

AFAIK, there's no plan to have a DLR based jscript in .net 4.0.

They do have one though, that used to ship with the early betas of Silverlight, but it's nowhere to be found nowadays.

For heavy usages, you'd get a performance boost. That being said, both IPY and IR have transitioned to a mixed interpreter / compiler state, where only the code called often gets compiled.

You could get debuggability of the scripts. Also you'd get interop with the other DLR based languages, or languages that support it. For instance you could have a «dynamic» C#4 object talking «natively» to a Jint object.