Microsoft Failures & Successes

<< Code Zest Top Coffee Goodies | Home | N-Layered != N-Tiered: Know the Difference! >>

Microsoft Failures & Successes

Technorati Tags: Opinion

I will give no explanation, because I think they speak for themselves, but here are some failures & Successes from Microsoft.

Failures

(not wasting my time on these, they just don’t let me easily and openly code good ORM, Patterns, etc. period)

Typed DataSets
SharePoint 2003 & 2007
New Entity Framework
MS CRM

Successes

C# 3.0
.NET 3.5
VS 2008
ASP.NET MVC Preview
LINQ

What is your list? What do you think?

Print | posted on Tuesday, October 07, 2008 5:41 PM

Comments on this post

# re: Microsoft Failures & Successes

Pretty much agreed on all, with the exception that I would split up SharePoint - leaving SharePoint 2003 in the failures column, while moving SharePoint 2007 in the successes column.

SharePoint 2003 really was quite weak, an anemic thing that overall was just too ambitious to completely realize (even for Microsoft) using the technologies available at the time. However, I think that SharePoint 2007 out of the box, especially if configured properly, really can be a useful repository for all types of information. What I've seen developers really complain about is when trying to extend or otherwise use the SharePoint object model...

Just my take.

Left by elèyn on Oct 07, 2008 10:37 PM

# re: Microsoft Failures & Successes

>>>What I've seen developers really complain about is when trying to extend or otherwise use the SharePoint object model...

But that's precisely my point of failure. What's the good of an application which prevents you from extending both the Database and object model in an open way that allows you to put in your own design patterns, and just lets you code straight OOP so you have control of quality, extensibility, anything rather than spend time finding work-arounds, hacks, or potential bugs that you have no control over in the development and testing phase.

The ultimate measure is, "Can we support any business requirements now and in the future easily". With that type of API, you will be spending more time figuring out wacky ways to use it rather than have used standard C#, .NET, VB, or whatever and had it done.

I find that packages that "do it all" end up being a huge wall than a help and you end up realizing hey, I could have gotten this stuff done a hell of a lot faster with just good OOP design in the first place up front when you have good developers. If you have to extend things in "funky" and "non traditional ways", you might as well throw it in the trash.

Left by Dave Schinkel on Oct 07, 2008 10:42 PM

# re: Microsoft Failures & Successes

Ah yes, but you are assuming that every business that wants to run SharePoint automatically wants adapt it in radical ways. To use an analogy, most people who install Microsoft Word really and truly just want to use it as a word processor. Sure, you can extend it (I think), and there's an API (I think)... but my point is, compared to SharePoint 2003, SharePoint 2007 really is usable out of the box as tool on a corporate LAN for collaboration, calendaring, document storage, not to mention for hosting Project Web Access (now, whether PWA is good is another story).

That said, trying to morph SharePoint into working in considerably different or vastly extended ways (e.g. creating a nation-wide document repository that also ties into other custom business apps) - for sure, the limitations (or inherent nature?) of the API will drive you nuts. In that respect, from what I hear, SharePoint blows...

What I'm thinking is that the list you were putting together is more like "Microsoft's failures/successes in development tools or platforms." Swinging that way, one could add:

-ASP.NET
-SQL Server

Swinging the other way (the mixed list you have now), one could add:

-Microsoft Office

Left by elèyn on Oct 08, 2008 7:55 PM