Never doubt thy debugger

What about the "Check for Updates" feature that is supposed to work in VS.  It's been there since the first VS and still doesn't "work" in VS2008.

Why depart from the paradigm of "Microsoft Update" which is simple and can even automatically download updates and keeps track of what is installed and applicable to the system.

It's a PITA to get the Hotfixes and keep track of them.  Why can't this be made easy for developers?  We already have enough to do - keeping on top of and learning new Technologies & Frameworks while still writing code and keeping projects on schedule!  We shouldn't have to worry about tracking every individual Hotfix for the Tools that we use and rely upon to do our job every day!

In general, I agree with the tenet of "if it ain't broke, don't fix it".  That being said there is something to be said for keeping your install as up to date as possible.

There are 5 main issues I see:

1. "Synchronization" between Developer workstations.  In other words, *everyone*, including the Build box (if using CI), has exactly the same configuration, environment, and Hotfixes.  This extends beyond the Development environment to Integration, Testing, and Production servers as well (however, those don't have VS installed but the principle is the same).  I have seen situations where a Hotfix applies and resolves a problem on one Developer workstation yet another workstation doesn't have the same problem.  Obviously, there must be some differences between the two.  However, because there are no real solutions for doing diffs on machines, the easiest solution is to apply the Hotfix to all workstations.  At least we know then there is 1 less difference between the group.

2. The 2nd issue is one of "side-effects".  A Hofix can often cause side effects to something that is seemingly unrelated but still affects your application or environment in some unforeseen way.  This is often discovered after much pain and agony spending hours and days trying to find out "what happened".

3. The 3rd issue is one of not knowing *exactly* if a Hotfix applies to your problem or not.  Often times we will run into an issue and find a Hotfix which seems to be applicable, install it and find that it didn't work.  Then continue on looking for the next possible solution.  The Hotfixes which didn't apply may never be removed.

4. Hotfixes often have other Hotfix dependencies either from a pre-requisite perspective or from "interacting with one another".  I've seen it happen that 2 different hotfixes have unexpected behavior when they are installed on the same machine.  Inother words, Hotfix A will cause Behavior A, Hotfix B will cause Behavior B, and the combination of Hotfix A and B will cause Behavior C (instead of A + B).

5. Lastly, in theory aren't all of these Hotfixes eventually rolled-up into a Service Pack anyway?  This means you'd have Hotfixes installed regardless of whether or not they apply to your system.

Things keep moving and after some discussion ( here and here I already expressed my view on the matter),