The beginning of a more in depth look at SQL Prompt 3.5

Well I said I’d post some more detailed information about the new version of SQL Prompt so I thought I’d better get down to it. I wanted to do this to coincide with getting the release candidate out there, which we’d planned on being this Friday, however we all had a bit of a rush of blood to the head at about 5 o’clock yesterday afternoon and decided to just jolly well release it, and frankly I’m glad. We were holding on only due to a couple of issues with the installer:

  1. If Visual Studio 2005 has never been used prior to installing SQL Prompt 3.5 then the SQL Prompt menu will not appear until Visual Studio is restarted. I can’t imagine too many people are going to run into this one, however if you have I apologise. The reason we encountered it is that we generally do install testing using virtual machines, which although I hate them with a raging passion, do have some uses.
  2. If you’re upgrading from 3.1 the tray application restarts during the installation but SQL Prompt won’t work in Query Analyzer until you manually restart it again. This is clearly going to affect a lot more people, however we decided that it wasn’t serious enough to hold up the RC any more.

I must admit it felt just slightly cavalier, however when you bear in mind that we’d been trying to fix just these two issues since last Thursday you can see why some of us might have been losing the will to live because they were holding us up from making any real progress and (1) in particular just isn’t that big a deal.

So… it’s out there in the wild, and already we’ve had one person who’s reported an issue with VS 2005 that’s likely to affect a number of people that we wouldn’t have found for some days, if at all, and which we’ll be fixing for the final release. And that’s why we released it: we reached the point where the user feedback was the thing we most needed. Getting a few hundred, or even a few dozen, people with widely varying system configurations to take an early look at a product is a really good way to shake out those last few pesky bugs.

There’s a whole philosophical debate then about whether or not this really constitutes a release candidate if we’re intending to make some changes, however what I’d say is this: we put out a release candidate when we think something is good enough to release (apart from the installer issues in this case), but I have never yet worked on a product here at Red Gate where we didn’t end up fixing at least a couple of bugs between the RC and the final release, so in this case we just did the pragmatic thing and kicked it out the door.

Anyway, I’ve deviated somewhat from the point, and now I’m (again) running out of time. In my last post I briefly listed some of the new features and enhancements we’ve added so over my next few posts I’m going to take a look at these things individually along with a few other improvements in more detail.