This month is an interesting T-SQL Tuesday party, as Glenn Berry hosts and asks us to think about the upcoming new release of SQL Server. Sometime later this year, I expect SQL Server 2022 to be released and Glenn is asking us to talk about our experiences.
If you haven’t looked at this new version, which is in RC0, you might take a few minutes to play with it and see if the language (or other) changes might be useful in your organization.
If you want to host a T-SQL Tuesday yourself, ping me on Twitter.
My Work with SQL Server 2022
I first saw some demos of SQL Server 2022 in 2021, at various conferences. Microsoft showed off some new capabilities, some of which were very interesting. I was lucky as an MVP to get some access to private, pre-CTP builds and experiment a bit with new features.
This spring Microsoft publicly released CTP 2.0, then 2.1, and now RC0. I’ve upgraded to 2.1 on my desktop, and was waiting the docker container to update before moving to RC0. just after Glenn’s invitation, I saw the container was up to date, so I pulled a new one.
I’ve run some of my old demos on the platform, just to check that they work. That’s really work to see if there are any regression bugs. I have rarely found this to be the case, but it has been interesting to do this with 2017, with the linux version, and more. I’ve only been lightly interested in this release, as a few of the changes aren’t applicable for the work I do with Redgate. Others might be interesting to the community, and I need to spend more time on them.
Really, I got mostly interested in the T-SQL language changes. I had been using Window functions and the OVER() clause quite a bit and find writing them cumbersome, but I was excited to see the SELECT..WINDOW clause. I’d also liked the STRING_SPLIT enhancements with an ordinal.
There was an article sent to SQL Server Central that got me to look at more of the features, and I think that while most of these aren’t changes I’ve been needing, they do improve and round out the language. I look forward to experimenting with them a bit.
I’ve reproduced a few performance demos, looking at the query optimization features, and those seem good, but for me and many other developers, these will just be things that should work. Not something to get too excited about. Unless you’re on call, then you might really want to upgrade and hope these fix some of your problems without creating other ones. Maybe the thing I’m most excited about is the granular UNMASK permission, which has been overdue for a couple of versions.
I don’t quite know what to think of this new version. While there are more changes than I expected, it feels like a lot of small changes and not much of a fundamental shift in the product. I suppose it’s a major release, but kind of like SQL Server 2014, this feels more evolutionary than revolutionary.