The Best Ever

  • Comments posted to this topic are about the item The Best Ever

  • SQL Server 7 gets a vote from me.  I used to nurse maid a SQL6.5 instance and had to upload data from a client on a weekly basis with a tight SLA.  SQL6.5 used to take 17 hours and then probably fail.
    The first time I tried SQL Server 7 it took under an hour.  It was quick enough for me to assume it had failed but it hadn't.  Also DTS made its first appearance.

    SQL2008 Gets a vote for the massive performance improvement with replication.

    SQL2017 its too early to say but the inclusion of Linux has to be a point in its favour even if very few people are using it in production yet.

  • The version that supports the SQL:92 DOMAIN functionality? The MS-SQL specific TYPE functionality is not the same particularly as it does not support all the functionality of DOMAINs - therefore doesn't help much in the long run in this regard.

    SQL 2016 resolves a lot of annoyances though, in particular I like the CREATE OR ALTER <x> functionality, even if SSMS does not support it yet when extracting generation code from the database.

  • SQL 2008 R2 for shared datasets in Reporting Services!
    In general, huge improvements were made in Reporting Services from 2005 to 2008.

  • 2008 would be #1 for me
    2016 would be #2

  • Latest is best
    Fond memories of SQL 2012 with the introduction of analytic functions

  • SQL 2008R2 gets my vote.  DB Backup compression in Standard Edition was a very welcomed enhancement for me.  Version was very stable. Quite honestly I would run it till the end of time if I had to and not blink.
    One of the biggest jumps in features and stability HAS to be SQL Server 6.5 to Version 7.  No more creating data files on devices... AND the database could automatically extend in size.

  • I've got to go with 2017. With SQL Server, Microsoft has done quite a good job of incremental improvements and removing cruft over the past 17 years.
    So pretty much any pro of the previous versions applies to 2017.

    Plus we (FINALLY) get STRING_AGG() and Linux support!

    (I'm sure there are a few other improvements in there too ; )

  • I hadn't spotted the introduction of the STRING_AGG function. That would save a lot of rather horrible and convoluted queries that do the same - with the likely further advantage that the STRING_AGG is not as glacially slow as the combination of the usual MS-SQL string handling functions are when combined together.

    Outputting data in a certain format tends to start religious wars about whether or not this is the job of the database or the application.

  • Probably, SQL Server 7 - DTS first implementation, and SQL 2005 with SSIS as an upgrade of DTS.

  • SQL Server 2016 SP1 - standard edition with most of the enterprise features, what is there not to like 😀

  • My choices reflect heavy use of ETL for data warehousing:  SQL 2005 for the introduction of the Unified Dimensional Model in SSAS.  Significantly if not radically changed the development of cubes.  Thumbs up to the same model for the introduction of SSIS.  Once the step learning curve overcome hard to look at a DTS package anymore (and I thought those were great at the time).  Next vote to SQL 2012 for the SSIS changes to the project deployment model.  Made the parent-child development method significantly easier.

  • Stefan LG - Tuesday, April 10, 2018 3:13 AM

    SQL 2008 R2 for shared datasets in Reporting Services!
    In general, huge improvements were made in Reporting Services from 2005 to 2008.

    This one for me and even on Standard edition.

  • Well, asking for my favorite SQL Server version isn't the same as asking for my favorite Smashing Pumpkins album release. SQL Server has a history of linear improvement. I may have some sentimental attachment to 2000 and then 2008, because I have spent most of my screen time with those versions. But I'm not one of those who are skeptical of the newest release; 2012, 2014, and 2016 have all been solid products with improving features, and if I had my choice of version for a new server, it would be 2017.

    "Do not seek to follow in the footsteps of the wise. Instead, seek what they sought." - Matsuo Basho

  • SQL Server 2016 is first
    SQL Server 2012 second
    SQL Server 2017 will probably be first now because it has both R and Python.

Viewing 15 posts - 1 through 15 (of 33 total)

You must be logged in to reply to this topic. Login to reply