Once a table is created, the ordinal position of the columns is fixed. This is especially problematic for large data warehouse tables, where new columns are expected to be added over time.

The only solution currently to re-order columns is to drop and recreate the table (or let the SSMS designer do this for you), but this is often not feasible for a large, highly used table, and it cannot be easily scripted.

Please add some ALTER syntax that will allow changing column order.

When you have:
1. SSMS set up as a default editor for *.sql files (and)
2. SSMS already started with "Connect to Database Engine" dialog window open
and you are trying open a .sql file by double-click, it starts another instance of SSMS and opens .sql file there.
My suggestion is to change the behavior to open the file in the exisiting instance of SSMS.

SQL Server Basic Availability Group should allow N number of databases to be supported. Currently this feature supports two replica's which is fine, but only support one database to part of the basic availability group. As always-on is strategic offering for availability and has been around since SQL 2012, the basic availability group feature in standard edition should allow more than 1 database. I would suggest N databases or whatever the Always-on technical limit is.
https://docs.microsoft.com/en-us/sql/database-engine/availability-groups/windows/basic-availability-groups-always-on-availability-groups?view=sql-server-2017
Other database vendors offer such feature outside of enterprise edition.

It's not great to have to manually reference Microsoft.SqlServer.SqlParser.dll when we need to parse TSQL code.

Open sourcing and releasing as a nuget package would open up a number of very interesting scenario that would empower the sql server community.
For us, we want to pre-process our scripts with macros and also format consistently but I imagine there would be other cool scenarios such as linters and code analyzers.

Would love to see sql server go the direction of nuodb or memsql and implement a fully distributed version of MS SQL SERVER.

Currently SSAS (2016 or 2017 ) model does not have the ability to show measure descriptions or a custom Text as a tooltip for users that are browsing cube.

I'd like the abilities to favorites stored procedures, tables, anything, for quick access.

Add support to In-Memory OLTP to ON DELETE CASCADE / SET NULL on foreign keys deletion

Please add the filter option on ssms to filter user defined table types by name.

SSMS 2018 Source Control

Do you we enable source control under SSMS 2018?

SQL Server 2017 introduces read-scale availability groups without a cluster.
Documented here:
https://docs.microsoft.com/en-us/sql/database-engine/availability-groups/windows/configure-read-scale-availability-groups?view=sql-server-ver15

Looking at the feature by edition (https://docs.microsoft.com/en-us/sql/sql-server/editions-and-components-of-sql-server-2017?view=sql-server-ver15#RDBMSHA), it mentions that SQL Server 2017 Standard Editon supports 'Read-scale availability group.' In Standard Edition to read from Secondary Replica of Read-scale Availability Group, you need to create a snapshot of the database. This point is not mentioned in any documentation, which confuses when people try to implement this feature.
I suggest making it clear in the documentation that the only way to read from Secondary is to create a database snapshot while using Standard Edition.

SSDT is great once set up, but the initial amount to get some real world code working is often a show-stopper.
Having to use variables for cross-database references is often over-engineered for simple projects. Please let us just use the simple notation [server].[db].[schema].[object] as in SQL Server. This is enough for 99% percent of the cases and the SQL Code does not have to be changed after import.

As of right now the ODBC drivers for SQL Server do not support Linux on ppc64 architecture. Many HANA systems are running on ppc64 today. Therefore it would be good to offer a driver for this architecture aswell.

The generally accepted way to truncate some part of a datetime is:
DATEADD(datepart,0,DATEDIFF(datepart,0,datevalue))
To save us from much typing, and probalby a good amount of CPU due to a good implementation, I propose we create a new function exactly for that:
DATETRUNC(datepart,datevalue)
which rids the datevalue of any component LESS than datepart (ie, truncates it to the dateart level)

Provide a method to force a rollover XE and Audit rollover file targets on demand.

This functionality will allow one to reduce event latency for batch consumers without incurring event loss. Currently, one must stop/start the trace/audit or restart the instance in order to create a new rollover trace file.

Since the new SQL documentations online (e.g. https://docs.microsoft.com/en-us/sql/relational-databases/system-dynamic-management-views/sys-dm-os-nodes-transact-sql?view=sql-server-2017) there is no hint since which SQL Server Version a DMV or specific T-SQL command (e.g. IIF()) was added.

Hi

I am trying to use the Database diagram option which is supposed to be within the folder of the database, but nothing is there, i can only view System tables,FilesTables,GraphTables and then my tables.

I would like to reuse a Variables type from a Columns Type. There would be no need to check the ColumnType anymore - and if a column is altered the Stored Proc should be recompiled.