Support for SQL Server #544
Conversation
* Removed support for python 3.7 * Added common and dialects packages to handle the new SQLAlchemy 2.0+
* The connections were hanging sometimes during the tests.
* Revert changes automatically applied by IDE * Fix formatting of CI caused by IDE changes * Codebase cleanup for SQLAlchemy 2.0 * Fix EOL in mkdocs
|
Do we want to extend Databases with new backends when library is basically on life support? 🤔 |
|
@rafalp the lib is actually very stable and works for 99% of the projects as is. With the latest 2PR being reviewed we go to the next level. Unless @aminalaee and @Kludex feel that Encode doesn't want to continue with this. If that is the case, there is always a way to continue anyway :) but I believe Encode wants to keep maintaining this :) |
I'm also Encode member and I was on meetings where we've discussed how to carry on with databases with my proposition being to keep it around because its great as lightweight async DBAL alternative to solve-it-all SQlAlchemy 2 ;) I think we should update the readme to make it more obvious that the plan is to keep the library around for a while, especially when people are seeing PRs removing mentions of Databases from Starlette's docs and asking questions. |
|
@rafalp I didn't mean any disrespect about you ahahah. I respect Encode tremendously :D. A lot of what I do is inspired by the great work you all do. I have more than interest to use databases and the reason why I "created" the driver for SQL Server is because I'm currently working on projects that use heavily SQL Server and the amount of code being used there is so ugly that databases solves it in a simple fashion. I never wanted to fork databases in order to put SQLAlchemy 2 and that is why I tried to open the PR here with the migration working but I also believe Encode knows what is the best course of action and I will always support the decisions 👍🏼 If you are Encode, I'm sure you are a fantastic developer! :D This PR althoug it looks big, it is mostly because I branched of the one implementing SQLAlchemy 2.0 but not yet merged. If Encode really wants to stop maintaining Databases (which according to you, it doesn't look like it). I offer myself to continue with the package. |
|
TBH I've always been thinking that Databases should support external drivers, so adding new databases to it should not changing the core library. I'm not familiar with SQL Server but this PR seems to mix SQLAlchemy 2 support changes with SQL Server changes. |
No, this is the same level of integration that is done for Postgres and MySQL but supporting SQLServer as well. SQL Alchemy has support for it already. What I'm proposing is simply to also add the same layer to MSSQL since it's widely used. The PR it looks big beause the SQL Alchemy 2.0 changes were not yet merged. It is here. Because I wanted to use already SQLA 2.0, I branched off that branch and added the SQLServer driver. Once that PR is merged, if is merged, you should only see 3 or 4 files only here to review |
@aminalaee @Kludex
This is a PR proposing support for SQL Server. A lot of companies still use it and with
aioodbcand databases patterns, this was possible to achieve as well. This could be merged maybe after this one? The reason why after its only because I branched off that branch to create this new one and once it's merged, it is way less files to review.