Correct paths to vsix

* poss fix

* poss fix

* Update range.fs

* Update range.fs

* fix build

* docs and limit fix

* Update src/fsharp/range.fs
Co-Authored-By: Ndsyme <dsyme@users.noreply.github.com>

* Update range.fs

* possible fix for signature issue

* fix de-duplication

Fixes #6068, where you could cause a `System.OverFlowException` by using a key and value that leads to calling `abs` on `Int32.MinValue`. This is invalid on .NET. The fix is to not call `abs`, since there's really no reason to make a hash code always positive.

* remove allocations from jaro

* improve perforamce

* fix build

* remove string concat allocations from FilterPredictions

* fix build

* move to struct tuple and remove the concat completely

* undo

* Dive into types in type test expressions in AST visitor

* Add test

* Update VisualFSharp test

* Update fsproj

* Cleanup test

* Replace isModule with ModuleOrNamespaceKind in SymModuleOrNamespace

* Fix structure tests

* Review fixes

* Improve workflow

* back to 32bit msbuild

* move out nugets, change targets in makefile

* Detect framework

* Done

* feedback

* fix assembly path for signing

* don't unconditionally deploy the templates VSIX on build

* ensure insertion manifest is written to the correct location

* correct package version

Initial benchmarking

* Fix #4946  --- Update FSC and FSI to recognize upto .net framework 472

* support really old frameworks

* Update test to satisfy net472 as well as earlier versions of dotnet

* comparer

* fix options ui logic and syncing between IDE instances

* add comments

* feedback

* add notes on FsLexYacc

* add notes on FsLexYacc

* add notes on FsLexYacc

* add notes on FsLexYacc

* add notes on FsLexYacc

* convert to SDK projects

* fix ci_part4

* convert to SDK projects

* Done

* Init command

* ws

* Update FSharpBuild.Directory.Build.props

* feedback

* Don't needless allocate a Result type during to to definition

* prevent rebuild

* prevent rebuild

* no-op change

* Remove some asses from the codebase

* Remove more asses

* Fix #5531

This is a working fix for the example of #5531.
However there are still some points open before I think it's a good idea to merge this:
- Discuss the change of this proposal, is it OK to prefer overrides of methods and so ignore base implementations?
- Extend the check to the complete inheritance graph instead of a single "look-back"
- Only ignore a methinfo if the signature of both match (so respect different overloads)

@dsyme It would be great if you could have a look whether this check is allowed at this location or should appear earlier.

* Fix left points.
Signed-off-by: Nrealvictorprm <mueller.vpr@gmail.com>

* another approach to fix

* remove old approach

* Remove approach again and apply different approach.
Signed-off-by: Nrealvictorprm <mueller.vpr@gmail.com>

* Adjusting predicate to reflect correct behaviour.

* Revert to old List.merge strategy.

* Apply review and add tests
Signed-off-by: Nrealvictorprm <mueller.vpr@gmail.com>

* Try fixing weird CI failure
Signed-off-by: Nrealvictorprm <mueller.vpr@gmail.com>

* Fix test.
Signed-off-by: Nrealvictorprm <mueller.vpr@gmail.com>

convert src/fsharp to be SDK projects

* small fixes

* fix test

* reduce memory allocations

* remove memory allocations

* Fix and tests, unbuilt and untested

* Correct one of the tests

Got the order wrong for the output of `countBy`.

* Repair the groupBy test

groupBy test was bad: it assumed (ridiculously) that groupBy wouldn't iterate the original sequence even though it somehow knew the groups in the sequence.

* remove massive string allocations

* change NormalizedSource Source and Seq to Array computation expression

* fix line end according to the source positions

* remove pos use i

* fix some tests

* CP: Use Framework valuetuple rather than netstandard1.0 one (#5872)

* Use Framework valuetuple rather than netstandard1.0 one (#5867)

* remove references from template

* update templates