Consider removing RelaxedLayout and friends

These three parsing-related language extensions were introduced some time ago yet are likely not used in the wild as they still are not registered in Cabal's language extension list (which Hackage uploads are checked against, see #8176 (closed)),

• RelaxedLayout (introduced in 7.2, prompted by #1060 (closed))
• AlternativeLayoutRule (introduced in 7.0, reason unknown)
• AlternativeLayoutRuleTransitional (introduced in 7.0, reason unknown)

As far as I can tell no one knows the status of these, nor are they listed in the users guide.

We should ask the community whether they would mind seeing these extensions vanish and remove them if not. If so then we should see to it that they are registered properly with Cabal so they can be used on Hackage.

Trac metadata

Trac field	Value
Version	7.10.3
Type	Task
TypeOfFailure	OtherFailure
Priority	normal
Resolution	Unresolved
Component	Compiler (Parser)
Test case
Differential revisions
BlockedBy
Related
Blocking
CC
Operating system
Architecture

mentioned in issue #8176 (closed)

mentioned in issue #10935 (closed)

changed milestone to %8.2.1

changed weight to 5

added Ttask Trac import parser labels

More information about AlternativeLayoutRule and AlternativeLayoutRuleTraditional: https://prime.haskell.org/wiki/AlternativeLayoutRule.

It is used for multiline mode in GHCi (:set +m, #4316).

changed milestone to %8.4.1

Given that 8.2.1-rc1 is imminent, I'm bumping these off to the 8.4

changed milestone to %8.6.1

Perhaps we can do this in 8.6.

This ticket won't be resolved in 8.4; remilestoning for 8.6. Do holler if you are affected by this or would otherwise like to work on it.

changed milestone to %8.8.1

These won't be fixed for 8.6, bumping to 8.8.

changed milestone to %8.10.1

Bumping milestones of low-priority tickets.

added Pnormal label

removed milestone

Bump, 7 years later. We should make progress here. Recently it slowed down @knothed's work in !9229 (closed), so perhaps he would be interested in removal of these extensions?

Cursory look at https://hackage-search.serokell.io/?q=AlternativeLayoutRule suggests -XAlternativeLayoutRule has never really been used. How would it, even? No one knows what it does. Simon didn't have a clue almost 10 years ago, either, as per the mailing list thread that Ben quoted (https://mail.haskell.org/pipermail/ghc-devs/2013-September/002267.html).

Yes, I didn’t even know these existed, just bumped into them when looking through the Lexer and realised that they make my life harder when parsing Or-Patterns.

So I’m in favour of removal.

mentioned in merge request !9229 (closed)

marked this issue as related to #4316

Happened across this today. I have figured out what this flag does. Here might be some text for the manual (though I'm also quite happy just to remove this apparently-never-used feature).

The layout rule means that, while processing a Haskell file, there is always a column number
remembered as the current level of indentation. (At the top level, this number is 0.) When
a line begins at this column number, a semicolon is implicitly inserted, and when a line
begins to the left of this column, a close-brace is implicitly inserted.

With ``-XRelaxedLayout``, if the character that appears to the left of the existing indentation
column number is a ``{``, then the close-brace is not inserted. As usual, in between explicit
braces, the layout rule is not in effect. After the close-brace that matches the open brace,
the old indentation level column number is still in effect.

As an example, with ``-XRelaxedLayout``, you can write ::

  f b = do
        b1 <- b
        q <- if b1 then return True else do
     { b2 <- b;         -- "{" is to the left of the "q" on the line above
    return (b1 || b2) } -- this is even further left
        return q        -- this must line up with the "q <-" above

This extension might be useful if you end up drifting to the right with many nested layout blocks. It allows a reset. Note that layout can resume within the explicit braces. Probably isn't worth it.

mentioned in issue #25322