add ordNub somewhere in containers

[gbaz]

After dropping the standard ordNub (i.e. this: https://github.com/nh2/haskell-ordnub) code into a codebase for the zillionth time I went and checked the history of why it never made it into a standard lib. It can't go into base because it requires containers or reimplementing Set just for this purpose, which is silly. It doesn't necessarily belong in the Data.Set module. But perhaps it could fall into a new module, like Data.SetUtils or Data.ContainerUtils or the like? Anyway, this package is the most natural place for it to live, even if it doesn't quite fit, and everyone ends up using it everywhere, so it would be really nice to just have a standard for it...

[treeowl]

There are actually two containers supporting `nub`-like operations with different constraints: `Data.Set` and `Data.IntSet`. Does it make sense to put these operations in? I dunno. It feels like a bit of a strange place. I suggest you raise the question of where these things belong (and what they should be called) on the libraries list.

…

On Thu, Oct 5, 2017 at 5:28 PM, gbaz ***@***.***> wrote: After dropping the standard ordNub (i.e. this: https://github.com/nh2/ haskell-ordnub) code into a codebase for the zillionth time I went and checked the history of why it never made it into a standard lib. It can't go into base because it requires containers or reimplementing Set just for this purpose, which is silly. It doesn't necessarily belong in the Data.Set module. But perhaps it could fall into a new module, like Data.SetUtils or Data.ContainerUtils or the like? Anyway, this package is the most natural place for it to live, even if it doesn't quite fit, and everyone ends up using it everywhere, so it would be really nice to just have a standard for it... — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#439>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABzi_TE_UVwYDiR8D1EDnZfOvzlMowMZks5spUoBgaJpZM4PvudX> .

[nh2]

Libraries discussion: https://mail.haskell.org/pipermail/libraries/2017-October/028269.html

[gbaz]

As I wrote there:

"The discussion period is well over and it seems the consensus is for
adding this, basically as David proposes. -- i.e. ordNub, and
ordNubOn, but not ordNubBy. (and also the -1 variants and the int-
variants). I think the only module name proposed to add them to is
Data.Containers.ListUtils, so I guess that stands.

David -- are you up for adding these, or would you like someone else
(i.e. me) to prepare a patch?"

[treeowl]

Either way.

…

On Dec 18, 2017 9:20 PM, "gbaz" ***@***.***> wrote: As I wrote there: "The discussion period is well over and it seems the consensus is for adding this, basically as David proposes. -- i.e. ordNub, and ordNubOn, but not ordNubBy. (and also the -1 variants and the int- variants). I think the only module name proposed to add them to is Data.Containers.ListUtils, so I guess that stands. David -- are you up for adding these, or would you like someone else (i.e. me) to prepare a patch?" — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#439 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ABzi_de7hPTq0MjBlqoOo4qOTu2qvs0iks5tBx13gaJpZM4PvudX> .

[treeowl]

We'd better get to this.

[gbaz]

See PR