scala.util.Try

Tony Morris @dibblego

@jsuereth @mergeconflict No, not sure why you think that. The current implementation of Try is has no redeeming qualities.

Josh Suereth @jsuereth

@dibblego @mergeconflict well, scala blends inheritance + FP. Doesn't seem your first choice of design goals.

Tony Morris @dibblego

@jsuereth @mergeconflict Scala makes some awful mistakes and these have been discussed at length on many occasions with honest people.

Tony Morris @dibblego

@jsuereth @mergeconflict I long ago accepted the mistakes of Scala and not long after, that they were unlikely to change.

Tony Morris @dibblego

@jsuereth @mergeconflict The Try data type is a different type of mistake, especially that although many of the Scala libraries are poor...

Tony Morris @dibblego

@jsuereth @mergeconflict ...they usually have some redeeming, even if degenerate, quality. That cannot be said for this latest case study.

Josh Suereth @jsuereth

@dibblego @mergeconflict Yeah, I'd love to understand what the problem is, besides having too loose a definition of arrows in your category.

Josh Suereth @jsuereth

@dibblego @mergeconflict Which is a general issue with CT and Scala's A => B types. You can't enforce a categorical limit A => B well.

Tony Morris @dibblego

@marius @mergeconflict I agree, but I've had to do it before (especially in JDBC). It's possible to do without bugs (scala.util.Try).

marius eriksen @marius

scala.util.Try (or, Thar be Trolls) http://t.co/xYkWA0I1 cc/ @jsuereth @heathercmiller @mccv @dibblego @mergeconflict

Daniel Spiewak @djspiewak

Apparently, I'm a member of the "scalaz faction" of the community. Cool.

Jeremy ☁️ @jeremycloud

@djspiewak you don't like Try? I'd be curious to hear why.

Kenji Yoshida @xuwei_k

scala.util.TryがFunctor則を満たさない http://t.co/2SVWQ2hF と話題になってて、そもそもTryはtwitterから持ってきたものだけれども、そのtwitterの中の人がブログに意見書く https://t.co/TlgPrKWx という展開

Steve J (taking a tweeting break) @stevej

@djspiewak @jeremycloud we didn’t like Either, we felt the name didn’t communicate what we wanted but decided on more than a typedef

Tony Morris @dibblego

@djspiewak Haha, the lengths some people will go to in order to fake learning amazes me. You're in the club dude! The faction!

Steve J (taking a tweeting break) @stevej

@djspiewak @jeremycloud like most code it had humble origins. /cc @marius

Steve J (taking a tweeting break) @stevej

@djspiewak @jeremycloud I don’t know the path Try took to get into stdlib, I haven’t been paying attention since May.

marius eriksen @marius

@djspiewak @stevej @jeremycloud i'd say it's sufficiently useful to promoted to a distinguished construct. it is simply useful ... (cont'd)

marius eriksen @marius

@djspiewak @stevej @jeremycloud its distinguished feature is just the way it composes. it naturally represents a future's state.

marius eriksen @marius

@djspiewak @stevej @jeremycloud (what do you think about scala.concurrent.Future?)

Daniel Spiewak @djspiewak

@marius @stevej @jeremycloud Either composes just fine. Better than Try actually, since it doesn't have the soundness issues.

Daniel Spiewak @djspiewak

@marius @stevej @jeremycloud I haven't looked at it yet, to be honest.

Steve J (taking a tweeting break) @stevej

@djspiewak @marius @jeremycloud what is the best venue to discuss soundness concerns? asking for a friend.

marius eriksen @marius

@djspiewak @stevej @marius @jeremycloud probably the sips mailing list? https://t.co/zM0UpQBc

Jeremy ☁️ @jeremycloud

@djspiewak @stevej @marius the funny thing is that we are using Try right now by tweeting about Try on twitter.