2.11.8 issues

Nils Kilden-Pedersen

unread,

Mar 9, 2016, 11:20:05 PM3/9/16

to scala-user

1) Future was returned, but the toString looks a little… shall we say… odd?

res6: scala.concurrent.Future[scala.collection.Map[(String, Short),Int]] = List()

scala> res6.toString
res11: String = List(scala.concurrent.impl.CallbackRunnable@3815c525, scala.concurrent.impl.CallbackRunnable@58a887fb)

2) REPL no longer limits the output of a println, causing massive console flooding on large collections. Perhaps this is intentional and can be disabled?

3) REPL may have been improved, but now really sucks in Cygwin. No history with up/down arrows and random cursor moves. Sorry, I hate to be the bearer of bad news, particularly when people have specifically been working to improve the REPL.

Seth Tisue

unread,

Mar 10, 2016, 12:54:52 AM3/10/16

to scala-user

can you open JIRA tickets on all three issues at https://issues.scala-lang.org …? that'd be the best place to discuss the details, separately.

Viktor Klang

unread,

Mar 10, 2016, 1:30:54 AM3/10/16

to Nils Kilden-Pedersen, scala-user

How did you create the Future?

--
Cheers,
√

--
You received this message because you are subscribed to the Google Groups "scala-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scala-user+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Viktor Klang

unread,

Mar 10, 2016, 4:34:51 AM3/10/16

to Nils Kilden-Pedersen, scala-user

Seems caused by this commit: https://github.com/scala/scala/commit/c201eac291682a9bdb9ca2790403084b4f36da76

Fix is to override toString in AbstractPromise to provide a default which does not leak implementation details. And optionally doing a final override of toString in DefaultPromise.

--

Cheers,

√

Jason Zaugg

unread,

Mar 10, 2016, 7:41:06 AM3/10/16

to Viktor Klang, Nils Kilden-Pedersen, scala-user

On Thu, Mar 10, 2016 at 7:34 PM Viktor Klang <viktor...@gmail.com> wrote:

Seems caused by this commit: https://github.com/scala/scala/commit/c201eac291682a9bdb9ca2790403084b4f36da76

Fix is to override toString in AbstractPromise to provide a default which does not leak implementation details. And optionally doing a final override of toString in DefaultPromise.

I think this is the same issue I raised as https://issues.scala-lang.org/browse/SI-9488. I didn't spot that it was a regression in 2.11 at the time, I thought it was only in 2.12.x

-jason

Jason Zaugg

unread,

Mar 10, 2016, 7:46:50 AM3/10/16

to Seth Tisue, scala-user

Regarding output truncation (or lack thereof), can you include a transcript like https://gist.github.com/retronym/572547cb93002192a6cc in your ticket?

-jason

On Thu, Mar 10, 2016 at 3:54 PM Seth Tisue <se...@tisue.net> wrote:

can you open JIRA tickets on all three issues at https://issues.scala-lang.org …? that'd be the best place to discuss the details, separately.

Viktor Klang

unread,

Mar 10, 2016, 7:56:33 AM3/10/16

to Jason Zaugg, Nils Kilden-Pedersen, scala-user

We missed it because it wasn't linked to the changes in 2.12 (I can't recall seeing the commit I linked to earlier in this thread)

--

Cheers,

√

Nils Kilden-Pedersen

unread,

Mar 10, 2016, 8:46:39 AM3/10/16

to scala-user

On Wed, Mar 9, 2016 at 10:19 PM, Nils Kilden-Pedersen <nil...@gmail.com> wrote:

1) Future was returned, but the toString looks a little… shall we say… odd?

res6: scala.concurrent.Future[scala.collection.Map[(String, Short),Int]] = List()

scala> res6.toString
res11: String = List(scala.concurrent.impl.CallbackRunnable@3815c525, scala.concurrent.impl.CallbackRunnable@58a887fb)

As noted, probably the same as:
https://issues.scala-lang.org/browse/SI-9488

2) REPL no longer limits the output of a println, causing massive console flooding on large collections. Perhaps this is intentional and can be disabled?

This may not be a bug. I'm unable to reproduce something right now, that wasn't also the case in 2.11.7

3) REPL may have been improved, but now really sucks in Cygwin. No history with up/down arrows and random cursor moves. Sorry, I hate to be the bearer of bad news, particularly when people have specifically been working to improve the REPL.

https://issues.scala-lang.org/browse/SI-9695

som-snytt

unread,

Mar 11, 2016, 1:59:37 PM3/11/16

to scala-user

Random footnote to the issue: the checkstyle tool has a notion of listing classes known to be immutable, but I recently wondered how that applies to subclasses. That is, a subclass is mutable but immutable with respect to the superclass API. Then what is the status of toString? Is it permitted to vary over time? At stake is idempotency, as opposed to memory model concerns.

Nils Kilden-Pedersen

unread,

Mar 11, 2016, 2:40:35 PM3/11/16

to som-snytt, scala-user

On Fri, Mar 11, 2016 at 12:59 PM, som-snytt <som....@gmail.com> wrote:

Random footnote to the issue: the checkstyle tool has a notion of listing classes known to be immutable, but I recently wondered how that applies to subclasses. That is, a subclass is mutable but immutable with respect to the superclass API. Then what is the status of toString? Is it permitted to vary over time? At stake is idempotency, as opposed to memory model concerns.

Generally, a class can only be assumed immutable if it's final.

On Thursday, March 10, 2016 at 4:41:06 AM UTC-8, Jason Zaugg wrote:
On Thu, Mar 10, 2016 at 7:34 PM Viktor Klang <viktor...@gmail.com> wrote:
Seems caused by this commit: https://github.com/scala/scala/commit/c201eac291682a9bdb9ca2790403084b4f36da76

Fix is to override toString in AbstractPromise to provide a default which does not leak implementation details. And optionally doing a final override of toString in DefaultPromise.

I think this is the same issue I raised as https://issues.scala-lang.org/browse/SI-9488. I didn't spot that it was a regression in 2.11 at the time, I thought it was only in 2.12.x

-jason

--

Simon Ochsenreither

unread,

Mar 13, 2016, 5:48:18 PM3/13/16

to scala-user, som....@gmail.com

Looks like we will need 2.11.9 rather sooner than later. I'd probably implement the fix for https://issues.scala-lang.org/browse/SI-9539 too.
(And fix the rest of these issues with platform-defined encodings in 2.12: https://issues.scala-lang.org/browse/SI-6407)

Seth Tisue

unread,

Mar 14, 2016, 6:36:56 PM3/14/16

to scala-user

nothing's decided yet, but there's certainly a chance we'll end up building 2.11.9 sooner than "late 2016", so if anyone wants fixes in, now would be a good time to be working on them. (or asking, via JIRA comments or what have you, if a fix would have a chance if it were submitted.)

David Barri

unread,

Mar 14, 2016, 9:26:42 PM3/14/16

to scala-user

It'd be great to have https://github.com/scala/scala/pull/4919 backported. I consider exhaustivity analysis bugs critical and would really like not having to wait til late 2016 to trust my code :)

Viktor Klang

unread,

Mar 23, 2016, 4:13:19 PM3/23/16

to Nils Kilden-Pedersen, scala-user

https://github.com/scala/scala/pull/5056

--

You received this message because you are subscribed to the Google Groups "scala-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to scala-user+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Cheers,

√

Reply all

Reply to author

Forward