Re: [scala-ide-user] Scala IDE severely crippled in Eclipse

[Simon Schäfer]

Hi Allan,

sad to hear that you have so many problems with Scala IDE. Unfortunately
most of your points do not share enough information about how to
reproduce the problems, thus we can't fully help you.

On 06/04/2013 07:12 PM, Allan Clarke wrote:
> Greetings.
> The Scala IDE (3.0.0-v-2_10) in Eclipse (Juno, 4.2) using Scala 2.10.1
> is having all kinds of problems. (I do have one Java class and the
> rest are Scala)
Some users have experienced worse performance/usability on Juno than for
Indigo. Thus, if it is possible for you I suggest to move back to Indigo
until Juno is stabilized.

> I'd like to recommend that we start using Scala in production, but
> these are serious roadblocks.
> * The IDE has lost track of whether breakpoints are set and where they are
When does this happen? I experience this sometimes when lines of code
are changed where breakpoints pointed to.
> * The IDE has shown all values in the Variables debug window as null
Again, when does this happen? Is it the case for Scala code and or for
code written in Java?
> * The IDE has repeatedly wanted to import the same file. Importing the
> second time clears the error, then I can delete the redundant import.
Yes, that is a known problem. Unfortunately I don't know of anyone
working on this at the moment.
> * None of the refactoring appears to work (choosing the menu items
> does nothing)
What do you mean with "does nothing"? Is the window not activated, are
no entries shown, do the refactorings produce wrong results?
> * There are very frequent stack overflow errors when the mouse is
> moved over the main menu bar
Haven't seen them yet, maybe that is because you use Juno. But I don't
know that. Can you please report the stacktrace?

Greetings,
Simon
> I can't believe that any of this is as intended. I feel like I am
> being about 40% efficient in my daily work due to these problems.
> Comments?
> Allan
> --
> You received this message because you are subscribed to the Google
> Groups "Scala IDE User" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to scala-ide-use...@googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[Rich Oliver]

On Wednesday, June 5, 2013 1:05:50 AM UTC+1, Simon Schäfer wrote:

Hi Allan,

Some users have experienced worse performance/usability on Juno than for
Indigo. Thus, if it is possible for you I suggest to move back to Indigo
until Juno is stabilized.

I've been using 4.2.2 since it came out. I haven't done a careful comparison with 3.8 or .3.7 but performance is supposed to be considerably better than the earlier versions of 4.2 and that seems to be my experience.

[Allan Clarke]

I may have to report more information as I hit them. Some of them have happened only once or twice in about 40 hours of use - I'll classify these as RARE below. Others happen within an hour of use - I'll call these FREQUENT below.

 

[snip] 

> * The IDE has lost track of whether breakpoints are set and where they are
When does this happen? I experience this sometimes when lines of code
are changed where breakpoints pointed to.

 

RARE - this happened only once and it affected all breakpoints. I do not believe that it was related

to deleting/moving a line that contained a breakpoint.

> * The IDE has shown all values in the Variables debug window as null
Again, when does this happen? Is it the case for Scala code and or for
code written in Java?

 

RARE - this happened twice. The first time ALL of the variable values showed null. The program did not see them as null, just the debugger view. The second time, only a few variables showed null as the value.

 

FREQUENT - Within an hour of use, I have seen the Variables window list variables and have a blank line in the list. When this happens, one or two varaiables at the end of the list are not expandable. It is in this state until I press continue and hit the breakpoint again (this usually corrects things).

> * The IDE has repeatedly wanted to import the same file. Importing the
> second time clears the error, then I can delete the redundant import.
Yes, that is a known problem. Unfortunately I don't know of anyone
working on this at the moment.
> * None of the refactoring appears to work (choosing the menu items
> does nothing)
What do you mean with "does nothing"? Is the window not activated, are
no entries shown, do the refactorings produce wrong results?

 

Let me get back to you on this. Some work some don't - meaning I choose the refactoring menu item and the menu disappears after selection as expected, but then nothing happens. 

> * There are very frequent stack overflow errors when the mouse is
> moved over the main menu bar
Haven't seen them yet, maybe that is because you use Juno. But I don't
know that. Can you please report the stacktrace?

 

 

I'll report the stacktrace on this forum thread.

 

By the way, my setup is a golden opportunity for the Scala IDE team. I can't count how many times I've wanted a context to reproduce flakey issues in the past. I'll be happy to demo problems, etc.

[Allan Clarke]

Info on the overflow:

 

An internal error occurred during: "Compute launch button tooltip".

java.lang.StackOverflowError
 at org.eclipse.jdt.internal.core.JarPackageFragmentRoot.hashCode(JarPackageFragmentRoot.java:212)
 at org.eclipse.jdt.internal.core.PackageFragment.hashCode(PackageFragment.java:384)
 at org.eclipse.jdt.internal.core.ClassFile.hashCode(ClassFile.java:588)
 at org.eclipse.jdt.internal.core.JavaElement.hashCode(JavaElement.java:471)
 at java.util.HashMap.hash(Unknown Source)
 at java.util.HashMap.getEntry(Unknown Source)
 at java.util.HashMap.get(Unknown Source)
 at org.eclipse.jdt.internal.core.JavaModelCache.getInfo(JavaModelCache.java:137)
 at org.eclipse.jdt.internal.core.JavaModelManager.getInfo(JavaModelManager.java:1950)
 at org.eclipse.jdt.internal.core.BinaryType.getElementInfo(BinaryType.java:284)
 at org.eclipse.jdt.internal.core.JavaElement.getElementInfo(JavaElement.java:241)
 at org.eclipse.jdt.internal.core.BinaryType.getDeclaringType(BinaryType.java:238)
 at org.eclipse.jdt.internal.core.ClassFile.getOuterMostEnclosingType(ClassFile.java:688)
 at org.eclipse.jdt.internal.core.ClassFile.openBuffer(ClassFile.java:629)
 at org.eclipse.jdt.internal.core.Openable.getBuffer(Openable.java:288)
 at org.eclipse.jdt.internal.core.ClassFile.getBuffer(ClassFile.java:359)
 at org.eclipse.jdt.internal.core.JavaElement.openWhenClosed(JavaElement.java:540)
 at org.eclipse.jdt.internal.core.BinaryType.getElementInfo(BinaryType.java:286)
 at org.eclipse.jdt.internal.core.JavaElement.getElementInfo(JavaElement.java:241)
 at org.eclipse.jdt.internal.core.BinaryType.getDeclaringType(BinaryType.java:238)
 at org.eclipse.jdt.internal.core.ClassFile.getOuterMostEnclosingType(ClassFile.java:688)
 at org.eclipse.jdt.internal.core.ClassFile.openBuffer(ClassFile.java:629)
 at org.eclipse.jdt.internal.core.Openable.getBuffer(Openable.java:288)

<a lot more like the above>

[Simon Schäfer]

Hi Allan,

what you can do to get more information about errors is to enable full logging. You can set it to DEBUG in Window -> Preferences -> Scala -> Logging and access it in Scala -> Open the Log. The next time when something fails to work (for example the refactoring component that simply doesn't work) please take a look at the log, maybe it gives an information about the error.

For now, especially for your RARE failures, there is too less information to find out what went wrong. If you find out a special case that does not work in your environment it would be nice if you could submit a bug at https://www.assembla.com/spaces/scala-ide/support/tickets

Simon

[Simon Schäfer]

The stacktrace only shows a failure of a component in JDT. Thus, this is probably not the fault of Scala IDE.

--