> I don't know any other way to make this test not hang as long as #caseError is broken.

Well, hypothetically, you could check thisContext stack :P

> General errors should be general and terse, not including the #printString's of, e.g., a 1000-element Array.

Absolutely.

I will try to merge this into ObjectTest as soon as I have an okay for the linked Merge Request :-)

Best,
Christoph
________________________________
Von: Chris Muller <ma.chris.m at gmail.com>
Gesendet: Dienstag, 17. März 2020 01:04:52
An: Thiede, Christoph
Cc: The general-purpose Squeak developers list; packages at lists.squeakfoundation.org
Betreff: Re: [squeak-dev] The Trunk: Tests-cmm.428.mcz


> simply to expose the bug

Wait, now I see the actual bug you are talking about. It took me some time to find this out, and when this is fixed, no one (except you :)) will be able to understand what this test should do. We should definitively add some more explaining comments to this.

I thought about that but then decided to simply leave the test generic, that #caseError is just working and not failing for ANY reason, not just the bug today.  As you said, once it's fixed, any explanations here become ancient history that's of questionable usefulness to future readers.

Also, an ideal test should never hang up/time out but fail instead, should it? :)

Once it's fixed, it won't.   :)   Timeout was introduced because tests were hanging the CI server.  I don't know any other way to make this test not hang as long as #caseError is broken.


In this case, I see we need an extra test class for it. (Well, personally I would've used #newSubclass, but I see this is a matter of taste ^^). The "regular" #caseOf: tests should go to ObjectTest.

Sure.  Please feel free to move it to ObjectTest, include your tests, and improve it if you wish.  I only care about getting #caseError reverted.  General errors should be general and terse, not including the #printString's of, e.g., a 1000-element Array.

Best,
  Chris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20200317/99bef882/attachment.html>