Hi Christoph --

I think that #errorMessage, #errorClass, and #errorCode might all be useful when deriving a #messageText. Do you still think it is a good idea to drop #errorMessage in favor of #messageText? Well, the current implementation of #messageText reads that #errorCode is a good #messageText. This information would get lost.

In this example, #errorMessage is very different from #errorCode (i.e. #messageText):

Compiler evaluate: '((((('

errorMessage:  'expression expected' 
errorCode:  a Text for '((((("expression expected ->"' 

Best,
Marcel

On 2020-09-14T12:27:25+00:00, christoph.thiede at student.hpi.uni-potsdam.de wrote:

> Hi all,
> 
> 
> please have a look at the attached changeset. It deprecates the IMO redundant instance variable 'errorMessage' in SyntaxErrorNotification and uses super's 'messageText' instead. The two single senders in the Trunk image are patched as well. I wonder at all how the usage in DecompilerTests could have worked before, giving the prior implementation of #messageText in SyntaxErrorNotification ...
> 
> 
> Please review and merge if you like it.
> 
> 
> Best,
> 
> Christoph
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20200914/8578cf83/attachment.html>
> -------------- next part --------------
> An embedded and charset-unspecified text was scrubbed...
> Name: SyntaxErrorNotification errorMessage.1.cs
> URL: <http://lists.squeakfoundation.org/pipermail/squeak-dev/attachments/20200914/8578cf83/attachment.ksh>
> 
>