Bug closed for "not replied" without asking a question?

Feedback Case Number: feedback://showreport?report_id=62398

It just got closed because I apparently have not replied to it. Oddly, all I remember is that Robin added a comment about having made a similar project. There was no question in that comment that I was asked to reply to.

Now, it got closed because I didn’t provide a reply to … huh?!?

I mean, WTF? I provide a demo that demonstrates compiler bugs. Someone changes my demo and says: Look, no more errors. Do I really have to explain that changing the demo to make it compile doesn’t invalidate the bug report?

I think Robin was showing you a workaround , but doing it that way, causes the case to AUTOMATICALLY close if you don’t respond in 10 days… I suspect he did not mean that to happen.

I think feedback needs a mechanism (if one does not already exist) so that Xojo personnel could mark a response like that as comment rather than a question that requires a response before they post it.

Just request that the case be re-opened.

-Karen

In one of my Feedback cases I didn’t get an email after the case was set to Information Required. It wasn’t possible to reproduce the behaviour.

1 Like

I’ve got that too, sometimes you get an email, and sometimes not. and you’re surprised to get it closed because of no response…

1 Like

I think @Robin_Lauryssen-Mitchell picked the wrong type of answer and this automatic closures without further emails is really annoying.

2 Likes

The errors are easy to reproduce with the given sample:

  1. The compiler should pick the one closer to current place, e.g. the one in Class1.
  2. Someone could debug the compiler to see why it stumbles on self here.
  3. The compiler should pick AConstant item within same class first.

Right. Thanks.

Also, the correct term is “scope”. Scopes are layers, as seen from the innermost point where the compiler is currently translated, e.g. inside a line of code.

Any decent programming language follows the rule of picking by searching for the identifier name from closest to farthest scope. Even the older compiler (in Real Studio 2012) got that right. A later compiler broke this, and I don’t even understand why. It makes no sense to break a commonly accepted concept like this.

I requested reopening.