Actually I believe this is not in fact what the 3.10 team has requested, although maybe Edgar or Ralph will correct me. It is my understanding that they instead wish you to add the issue to the list on this page:

http://wiki.squeak.org/squeak/5934

Ken

On Mon, 2007-08-06 at 13:27 +0200, Damien Cassou wrote:

Hi Bryce,

2007/8/6, bryce@kampjes.demon.co.uk bryce@kampjes.demon.co.uk:

...

I've packaged up a minor fix to printing Arrays containing metaclasses. What is the appropriate way to submit this to be included in the next release? There's even a test covering the bug.

Please send a mail to the 3.10 mailing list if you are subscribed. Also, add an item to http://wiki.squeak.org/squeak/5934

Lukas Renggli writes:

...

The bug matters to me because I print arrays into error logs then want to evaluate them back into Squeak to replay the logs to reproduce bugs.

Shouldn't you use #storeString instead of #printString for that matter?

printString does what I want for Arrays with all elements except metaclasses including classes. It's a log file, it's got two uses, one is to read to see what happened the other is to replay to reproduce bugs.

{Array . 42} storeString '((Array new: 2) at: 1 put: Array; at: 2 put: 42; yourself)' {Array . 42} printString '{Array . 42}'

I quickly tried using storeString and it's not very readable for small examples, for the longer ones with 5-10 arrays containing an integer (bytecode address) and a class (the receiving class) it would be hopeless.

Bryce

P.S. I just used printString because it worked, I only tried storeString after reading your email.