3.9-7061 showstopper bug - Mantis #5231

Scott Wallace SqueakList at pacbell.net
Wed Oct 18 12:09:14 UTC 2006


Lack of energy, ha!  If there's anyone who has *not* lacked for  
energy on behalf of the community over recent years, it has been you,  
Marcus!

Anyway... in truth I also lacked the energy to understand the bug  
deeply.  I could see where it was happening, and I could see how to  
avoid it for all the existing use cases, and I was unwilling to live  
with the bug, so I bludgeoned it into submission with unspeakable  
workaround code.

Much better than incorporating my workaround code would be for  
someone to make the effort to understand this problem completely and  
to proffer a complete, palatable -- perhaps even elegant -- fix.   
It's just one short method, after all ;=)

For me, the ugly workaround is much more tolerable than living with  
the bug, but there has to be a nicer solution.

Cheers,

   -- Scott


On Oct 15, 2006, at 2:05 AM, Marcus Denker wrote:

>
> On 14.10.2006, at 14:33, Scott Wallace wrote:
>
>> The bug was not only present in 3.8, it was already present in 3.7.
>>
>> It arrived in update "5706KCP170CompilerPrtclRefactor", in a  
>> method timestamped "NS 1/28/2004".  So, amazingly, the community  
>> has been tolerating this showstopping bug in its "stable" releases  
>> for more than 2.5 years.
>>
>> I personally was unwilling to live with the bug for even a day,  
>> once I encountered it;  so for me, as for Tim, this was a  
>> showstopper, no irony intended.   Not only can the .changes file  
>> get quickly filled with arbitrarily large amounts of arbitrary  
>> text, but the unformatted logging of arbitrary text causes havoc  
>> with attempts of the ChangeList browser to parse chunks from  
>> the .changes file.  So to me this was a completely intolerable bug.
>>
>> I quickly tracked down the cause to the "Kernel Cleaning Project,"  
>> in the change-set mentioned above.  I couldn't immediately devise  
>> an obvious, elegant fix, but I did settle on a workaround that did  
>> the job, however unappetizingly.  Since that time there have been  
>> two further advances on the method, one for TinLizzie (for Tweak  
>> compatibility) and the latest for OLPC.
>>
>> I attach the fileout that constitutes the latest OLPC incarnation  
>> of this method, which is more robust than the Squeakland version.   
>> I don't know how this would integrate with 3.9 -- Marcus must have  
>> had good reasons for not porting it 3.9 -- but I offer it FWIW.
>
> No, there was no good reason... just a lack of energy.
>
>     Marcus




More information about the Squeak-dev mailing list