Re: [Chicken-hackers] ugarit and tuples start breaking on 2012/03/03

[Felix]

From: Christian Kellermann <address@hidden>
Subject: Re: [Chicken-hackers] ugarit and tuples start breaking on 2012/03/03
Date: Tue, 6 Mar 2012 12:38:21 +0100

> * felix winkelmann <address@hidden> [120306 12:16]:
>> From: Peter Bex <address@hidden>
>> Subject: Re: [Chicken-hackers] ugarit and tuples start breaking on 2012/03/03
>> Date: Tue, 6 Mar 2012 11:17:11 +0100
>> 
>> > On Tue, Mar 06, 2012 at 08:23:17AM +0100, Felix wrote:
>> >> > Hi,
>> >> > 
>> >> > Something changed in the core that makes ugarit and tuples break on
>> >> > 2012/03/04.  They were not breaking the day before.
>> >> > 
>> >> > You can see the salmonella diff here:
>> >> > http://tests.call-cc.org/master/linux/x86/2012/03/04/yesterday-diff/
>> >> > 
>> >> > The installation error messages for ugarit and tuples are here:
>> >> > http://tests.call-cc.org/master/linux/x86/2012/03/04/salmonella-report/install/tuples.html
>> >> > http://tests.call-cc.org/master/linux/x86/2012/03/04/salmonella-report/install/ugarit.html
>> >> > 
>> >> > The core commits range in which the cause of the problems may be is
>> >> > b7f7e36 - b8363cb.
>> >> 
>> >> Some use of the "assert" macro now has line-number information and tries
>> >> to string-append the "msg" argument to the file/lineno info string.
>> >> Apparently "assert" is called with a non-string "msg" argument, but I
>> >> can't disentangle the complex macros used in "contracts".
>> > 
>> > Possibly the simplest fix is to use "conc" instead?
>> 
>> Well, this simply seems to be an incorrect use of "assert". It should
>> check the second argument for being a string. It happens to work because
>> "error" accepts an optional "location" as initial argument.
> 
> I agree and have attached a patch.

Thanks, signed off, pushed and untested.


cheers,
felix