chicken-hackers
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Chicken-hackers] gcc compile farm


From: Mario Domenech Goulart
Subject: Re: [Chicken-hackers] gcc compile farm
Date: Wed, 10 Aug 2011 15:17:10 -0400
User-agent: Gnus/5.13 (Gnus v5.13) Emacs/23.1.91 (gnu/linux)

Hi Felix,

On Wed, 10 Aug 2011 01:58:20 -0400 (EDT) Felix <address@hidden> wrote:

> From: Mario Domenech Goulart <address@hidden>
> Subject: Re: [Chicken-hackers] gcc compile farm
> Date: Tue, 09 Aug 2011 09:37:53 -0400
>
>> On Fri, 05 Aug 2011 08:54:26 +0200 (CEST) Felix <address@hidden> wrote:
>> 
>>>>> I registered at the GCC Compile Farm:
>>>>>
>>>>>   http://gcc.gnu.org/wiki/CompileFarm
>>>>>
>>>>> Contact me if you're interested in getting chicken-stuff tested on one
>>>>> of their machines.
>>>> 
>>>> That's very interesting.  Would running salmonella on that farm be
>>>> possible?  I'm asking because salmonella is pretty heavy weight in terms
>>>> or processor/memory usage.
>>>
>>> I think running it regularly will probably become annoying for other
>>> users.
>> 
>> Ok.
>
> Well, that doesn't mean one could give it a try. I don't really now
> whether it may be a problem or not.

Alright.  Maybe we can try one of the lightweight approaches.  If it
disturbs the other users, we stop, of course.


>>> That's a nice option. Perhaps it could also allow building only *changed*
>>> eggs since a given timestamp?
>> 
>> That's a good approach for testing eggs, indeed.  OTOH, by skipping
>> unchanged eggs, we could miss some opportunities to test the compiler.
>
> Right. Perhaps it's not such a great idea.

Actually it is quite good for testing eggs.  It would be very
lightweight considering the current frequency of egg code changes.

Another interesting aspect is that it would be able to catch bugs
related to dependencies on specific egg versions, since the local egg
repository won't be set empty.  For example, consider an egg A which
depends on an egg B whose version should be greater than 2.  If A's meta
file doesn't specify the minimum requirement for B's version, that's a
bug.

If we don't set the local egg repository empty and B version 1 is
installed, when A is tested, there is some chance it'll raise an error
because it won't find a feature provided by the installed B version
(which older than the required, but not specified, version).

That's something that the current salmonella is not able to test.

Best wishes.
Mario
-- 
http://parenteses.org/mario



reply via email to

[Prev in Thread] Current Thread [Next in Thread]