chicken-hackers
[Top][All Lists]
Advanced

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

Re: [Chicken-hackers] Should we make a 4.12 release?


From: Mario Domenech Goulart
Subject: Re: [Chicken-hackers] Should we make a 4.12 release?
Date: Wed, 04 Jan 2017 11:18:07 +0100

Hi again,

On Wed, 04 Jan 2017 11:03:27 +0100 Mario Domenech Goulart <address@hidden> 
wrote:

> On Tue, 03 Jan 2017 23:50:19 +0100 address@hidden wrote:
>
>>> On Tue, 03 Jan 2017 21:35:46 +0100 address@hidden wrote:
>>> 
>>> >> Salmonella relies on -prefix to install eggs into a fresh local egg
>>> >> repository.  That's how it is able to catch dependency problems, for
>>> >> example.
>>> >
>>> > Shouldn't you use a separately built chicken insallation for that,
>>> > with a pristine egg repository for each run?
>>> 
>>> That would work, but it would make the use of salmonella very
>>> inconvenient, as salmonella would depend on a fresh chicken installation
>>> to work.  Not sure if I understand your proposal correctly, though.
>>> 
>>> For example, we have two common scenarios where salmonella is used:
>>> 
>>> * by egg authors, to test eggs:
>>> 
>>>   $ cd egg-source-dir
>>>   $ salmonella
>>> 
>>> * by our test farm, where we build the whole set of eggs for different
>>>   combinations of chicken branches, C compilers, etc.
>>> 
>>> Would it be feasible to use salmonella and C5 in these cases?
>>
>> I think it would. To truly test in a clean environment I would suggest to
>> use a fresh chicken installation for case 1 as well. Note that this can
>> be done by pure copying/hardlinking of files, assuming the user didn't
>> modify installed files. But I didn't know of case 1 and am not sure what
>> advantages it provides.
>
> Case 1 is the recommended check before one submits an egg to the egg
> index.  It detects dependency problems, malformed .meta fields, runs
> tests etc.  Basically, all the checks that are performed daily to all
> eggs, but _before_ eggs are published.  That saves us from reporting
> trivial bugs right after eggs are added to the coop and guarantees that
> eggs can be at least installed when they are available from the official
> egg index.

Regarding detection of dependency problems, I have a feeling that this
is going to be quite handy when porting eggs to C5, as many frequently
used core units in C4 are eggs in C5 (e.g., srfi-1, srfi-13).

All the best.
Mario
-- 
http://parenteses.org/mario



reply via email to

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