pcp
[Top] [All Lists]

Re: [pcp] qa/1068 and 1069 failing

To: Nathan Scott <nathans@xxxxxxxxxx>, Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] qa/1068 and 1069 failing
From: Marko Myllynen <myllynen@xxxxxxxxxx>
Date: Mon, 4 Jan 2016 12:56:50 +0200
Cc: pcp@xxxxxxxxxxx
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <1364880211.3788367.1451892440721.JavaMail.zimbra@xxxxxxxxxx>
Organization: Red Hat
References: <5674A9A1.3000200@xxxxxxxxxxxxxxxx> <56828719.5060309@xxxxxxxxxx> <1364880211.3788367.1451892440721.JavaMail.zimbra@xxxxxxxxxx>
Reply-to: myllynen@xxxxxxxxxx
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.4.0
Hi,

On 2016-01-04 09:27, Nathan Scott wrote:
> ----- Original Message -----
>> On 2015-12-19 02:49, Ken McDonell wrote:
>>> This is on grundy (PCP 3.11.0 ia64 SUSE SLES11 SP1) and may be related
>>> to an old version of Python that we need to guard against (either in the
>>> qa test, or better in the PCP code at run time).
>>>
>>> kenj@grundy:~> pcp python -V
>>> Python 2.6
>>>
>>> This is outside my sphere of competence, so I need some assistance from
>>> those who know.
>>
>> OrderedDict is a hard requirement for pmrep but the surprising part is
>> that it is available on RHEL 6 which has python-2.6.6-64.el6.x86_64 but
>> upstream says it was introduced in 2.7:
>>
>> https://docs.python.org/2/library/collections.html
>>
>> I guess with older Python versions a user (not QA) could use
>> https://pypi.python.org/pypi/ordereddict.
> 
> Sounds like we need a configure.ac update here then, and conditional pmrep
> build?  Or can we re-engineer this not to make use of OrderedDict (as David
> did over in commit dc60f3389b1b86606b39a92d11164900e507cebc IIRC).

I think we should wait a bit with that to see will it even be needed as
if we have pmfg / enhanced pmcc / or such available a bit later we could
instead migrate using it and drop/change this code then.

Thanks,

-- 
Marko Myllynen

<Prev in Thread] Current Thread [Next in Thread>