pcp
[Top] [All Lists]

Re: Little pmrep regression

To: Nathan Scott <nathans@xxxxxxxxxx>
Subject: Re: Little pmrep regression
From: Marko Myllynen <myllynen@xxxxxxxxxx>
Date: Sun, 13 Dec 2015 17:11:18 +0200
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <198056120.39524113.1449797826555.JavaMail.zimbra@xxxxxxxxxx>
Organization: Red Hat
References: <198056120.39524113.1449797826555.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 2015-12-11 03:37, Nathan Scott wrote:
> 
> +Traceback (most recent call last):
> +  File "/bin/pmrep", line 1320, in <module>
> +    P.connect()
> +  File "/bin/pmrep", line 1313, in connect
> +    self.context = pmapi.pmContext.fromOptions(self.opts, sys.argv, 
> self.opts.pmGetOptionContext())
> +  File "/usr/lib64/python3.3/site-packages/pcp/pmapi.py", line 1086, in 
> fromOptions
> +    if (typed <= 0):
> +TypeError: unorderable types: NoneType() <= int()

right, I didn't see this with Python 2.

> Looks like commit 876645e03864 introduces a possibility for a None
> to be passed into the typed param of pmapi.pmContext.fromOptions()
> ... backing out the part of the change below resolves it for me.

Hmm, yes, that helps there but I just noticed that this venture with
command line parsing has caused a far more serious regression, see my
follow up patch coming shortly.

Thanks,

-- 
Marko Myllynen

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