pcp
[Top] [All Lists]

Re: [pcp] Lots of Python failures on Centos 5.10

To: Ken McDonell <kenj@xxxxxxxxxxxxxxxx>, David Arnold <davida@xxxxxxxxx>
Subject: Re: [pcp] Lots of Python failures on Centos 5.10
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Tue, 16 Sep 2014 04:56:00 -0400 (EDT)
Cc: pcp@xxxxxxxxxxx
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <541206BE.9060503@xxxxxxxxxxxxxxxx>
References: <540CDD93.1040606@xxxxxxxxxxxxxxxx> <54114E71.50900@xxxxxxxxxxxxxxxx> <E7E51348-83C3-4C07-92CE-0198FCD69D3A@xxxxxxxxx> <541206BE.9060503@xxxxxxxxxxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: /ATWV4G8jvhRuDjZqyVIO2662Yvs5g==
Thread-topic: Lots of Python failures on Centos 5.10
Hi,

----- Original Message -----
> On 11/09/14 21:39, David Arnold wrote:
> > ...
> > [lots of helpful info omitted]
> 
> Thanks Dave.
> 
> This is exactly the sort of insight I was seeking and did not have.
> 
> My _everywhere_ assertion may have been a little premature.  There are 23
> uses in the QA test program I was looking at, and I extrapolated from that.
> 
> In fact there are only 10 files involved, so conditional post-processing in
> the build seems entirely feasible.
> 
> I'll give it a try.

Sorry 'bout this guys, this is all fallout from the python3 porting efforts -
I didn't realise 2.4 would have these issues (will introduce a RHEL5 box into
my QA for next release & beyond).

I'll take a closer look at your latest Ken, and see where we are at with v3.
As Dave mentioned, python3 is especially strict on the new exception syntax
- but the issues run deeper than that, from perusing some of those errors.

So, we may have to go with the "RHEL5 python is too old" option (as we did for
RHEL5 Qt already)... not sure yet.  Thanks for hacking on this in my absence!

cheers.

--
Nathan

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