pcp
[Top] [All Lists]

Re: pcp updates: python fix, build, qa, and docs tweaks (+lberk+mgoodwin

To: "Frank Ch. Eigler" <fche@xxxxxxxxxx>
Subject: Re: pcp updates: python fix, build, qa, and docs tweaks (+lberk+mgoodwin+minnus merges)
From: Nathan Scott <nathans@xxxxxxxxxx>
Date: Wed, 16 Sep 2015 20:51:07 -0400 (EDT)
Cc: PCP Mailing List <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <y0mlhc61j3o.fsf@xxxxxxxx>
References: <1694554334.33104859.1442384154493.JavaMail.zimbra@xxxxxxxxxx> <1055972703.33105630.1442384264176.JavaMail.zimbra@xxxxxxxxxx> <y0mlhc61j3o.fsf@xxxxxxxx>
Reply-to: Nathan Scott <nathans@xxxxxxxxxx>
Thread-index: w6ZzUDM04H8tUzBR38P76wVg8fvIFA==
Thread-topic: pcp updates: python fix, build, qa, and docs tweaks (+lberk+mgoodwin+minnus merges)

----- Original Message -----
> [...]
> The tarball you elected represents close to a git checkout of the

By "you elected" I guess you mean the comment in the rpm specfile?
That's a link Martin sent me several months back - possibly needs
an update, will follow up on that.

> They do not represent ideally packaged-for-deployment files.

Martin releases both forms, specfile comment needs updating methinks.

> Packaging a small & quick-to-load deployable vector was one of the
> little contributions of the pcp-webjs copy of vector.

Martin releases both forms, & we should be encouraging Vector to be
independently successful, not trying to "add value" while copying it
into pcp-webjs.

I remain convinced that your second webjs copy of Vector adds no value
- and you've shown now via https://github.com/Netflix/vector/issues/83
(which shoulda been found >2 months ago) that this double handling can
be actively harmful.

cheers.

--
Nathan

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