On Mon, Apr 16, 2012 at 11:12:56AM +0800, Osier Yang wrote:
On 2012年04月16日 10:18, Daniel Veillard wrote:
>On Fri, Apr 13, 2012 at 11:59:32PM +0800, Daniel Veillard wrote:
>>On Fri, Apr 13, 2012 at 09:30:45AM -0600, Eric Blake wrote:
>>>On 04/13/2012 09:16 AM, Daniel Veillard wrote:
>>>>This is based on recent developments on patch checker and the
>>>>goal is to keep a list of pending patches needing review on the
>>>>project web site. The page template in git just hold a pointer
>>>
>>>s/hold/holds/
>>>
>>>>to the web page.
>>>>
>>>>diff --git a/docs/pending.html.in b/docs/pending.html.in
>>>>new file mode 100644
>>>>index 0000000..dfbe647
>>>>--- /dev/null
>>>>+++ b/docs/pending.html.in
>>>>@@ -0,0 +1,13 @@
>>>>+<html>
>>>>+<head>
>>>>+<meta http-equiv="Content-Type" content="text/html;
charset=UTF-8" />
>>>>+<title>Pending patches needing review</title>
>>>>+</head>
>>>>+<body>
>>>>+<h1>Pending patches needing review</h1>
>>>>+<p> A list of pending patches needing review upstream is
available
>>>>+ on<a
href="http://libvirt.org/pending.html">the
project pending
>>>>+ patches page</a>.</p>
>>>
>>>Self-referential, at least when installed on
libvirt.org (but makes
>>>sense when installed locally to end-user machines). And I suppose you
>>>will eventually be changing libvirt.org's page to be generated live
>>>based on your patch checker tool. Works for me.
>>
>> Yup, that's the intent, though I didn't set the crontabs yet :-)
>> But I have put a first page there
http://libvirt.org/pending.html.
>
> Okay, that's live now, it should update 3 times a day, that seems to
>work from now. I could improve the XSLT stylesheet to make things more
>useful, for example sorting by author and then by date rather than
>the lengthy by date list. Opinion welcome, fixing this kind of things
>should be simple. Improving the list to avoid false positive is a bit
>harder but I'm thinking about various small improvements :-)
>
How to tell patchchecher not to list the patches which were not reviewed
in vN (set), but finally got ACK and pushed in later vM (set)? (M > N).
One idea is to mark the vN with specific tag, (especially for the
patch sets which are for same purpose, but have different names
between vN and vM, there is no way to do it totally automatically).
And thus patchchecker could skip it, but it increases the patch authors
load.
patch checker already tries to do this, it detected around 500
obsolete patches which got resent one way or another. Basically if
the author is the same and the patch title is similar, it will assume
the later patch is obsoleting the former one.
And it might be better to send the not reviewed list to libvir list
days before each release. The problem is we don't have a fixed release
date, so it can't be done automatically.
Not so sure I want to push that as email. We know where the list
is, sending a reminder to the URL when entering freeze is IMHO more
useful than copying the list.
Another advise is to indent the list if the patches are in same
thread,
it will be much easier for eyes. :-)
yes presentation should be improved, I did the minimum.
yes that is something I need to add as I stated, when people reply
with patches to be squashed in commit on ACK, patchchecker doesn't
detect it yet, that would be #1 improvement to make right now.
Daniel
--
Daniel Veillard | libxml Gnome XML XSLT toolkit
http://xmlsoft.org/
daniel(a)veillard.com | Rpmfind RPM search engine
http://rpmfind.net/
http://veillard.com/ | virtualization library
http://libvirt.org/