summaryrefslogtreecommitdiffstats
path: root/zarb-ml/mageia-webteam/attachments/20110308/fd14a85b/attachment-0001.html
blob: a5338e8aab3d69654d0f3f8be5802b261d3e4a60 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
<br><br><div class="gmail_quote">On 8 March 2011 16:07, Romain d&#39;Alverny <span dir="ltr">&lt;<a href="mailto:rdalverny@gmail.com">rdalverny@gmail.com</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div class="im">On Tue, Mar 8, 2011 at 16:54, nicolas vigier &lt;<a href="mailto:boklm@mars-attacks.org">boklm@mars-attacks.org</a>&gt; wrote:<br>
&gt; On Tue, 08 Mar 2011, Romain d&#39;Alverny wrote:<br>
</div><div class="im">&gt;&gt;  * through the CLI, would it auth against the maintdb app or through<br>
&gt;&gt; the buildsystem that, in turn, would notify the maintdb?<br>
</div>&gt; [...]<br>
<div class="im">&gt; Or the buildsystem can do the auth, and forward the request to maintdb<br>
&gt; private method without doing any check, but including in the request<br>
&gt; the login name of the user requesting the change so that maintdb can<br>
&gt; check the permission. This one is probably better if we want to allow<br>
&gt; changes using both website and CLI as it avoids doing the same checks<br>
&gt; in two different places.<br>
<br>
</div>Looks better that way, yes.<br>
<br>
So maintdb will need to know who is admin (can be checked later with a<br>
flag). Not sure, can be updated later:<br>
<a href="http://mageia.org/wiki/doku.php?id=web:maintdb#break_a_maintainer_package_association" target="_blank">http://mageia.org/wiki/doku.php?id=web:maintdb#break_a_maintainer_package_association</a><br>
<br>
&gt;&gt; <a href="http://mageia.org/wiki/doku.php?id=web:maintdb#specs_actions" target="_blank">http://mageia.org/wiki/doku.php?id=web:maintdb#specs_actions</a> [...]<br>
<div class="im">&gt;<br>
&gt; But I would change this :<br>
&gt;<br>
&gt; * POST <a href="https://maintdb/packages" target="_blank">https://maintdb/packages</a><br>
&gt;  user - string - optional - login of the user pushing the package<br>
&gt;<br>
&gt;  The user string should not be optional but mandatory.<br>
<br>
</div>Right. Fixed.<br>
<div class="im"><br>
&gt; * GET <a href="https://maintdb/packages" target="_blank">https://maintdb/packages</a><br>
&gt;  Returns a list such as:<br>
&gt;  packagename / maintainers login / last commit date / url to bugs report / url to package repo<br>
&gt;<br>
&gt;  I don&#39;t think we should return url to bugs report and url to package<br>
&gt;  repo here.<br>
<br>
</div>Ok. Then we must have clearly spec&#39;d URLs for both so we can use them for sure.<br>
<div class="im"><br>
&gt; For the &quot;last commit date&quot;, we don&#39;t have this info. The<br>
&gt; info we can have is &quot;last sucessful submit date&quot; (the last time the<br>
&gt; private method was called on that package).<br>
<br>
</div>Right, updated as well.<br>
<br>
<br>
Kosmas, would you have time this week to work on maintdb to port these<br>
interfaces to it or do you need help too?<br>
<br>
I can devote some time on it this week, but I&#39;ll have to figure out<br>
how to... well, I&#39;ve got a conflict (RubyGem version error: rack(1.2.1<br>
not ~&gt; 1.1.0)) to make it run.<br></blockquote><div><br> Romain,<br><br> Yes it seems very likely that I would get some time this week, as I&#39;m finishing a project that I had to do today.<br> By all means to help if you want to :-)<br>
 I&#39;ll be available tomorrow if you would like to have a look at the conflict that you have.<br><br> I will also go through the proposed changes and the spec that you have put in the wiki, and ask if I have any questions.<br>
<br> Kosmas<br></div></div>