1 |
-----BEGIN PGP SIGNED MESSAGE----- |
2 |
Hash: SHA256 |
3 |
|
4 |
On 27-06-2011 00:00, Peter Stuge wrote: |
5 |
> Jorge Manuel B. S. Vicetto wrote: |
6 |
>> I feel I should point out that catalyst is Release Engineering's |
7 |
>> team release tool and not a "toy" for people to tinker with. |
8 |
> |
9 |
> Toy? Saying that other people's use of catalyst is only play, while |
10 |
> releng is the only serious user, is really spitting in the face of |
11 |
> everyone who uses catalyst. Maybe not so helpful. |
12 |
> |
13 |
> Meeting you I think you seemed sensible enough that you of course |
14 |
> understand that catalyst is equally much a tool for all it's users. |
15 |
|
16 |
Peter, |
17 |
|
18 |
I didn't mean to imply that anyone else uses catalyst as a toy. What I |
19 |
was saying is that catalyst is a "crucial" tool to releng, whilst to |
20 |
some of the people currently working on it (having commit privileges to |
21 |
the repo) it could be a "toy". |
22 |
I did not and do not in any way want to diminish the importance of the |
23 |
tool to any of its users. At the same time, it should be obvious that |
24 |
releng is a "special" user of the tool. |
25 |
|
26 |
>> I appreciate the interest all of you are showing for the tool and I |
27 |
>> appreciate any improvements, but I and other releng team members need |
28 |
>> this tool to work for us to have releases. |
29 |
> |
30 |
> No problem. Like every other consumer of open source tools you simply |
31 |
> need to pick the version you choose to use carefully, so that it |
32 |
> works for you. This case is not different from any other tool issue. |
33 |
> |
34 |
>>>> For a significant change like this, |
35 |
>>> |
36 |
>>> "significant" is so subjective though. |
37 |
>> |
38 |
>> It did made a significant change to the dependencies of catalyst. |
39 |
> |
40 |
> No, not really. It added one dependency, which is hardly significant. |
41 |
> As has already been shown (by others than William, might I add) |
42 |
> further indirect dependencies are really a bug in the asciidoc |
43 |
> ebuild, and should be fixed there. Since you are all developers |
44 |
> (while I am not) you could actually *already* have eliminated the |
45 |
> point of contention - but noone has bothered and instead you're |
46 |
> writing email complaining about how a little bit of progress is |
47 |
> ruining your workflow. (This is how it looks anyway.) |
48 |
|
49 |
As you know the current dependency pulls more than 10 deps. I don't know |
50 |
if they're accurate or not, but they show up when you try to merge |
51 |
catalyst-9999. If those deps are wrong, we should try to get the team |
52 |
maintaining the asciidoc package fix them. |
53 |
|
54 |
>> I haven't checked it closely yet, |
55 |
> |
56 |
> Maybe that's actually wise, to determine how significant the change |
57 |
> is? |
58 |
> |
59 |
> |
60 |
>> William is not the only one to have concerns about this change. |
61 |
> |
62 |
> He was so far the only one who voice any, and they weren't so nicely |
63 |
> expressed. |
64 |
|
65 |
He wasn't happy to see something pushed through in a so short time span |
66 |
and in a way that seemed to go over others opinions. I also wasn't happy. |
67 |
|
68 |
>> this list probably had little if any releng members. As this is a |
69 |
>> releng tool, |
70 |
> |
71 |
> Either "your" catalyst is an open source project or it is not. If it |
72 |
> is not then you need to hide it away in a secret internal repo so |
73 |
> that noone else in the world can access it. Or you can just do what |
74 |
> the rest of the world does; verify your tools before expecting them |
75 |
> to work. |
76 |
|
77 |
It is an open source project and we want it to be - no argument about |
78 |
that. The commit privileges were restricted to releng or some of its |
79 |
members until a short time ago. |
80 |
As with other open source projects, releng is happy to see the tool |
81 |
improve and is open for new features and requests from the users. |
82 |
However, like any other open source project, there needs to be some |
83 |
consensus. |
84 |
|
85 |
> I understand that you want stable tools, but if you want frozen tools |
86 |
> then you need to do that on your own - because other catalyst users |
87 |
> can and will want to change things. Absolutely not very often, but |
88 |
> apparently often enough that it's a problem for releng to continue |
89 |
> be part of the catalyst community. Or? |
90 |
|
91 |
We're not interested in frozen tools, but we're also not ready to be |
92 |
kept in the sidelines or ignored about catalyst development. Some of the |
93 |
people now working on them are not building or responsible for the |
94 |
building of the official releases, I and a few others are. |
95 |
I have no interest in having catalyst forked, but for that, the |
96 |
developers that got access to catalyst need to realize they need to work |
97 |
with releng and can't ignore it. |
98 |
|
99 |
>> not having us around to "object" doesn't make it ok to commit |
100 |
>> changes without ensuring releng is ok with the changes. |
101 |
> |
102 |
> If so, that in itself is reason for forking, as was discussed. |
103 |
> |
104 |
> I would have zero bad feelings about that, because the wants and |
105 |
> needs simply seem to be different between releng and all other |
106 |
> catalyst users. |
107 |
|
108 |
Sometimes forks can be the best solution, but I really would like to |
109 |
avoid that. I don't see why we shouldn't talk to ensure we reach |
110 |
consensus that work for anyone. At times, a consensus might not be |
111 |
possible, but in those cases the decision will hopefully not cause |
112 |
enough problems or grief as to send people away. If we end up having to |
113 |
make decisions that split the user base of the tool, then people can |
114 |
think about doing a fork. |
115 |
So, I think a fork should be the last option and that we should work |
116 |
hard to reach decisions that everyone can live with. |
117 |
|
118 |
>> This is about making sure that the people interested as well as the |
119 |
>> direct consumers of the tool are ok with any proposed changes. |
120 |
> |
121 |
> You are neglecting every other user than releng. That means me. That |
122 |
> sucks. |
123 |
|
124 |
That is not my purpose. Furthermore, as I've tried to explain above, my |
125 |
previous mail was not about the users of the tool but about the recent |
126 |
people committing to the repo. |
127 |
|
128 |
>> I'm sure no one wants to risk causing a split that could lead to either |
129 |
>> releng assuming control of catalyst again or worse causing a fork in the |
130 |
>> code. |
131 |
> |
132 |
> Actually, forking is indeed the one and only productive step when |
133 |
> different users have different enough requirements and expectations. |
134 |
> |
135 |
> |
136 |
> //Peter |
137 |
> |
138 |
|
139 |
- -- |
140 |
Regards, |
141 |
|
142 |
Jorge Vicetto (jmbsvicetto) - jmbsvicetto at gentoo dot org |
143 |
Gentoo- forums / Userrel / Devrel / KDE / Elections / RelEng |
144 |
-----BEGIN PGP SIGNATURE----- |
145 |
Version: GnuPG v2.0.17 (GNU/Linux) |
146 |
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ |
147 |
|
148 |
iQIcBAEBCAAGBQJOB89zAAoJEC8ZTXQF1qEPA0IQAJssiqfYbS8bE9w3hjpSS3jn |
149 |
HpnUWIoLTkst3VSYO/jGr37mthbx7YbxtXm6V4EL0D4hhNSWfKUybTriTTl29yCY |
150 |
vbDqDrrb8JcJMTx2UAg7q8BjsPpxIo2xhcjQoVX4O1m9zvWeVKgXJIGW/sJ9UKyl |
151 |
ZhcgMGmXGQskraKmTHHxk6aOWgE7iQxmMMZooVm0IYY8TphesMnqncxJveleOPHx |
152 |
0kSTOqsm3QedjHq5MAJcLkPF3cFqJ5f3xfQ7tFWqDBTxDABaHp9Q7FF6J4jCa8p/ |
153 |
DyLW7gYK7NHUXYaemjntumgR/8/l4nGUm4OB+X6nMtSy40jQPBZEghMtu1/Kuy5r |
154 |
9ss82MWh8sn7rW4FusF05JxpSe/du6kh8lrVkCVvJWRnTslWjsmFLOW3RoTnymHI |
155 |
MZr/d+foLNdc9uxOcDiBbjYH0rdOPsqnoygCgMvyOOwKmnqDjnt567vrNh7iRUC/ |
156 |
f/rl5SfEi7cUWHU8OP2GgDsleC22BaNKC8Sb7hvEp4qWiQziBjKaDv/plFEwGWGq |
157 |
6qfbJ4+B3oX4TmCm/Rc2LesByyyk8kyNIVYA3wBRAvajn9qeJ9+Ki/Ezo7Aqcvkm |
158 |
9uzLIxEppegt8NqJzZLmjlCI6gplvRxHfsl10JH/fMZj8tyyP4r8zX/A56fq7763 |
159 |
ou2cZRzmYmLFp8haqTT2 |
160 |
=vgjW |
161 |
-----END PGP SIGNATURE----- |