1 |
|
<!doctype html public "-//w3c//dtd html 4.0 transitional//en"> |
2 |
<html> |
<html> |
3 |
<head> |
<head> |
4 |
|
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> |
5 |
|
<meta name="GENERATOR" content="Mozilla/4.76 [en] (X11; U; Linux 2.4.1 i686) [Netscape]"> |
6 |
|
<meta name="Author" content="Chris Hill"> |
7 |
|
<title>MITgcm CVS policy</title> |
8 |
</head> |
</head> |
9 |
<body> |
<body text="#000000" bgcolor="#FF99FF" link="#0000EF" vlink="#51188E" alink="#FF0000"> |
10 |
MITGCM CVS policies |
|
11 |
=================== |
<center> |
12 |
|
<h1> |
13 |
o Introduction |
MITgcm CVS policy</h1></center> |
14 |
|
|
15 |
This note describes policies that apply to the MITGCM CVS repository |
<h2> |
16 |
|
Introduction</h2> |
17 |
o Why have a policy? |
This note describes policies that apply to the MITGCM CVS repository. |
18 |
|
<h2> |
19 |
CVS itself is a liberal free-for-all product that can be used in a variety |
Why have a policy?</h2> |
20 |
of ways. It is designed to provide a system for storing arbitrary files |
CVS itself is a liberal free-for-all product that can be used in a variety |
21 |
in a way that allows the change history of the individual files to be |
of ways. It is designed to provide a system for storing arbitrary files |
22 |
tracked. If CVS is used without any other policy the result can be a |
in a way that allows the change history of the individual files to be tracked. |
23 |
collection of files each of which has complex, multiply branched set of |
If CVS is used without any other policy the result can be a collection |
24 |
interelated versions. This sort of CVS repository can be come like a |
of files each of which has complex, multiply branched set of inter-related |
25 |
library where books are simply stored in a huge heap. Although nothing is |
versions. This sort of CVS repository can be come like a library where |
26 |
actually lost, the task of finding a coherent collection of material soon |
books are simply stored in a huge heap. Although nothing is actually lost, |
27 |
becomes impossible. |
the task of finding a coherent collection of material soon becomes impossible. |
28 |
|
<p>The policies we employ address tree areas |
29 |
The policies we employ address two areas |
<ol> |
30 |
1. Maintaining an orderly and easily identifiable, coherent set of |
<li> |
31 |
evolving "products". |
Maintaining an orderly and easily identifiable, coherent set of evolving |
32 |
2. Allowing concurrent, on-going development of products. |
"products".</li> |
33 |
|
|
34 |
o Development trees and checkpoint trees |
<li> |
35 |
|
Allowing concurrent, on-going development of product components.</li> |
36 |
A directory within the MITGCM repository resides under either the |
|
37 |
development branch or the checkpoint branch. Files within each branch |
<li> |
38 |
follow different policies. |
Making the integration of achieved developments easy, rapid, organized |
39 |
|
and clear.</li> |
40 |
o Development tree policies |
</ol> |
41 |
|
|
42 |
Development trees are intended to be flexible areas where arbitrary files |
<h2> |
43 |
can be stored with multiple versions, many branches supporting multiple |
Development trees and checkpoint trees</h2> |
44 |
ongoing streams of development. Development trees have no policies in |
A directory within the MITGCM repository resides under either the development |
45 |
place to control complexity. Development trees might be associated with |
branch or the checkpoint branch. Files within each branch follow different |
46 |
a particular person, a certain project or a particular special piece of |
policies. |
47 |
work. These trees are intended to be useful areas for storing current |
<h2> |
48 |
work and for archiving partially finished work so that it doesn't get |
Development tree policies</h2> |
49 |
mislaid and s that some record of the development history can be easily |
Development trees are intended to be flexible areas where arbitrary files |
50 |
maintained. The only policy that applies to development trees is that |
can be stored with multiple versions, many branches supporting multiple |
51 |
this style of tree is not intended to be used for providing a |
ongoing streams of development. Development trees have no policies in place |
52 |
"checkpoint" distribution. Tagged configurations of tools built from this |
to control complexity. Development trees might be associated with a particular |
53 |
style of tree can be distributed, but because these trees do not have any |
person, a certain project or a particular special piece of work. These |
54 |
polcies regarding testing of functionality, platform coverage or |
trees are intended to be useful areas for storing current work and for |
55 |
documentation these trees are not allowed to form the basis of |
archiving partially finished work so that it doesn't get mislaid and so |
56 |
"checkpoint" distrbutions or formal model releases. Other policies can |
that some record of the development history can be easily maintained. The |
57 |
be defined by individuals users of these trees but there are no further |
only policy that applies to development trees is that this style of tree |
58 |
global policies. The MITGCM repository development/ subdirectory is |
is not intended to be used for providing a "checkpoint" distribution. Tagged |
59 |
reserved for holding development trees. Development trees also serve as |
configurations of tools built from this style of tree can be distributed, |
60 |
experimental areas for exploring new code management policies. |
but because these trees do not have any policies regarding testing of functionality, |
61 |
|
platform coverage or documentation these trees are not allowed to form |
62 |
o Checkpoint tree policies |
the basis of "checkpoint" distributions or formal "releases". Other policies |
63 |
|
can be defined by individuals users of these trees but there are no further |
64 |
Checkpoint trees are intended to provide structured storage areas for |
global policies. The MITGCM repository development_tree/ sub-directory |
65 |
holding code that is intended for open distribution and is to be readily |
is reserved for holding development trees. Development trees also serve |
66 |
downloaded. There are policies governing the operation of these trees |
as experimental areas for exploring new code management policies. |
67 |
which are designed to ensure that distributed codes are clearly |
<h2> |
68 |
identified and meet certain levels of quality. |
Checkpoint tree policies</h2> |
69 |
|
Checkpoint trees are intended to provide structured storage areas for holding |
70 |
1. Check-out |
code that is intended for open distribution and is to be readily downloaded. |
71 |
|
There are policies governing the operation of these trees which are designed |
72 |
Just do it! Two mechanisms are available. cvsanon for read only access |
to ensure that distributed codes are early identified and meet certain |
73 |
and regular cvs co .... for read/write access. |
levels of quality. |
74 |
|
<ol> |
75 |
2. Check-in. |
<li> |
76 |
|
Check-out</li> |
77 |
The code check in procedure for a "checkpoint" tree is as follows |
|
78 |
2.1 Check out the latest main branch revision. |
<br>Just do it! Two mechanisms are available. cvsanon for read only access |
79 |
2.2 Merge your changes into that revision. |
and regular cvs co .... for read/write access. |
80 |
2.3 Build and validate new code. |
<li> |
81 |
2.4 Check that there have been no further changes to the |
Check-in</li> |
82 |
repository. Repeat from 2.1 if repository has changed. |
|
83 |
2.5 Get clearance from other developers to check in your changes. |
<br>The code check in procedure for a "checkpoint" tree is as follows |
84 |
2.6 Check in your changed main branch. |
<ol> |
85 |
2.8 Build and validate the new changes. |
<li> |
86 |
2.9 Tag code as "checkpointNN". Add records to docs/tag-index. |
Check out the latest main branch revision.</li> |
87 |
2.10 Build and validate test cases (see testing). |
|
88 |
2.11 Create and install checkpointNN.tar.gz |
<li> |
89 |
|
Merge your changes into that revision.</li> |
90 |
3. Testing |
|
91 |
|
<li> |
92 |
Things in a checkpoint tree require a test case that |
Build and validate new code.</li> |
93 |
can be used to validate the component. |
|
94 |
|
<li> |
95 |
4. Checkpoint tagging |
Check that there have been no further changes to the repository. Repeat |
96 |
|
from 2.1 if repository has changed.</li> |
97 |
No code should be left in limbo. Checking in code and then |
|
98 |
leaving it in the repository untagged is bad. When you check |
<li> |
99 |
in code you are creating a new checkpoint. That means you don't |
Get clearance from other developers to check in your changes.</li> |
100 |
check in some code which you "know" works 100% and then go away |
|
101 |
for two weeks. When you start checking in code you make sure |
<li> |
102 |
you have time to do the process end-to-end as described in section |
Check in your changed main branch.</li> |
103 |
2. |
|
104 |
|
<li> |
105 |
5. Release tagging |
Build and validate the new changes.</li> |
106 |
|
|
107 |
Releases are only based on checkpoint tree code. Maintenance fixes |
<li> |
108 |
to releases are also maintained within the checkpoint tree. Files |
Tag code as "checkpointNN". Add records to docs/tag-index.</li> |
109 |
within a release must have accompanying documentation. The form of this |
|
110 |
documentation depends on the file type. |
<li> |
111 |
|
Build and validate test cases (see testing).</li> |
112 |
6. Branches |
|
113 |
|
<li> |
114 |
Branches are to be used for bug-fixes and code patches to releases |
Create and install checkpointNN.tar.gz</li> |
115 |
only. All other changes e.g. totally new features, bug-fixes to |
</ol> |
116 |
checkpoints are introduced by moving checkpoint levels forward. The |
|
117 |
only historical code maintenance that is employed is for fixes and |
<li> |
118 |
patches to formal releases - not checkpoints. |
Testing</li> |
119 |
|
|
120 |
o These policies are causing me a big problem, what can I do? |
<br>Things in a checkpoint tree require a test case that can be used to |
121 |
|
validate the component. |
122 |
The policies are not enforced by any mechanism other than mutual |
<li> |
123 |
agreement! If you think the policies are not appropriate then let us know |
Checkpoint tagging</li> |
124 |
and we can discuss changing them. However, if you simply ignore the |
|
125 |
policies regarding the checkpoint_release trees then your code may be |
<br>No code should be left in limbo (un-tagged) for extended periods. On |
126 |
removed and/or your access revoked. |
the other hand it's unnecessary to create a checkpoint tag for every little |
127 |
|
change. Checkpoint tags should be made after a particularly significant |
128 |
o What about bitkeeper |
code modification or otherwise on a regular basis, say bi-weekly. Very |
129 |
|
often we set a list of goals to be reached by the next checkpoint which |
130 |
We are looking at bitkeeper (www.bitkeeper.com). It looks cool, but |
sometimes takes more than two weeks to achieve. Obviously, in this case |
131 |
policies are still important. Any experience, suggestions let us know. |
a bi-weekly checkpoint would not be useful. |
132 |
Watch this space! |
<li> |
133 |
|
Release tagging</li> |
134 |
|
|
135 |
|
<br>Releases are only based on checkpoint tree code. Maintenance fixes |
136 |
|
to releases are also maintained within the checkpoint tree. Files within |
137 |
|
a release must have accompanying documentation. The form of this documentation |
138 |
|
depends on the file type. |
139 |
|
<li> |
140 |
|
Branches</li> |
141 |
|
|
142 |
|
<br>Branches are a useful tool for making changes prior to checkpoints |
143 |
|
without breaking other working versions but it must be understood that |
144 |
|
branches are short-lived and that releases and checkpoints not be made |
145 |
|
from a branch. Branches are especially useful for adding totally |
146 |
|
<br>new features. bug-fixes to checkpoints are introduced by moving checkpoint |
147 |
|
levels forward. The only historical code maintenance that s employed is |
148 |
|
for fixes and patches to formal releases - not checkpoints.</ol> |
149 |
|
|
150 |
|
<h2> |
151 |
|
Someone checked-in broken code so not my code doesn't work?</h2> |
152 |
|
You have several options: |
153 |
|
<ol> |
154 |
|
<li> |
155 |
|
Politely email everyone at support@mitgcm.org asking what has happened |
156 |
|
and that it be fixed?</li> |
157 |
|
|
158 |
|
<li> |
159 |
|
Figure out why the new code is broken, fix it, check it in and proudly |
160 |
|
send a message to support@mitgcm.org to show how constructive you are.</li> |
161 |
|
|
162 |
|
<li> |
163 |
|
Complain that the quality of work is too low and then do nothing to fix |
164 |
|
the code.<br> |
165 |
|
<BR></li> |
166 |
|
|
167 |
|
<br>We advise you to only use the third option if you are confident that |
168 |
|
your own contributions to the code are bug-free, well written, documented |
169 |
|
and fool proof. :)</ol> |
170 |
|
|
171 |
|
<h2> |
172 |
|
These policies are causing me a big problem, what can I do?</h2> |
173 |
|
The policies are not enforced by any mechanism other than mutual agreement! |
174 |
|
If you think the policies are not appropriate then let us know and we can |
175 |
|
discuss changing them. However, if you simply ignore the policies regarding |
176 |
|
the checkpoint_release trees then your code may be removed and/or your |
177 |
|
access revoked. |
178 |
|
<h2> |
179 |
|
What about bitkeeper</h2> |
180 |
|
We are looking at bitkeeper (www.bitkeeper.com). It looks cool, but policies |
181 |
|
are still important. Any experience, suggestions let us know. Watch this |
182 |
|
space! |
183 |
|
<p>Questions, comments e-mail: code.czars@mitgcm.org |
184 |
|
<br> |
185 |
|
<hr WIDTH="100%"> |
186 |
|
<table CELLSPACING=0 CELLPADDING=0 WIDTH="100%" NOSAVE > |
187 |
|
<tr NOSAVE> |
188 |
|
<td><font size=-1>Last modified on $Date$</font></td> |
189 |
|
|
190 |
|
<td> |
191 |
|
<div align=right><font size=-1>CVS: /u/gcmpack/mitgcm.org/../cvspolicy.html,v |
192 |
|
$Revision$</font></div> |
193 |
|
</td> |
194 |
|
</tr> |
195 |
|
</table> |
196 |
|
|
|
Questions, comments e-mail: code.czars@mitgcm.org |
|
197 |
</body> |
</body> |
198 |
</html> |
</html> |