72 | | This whole review process is to give and receive feedback, where we as a community can help making OpenVPN better - together. For example, the more experienced developers can share their advices to those who are less experienced. Those being less experienced can share their understanding of the code as well, which might shed some light to issues or solutions which are better. Non-developers can also participate by sharing their expertise on OpenVPN in general by giving feature ACKs (NACKs) when applicable. This way everyone can learn something and hopefully even get wiser. |
73 | | |
74 | | To explain the Signed-off-by and Acked-by process a bit further. In git you have to "data fields" which are populated automatically when commits are done, author and commiter. The commiter will be the one who applies patches from the mailing list or the one doing commits to a tree which are fetched via the remote access features in git. |
75 | | |
76 | | Then you have the author field which is either present in the patch file itself. This field might be missing if ''git format-patch'' has not been used to create the patch file sent to the mailing list. In these cases, it is expected that the author of the patch is the same as the sender - unless the commit message in the patch indicates something else. |
77 | | |
78 | | If the author information is present, the sender field will be treated differently as wel, in those cases where someone sends a patch on behalf of somebody else. |
| 72 | This whole review process is to give and receive feedback, where we as a community can help making OpenVPN better - together. For example, the more experienced developers can share their advice to those who are less experienced. Those being less experienced can share their understanding of the code as well, which might shed some light to issues or solutions which are better. Non-developers can also participate by sharing their expertise on OpenVPN in general by giving feature ACKs (NACKs) when applicable. This way everyone can learn something and hopefully even get wiser. |
| 73 | |
| 74 | To explain the Signed-off-by and Acked-by process a bit further. In git you have two "data fields" which are populated automatically when commits are done, author and committer. The committer will be the one who applies patches from the mailing list or the one doing commits to a tree which are fetched via the remote access features in git. |
| 75 | |
| 76 | Then you have the author field which is usually present in the patch file itself. This field might be missing if ''git format-patch'' has not been used to create the patch file sent to the mailing list. In these cases, it is expected that the author of the patch is the same as the sender - unless the commit message in the patch indicates something else. |
| 77 | |
| 78 | If the author information is present, the sender field will be treated differently as well, in those cases where someone sends a patch on behalf of somebody else. |
85 | | Then a patch author sends his patch to somewhere, he should make sure it contains a 'Signed-off-by: {Full name} <email@example.com>' line in the commit text. This is to officially announce that declare that this patch is ready for further processing. When doing git commits, you can add this indication automatically by doing 'git commit -s' (or --signed-off). |
86 | | |
87 | | The one who receives the patch and finds it good to go even further (if want you do some kind of "internal" review before sending it further to the public) will add his/hers "Signed-off-by" reference as well when sending further again. This way we can track who have been looking at the patch. And the more people who have eye-balled the patch and adds their "Signed-off" line to the patch, the better! By doing so, the reviewer indicates that the patch is good. |
| 85 | When a patch author sends his patch to somewhere, he should make sure it contains a `Signed-off-by: {Full name} <email@example.com>` line in the commit text. This is to officially announce that declare that this patch is ready for further processing. When doing git commits, you can add this indication automatically by doing ''git commit -s'' (or ''--signed-off''). |
| 86 | |
| 87 | The one who receives the patch and finds it good to go even further (if you want to do some kind of "internal" review before sending it further to the public) will add his/hers ''Signed-off-by'' reference as well when sending further again. This way we can track who has been looking at the patch. And the more people who have eye-balled the patch and added their ''Signed-off'' line to the patch, the better! By doing so, the reviewer indicates that the patch is good. |
144 | | The Acked-by: can be the one who does the final commit to the source code repository, but it can also be someone who don't do the final commit. The last Signed-off-by should normally be the person the one who does the final commit. This is usually the person who "touched" the patch by adding the Acked-by line(s). |
145 | | |
146 | | However, you should not see any commits which is written by (author) and Acked-by by the same person. Then the process have failed to give a qualified review. |
147 | | |
148 | | You may ask why we have this "bureaucracy" (which is a fair question!). It is simply to keep all who send in patches, those who reviews them and finally accepts them into the source code repository more accountable for their work. This process will document each accepted patch and give an indication that it has been through a certain set of reviews. This is to ensure that we don't accept bad code easily into the source code repository. This is just to keep this project as transparent and open as possible. |
149 | | |
150 | | For now, the only patches which which "breaks" this process in the git tree are the patches coming in from James' SVN BETA21 branch. But we trust that James tracks the patches he pulls in according to the standards he needs to feel comfortable with the OpenVPN code. |
| 144 | The Acked-by: can be the one who does the final commit to the source code repository, but it can also be someone who doesn't do the final commit. The last Signed-off-by should normally be the person who does the final commit. This is usually the person who "touched" the patch by adding the Acked-by line(s). |
| 145 | |
| 146 | However, you should not see any commits which are written by (author) and Acked-by by the same person. Then the process has failed to give a qualified review. |
| 147 | |
| 148 | You may ask why we have this "bureaucracy" (which is a fair question!). It is simply to keep all who send in patches, those who review them and finally accept them into the source code repository more accountable for their work. This process will document each accepted patch and give an indication that it has been through a certain set of reviews. This is to ensure that we don't accept bad code easily into the source code repository. This is just to keep this project as transparent and open as possible. |