Sign in Join

Live chat for Free Pascal, Lazarus, Delphi and every Pascal programmer.

Mailing lists / lazarus / 2013-06

Contributions by Cyrax and better SVN commit info

8 messages · 6 people ·

Reply by email Subscribe to this list Your subscription

Read-only archive. Use j and k to move between messages.

Juha Manninen

Looking at issue #24662, Cyrax has studied lot's of code for the patch. Cool! I was not happy with the data structure, I believe it is better now. I removed the generics type TTreeNodeList and replaced it with TList. It's only purpose was to prevent type-casts but they are now needed anyways.

*** Please all developers add the tracker issue number and name of the patch author to a commit message when such info exists! ***

It helps to study and search the commit history by author, not only by committer.
I think there is even an "Author" field in the commit data for this purpose but I don't know (yet) how to set it.

Using gitk I can easily count 16 commits where Cyrax provided a patch. Fix for #24662 makes it 17, but I don't know how many other commits there are without mention of the author.
Cyrax, do you have any estimate yourself of the number of your applied patches?

This is part of project management.
I know there are special tools for mapping issue tracking and source commits together, but our current tools (SVN and Mantis) can do the job just fine if they are used properly.

Regards,
Juha

Reply by email

Vincent Snijders
to Juha Manninen

2013/6/27 Juha Manninen

Show quoted text 10 lines
Looking at issue #24662, Cyrax has studied lot's of code for the patch. Cool! I was not happy with the data structure, I believe it is better now. I removed the generics type TTreeNodeList and replaced it with TList. It's only purpose was to prevent type-casts but they are now needed anyways. *** Please all developers add the tracker issue number and name of the patch author to a commit message when such info exists! ***

AFAIK, svn doesn't make any distinction between committer and author.

Vincent
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20130627/699b16ba/attachment.html>

Reply by email

Mattias Gaertner
to Vincent Snijders

?

Show quoted text 16 lines
Vincent Snijders hat am 27. Juni 2013 um 13:30 geschrieben: ? 2013/6/27 Juha Manninen < <mailto:> > ? ? > > Looking at issue #24662, Cyrax has studied lot's of code for the patch. Cool! >? ? I was not happy with the data structure, I believe it is better now. >? ? I removed the generics type TTreeNodeList and replaced it with TList. >? ? It's only purpose was to prevent type-casts but they are now needed >? ? anyways. > >? ? *** Please all developers add the tracker issue number and name of the >? ? patch author to a commit message when such info exists! ***

Some years ago I did that, but then we had to change the license and there were some strange "copyright" laws in Germany and other countries. Basically, if you add a svn comment that a patch was from a person, you grant that person the right to veto all license issues and the right to withdraw his patch. Some countries have a "fair use" clause that allows to ignore rights for small patches.
Germany does not have a "fair use" clause, in fact, German has no word for "fair".
So it is no problem to add a name to the contributors list, but it can be a problem to name someone for a specific patch.

It's ridiculous. I know.

? ?

? AFAIK, svn doesn't make any distinction between committer and author.

Mattias

?

Reply by email

Martin
to Mattias Gaertner
Mattias Gaertner wrote 30 lines
> Vincent Snijders hat am 27. Juni 2013 um 13:30 > geschrieben: > > > > 2013/6/27 Juha Manninen < > <mailto:> > > > > Looking at issue #24662, Cyrax has studied lot's of code for the patch. > Cool! >> I was not happy with the data structure, I believe it is better now. >> I removed the generics type TTreeNodeList and replaced it with TList. >> It's only purpose was to prevent type-casts but they are now needed >> anyways. >> >> *** Please all developers add the tracker issue number and name of the >> patch author to a commit message when such info exists! *** Some years ago I did that, but then we had to change the license and there were some strange "copyright" laws in Germany and other countries. Basically, if you add a svn comment that a patch was from a person, you grant that person the right to veto all license issues and the right to withdraw his patch. Some countries have a "fair use" clause that allows to ignore rights for small patches. Germany does not have a "fair use" clause, in fact, German has no word for "fair". So it is no problem to add a name to the contributors list, but it can be a problem to name someone for a specific patch.

Are you sure this is based on the name being in the commit message?

I know laws like this exist, but I thought it applies to all contributed code, never mind, if the contribution is documented in some form or not. Besides if patches come via mantis, then there also is documentation.

This is way many projects ask contributes to sign some form of agreement.

Reply by email

Mattias Gaertner
to Martin
Show quoted text 44 lines
Martin hat am 27. Juni 2013 um 14:57 geschrieben: On 27/06/2013 12:52, Mattias Gaertner wrote: > > >> Vincent Snijders hat am 27. Juni 2013 um 13:30 >> geschrieben: >> >> >> >> 2013/6/27 Juha Manninen < >> <mailto:> > >> > > Looking at issue #24662, Cyrax has studied lot's of code for the patch. >> Cool! >>> I was not happy with the data structure, I believe it is better now. >>> I removed the generics type TTreeNodeList and replaced it with TList. >>> It's only purpose was to prevent type-casts but they are now needed >>> anyways. >>> >>> *** Please all developers add the tracker issue number and name of the >>> patch author to a commit message when such info exists! *** > Some years ago I did that, but then we had to change the license and there > were > some strange "copyright" laws in Germany and other countries. > Basically, if you add a svn comment that a patch was from a person, you > grant > that person the right to veto all license issues and the right to withdraw > his > patch. Some countries have a "fair use" clause that allows to ignore rights > for > small patches. > Germany does not have a "fair use" clause, in fact, German has no word for > "fair". > So it is no problem to add a name to the contributors list, but it can be a > problem to name someone for a specific patch. > > Are you sure this is based on the name being in the commit message? I know laws like this exist, but I thought it applies to all contributed code, never mind, if the contribution is documented in some form or not. Besides if patches come via mantis, then there also is documentation.

True. It still makes it simpler to change single files or directories.

This is way many projects ask contributes to sign some form of agreement.

Yes, but when we change the license we have to ask to sign everyone again.

I don't make want to make a big fuss about it. I just pointed out, that the American way to name contributors has a different law effect in other countries. And since there were some ridiculous cases in Germany, I chose the safe way.

I can add the contributors again, if you want.

Mattias

Reply by email

Juha Manninen
to Mattias Gaertner

Amazing. I did not realize there are such issues.
I don't know how important it is to map commits with Mantis reports. It would be nice for tracking reasons for a certain commit and for statistics later.
Should we instead use a separate tool for it? No because nobody would use it. I would prefer the commit message if it is possible for legal reasons.

Juha

Reply by email

Sven Barth
to Juha Manninen
"Juha Manninen" wrote 8 lines
Amazing. I did not realize there are such issues. I don't know how important it is to map commits with Mantis reports. It would be nice for tracking reasons for a certain commit and for statistics later. Should we instead use a separate tool for it? No because nobody would use it. I would prefer the commit message if it is possible for legal reasons.

In FPC we have the possibility to specify the bug report ID using "Mantis #12345" which will then create a svn property containing the ID. Don't know wether this is enabled for Lazarus as well and for details you should ask Florian.

Regards,
Sven
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lazarus.freepascal.org/pipermail/lazarus/attachments/20130627/2cfc5d28/attachment.html>

Reply by email

Hans-Peter Diettrich
to Mattias Gaertner: ? Some years ago I did that, but then we had to change the license and there were some…
Show quoted text 8 lines
>> *** Please all developers add the tracker issue number and name of the >> patch author to a commit message when such info exists! *** Some years ago I did that, but then we had to change the license and there were some strange "copyright" laws in Germany and other countries. Basically, if you add a svn comment that a patch was from a person, you grant that person the right to veto all license issues and the right to withdraw his patch.

I cannot find such legal stuff nor license conditions.

Some countries have a "fair use" clause that allows to ignore rights for small patches.

Patches are not self-contained work, and consequently are not protected as such.

Germany does not have a "fair use" clause, in fact, German has no word for "fair". So it is no problem to add a name to the contributors list, but it can be a problem to name someone for a specific patch.

I see no problems. The only question is about ownership of a software project - is it the first contributor only, or all contributors together? In practice it doesn't make any difference, because free software licenses grant *everybody* the right to modify the software. It's only polite to name the contributors, not an obligation.

DoDi

Reply by email