Re: [ros-dev] Gudie Lines for SVN Write access

[Magnus Olsen]

My point is we starting getting alot of people, and yes the currenct rules works but not
for long. 
we are now lest 35 people with commit access.  And we need a system how u getting svn 
access it must be bit harder getting svn write access. 

1. I know we got bugzlla, I wrote Bugzila or mailing list  

2. A patch need always be tested and exaim before it been accpect
I can not count how many time this year I got bad patcher from people
and I need test the code, I got some patcher that doing BSOD in some
case. Yes U need always test and examin the patcher from people that
does not have svn write access. 

3. Not with the time. we are goving alot. We need start thing about futer 
so we do not land up with example over 100 commicters. 

Yes it should be good with full name and commit name on 
reactos website some where. So public can get know how is
how.

I still stand we need these gudlines I worte. 

  ----- Original Message ----- 
  From: Casper Hornstrup 
  To: 'ReactOS Development List' 
  Sent: den 16 October 2005 00:26
  Subject: RE: [ros-dev] Gudie Lines for SVN Write access


    1.. We have a Patches component in Bugzilla for this. 
    2.. The committer should do minimal testing even if he didn't create the patch. At
least check that ReactOS can still be built and boot. 
    3.. It is enough to have a well-known committer recommend you get repository access.
This is usually the person that has committed most of your patches. I see no reason for
this to change since it is simple and works quite well. 
   

  I can maintain a page in wiki containing full name and username of each committer if
needed.

   

  Btw. Christoph von Wittich has posted to the mailing list several times.

   

  Casper

   


------------------------------------------------------------------------------

  From: ros-dev-bounces(a)reactos.org [mailto:[e-mail]] On Behalf Of
Magnus Olsen
  Sent: 16. oktober 2005 00:01
  To: ReactOS General List; ReactOS Development List
  Subject: [ros-dev] Gudie Lines for SVN Write access

   

  Hi I should love to see some new gudieles 

  how to get svn access. Today it seam anyone

  that provide with a patch that is coder can 

  getting svn access. But I think we need 

  start think how we should handle it,

  Here is some ideas how it can be

  1. A maling list with patch or in Bugzila
      I love see a maling list with the patch

   

  2.  When people have submit the patch

  to us we start examing it see if it any godd
  (we are doning that already) 

  3. To get SVN write access u need lest 
  provide patch in regual basic under 

  6 month lest, Then after 6 month the provider
  can write to mailing list see if can getting 
  SVN write access before he grant SVN 

  write access, the full name and mailing 

  address must be provided, and we should 
  have a vote if that provder can getting 
  SVN write access. no accpect from  all 

  this rules. 

   

  Yestday some was granted svn access
  his name was not on the mailing list
  why he got one, he did not write either 
  on the mailing list asking for svn access
  that why I want see new guide lines how

  svn write access handles. so every one 
  know how he is and why he was granted

  to be granted after few patcher are not

  accpect in my eys.  For we are starting
  getting alot with people with svn write
  access. And new guide lines must be

  create. 

  BestReagds 
  Magnus Olsen

   



------------------------------------------------------------------------------


  _______________________________________________
  Ros-dev mailing list
  Ros-dev(a)reactos.org
  http://www.reactos.org/mailman/listinfo/ros-dev


------------------------------------------------------------------------------


  No virus found in this incoming message.
  Checked by AVG Anti-Virus.
  Version: 7.0.344 / Virus Database: 267.12.0/134 - Release Date: 2005-10-14

[Ged]

Magnus Olsen wrote:
>   And we need a system how u getting svn 
>  access it must be bit harder getting svn write access. 
I have to disagree with this. We want to attract new developers. Making 
it difficult to get commit access won't do this.
Committing to bugzilla and relying on other devs is somewhat annoying 
and can stop people committing.

Ged.

[Aleksey Bragin]

Well, Ged - do you know that sometime ago we were discussing the  
process of one commiter, who would review ALL patches from ALL devs.
Certainly ReactOS wasn't ready for that decision back then, and IMHO  
this idea of 1 (one) committer per whole project is unacceptable.

But for example for WINE - it's not a big problem, everyone is  
commiting patches which Alexandre reviews and commits to the tree.
Linux - for years it was only Linus reviewing patches sent to him. I  
don't know if it's still this way, but anyway.


But personally I'm not against giving more people commit-right. I'm  
for it, but candidate to devs should meet those minimum criterias.
And, intention to give them commit access should be posted on mailing- 
list, and if no objections come within some days (week? a few days?)  
the person should be given commit-right.


What do you think?


WBR,
Aleksey Bragin.


On Oct 16, 2005, at 3:28 PM, Ged wrote:

>   Magnus Olsen wrote:
>
> >   And we need a system how u getting svn access it
> > must be bit  
> >  harder getting svn write access.
> >  
>  I have to disagree with this. We want to attract new developers.  
>  Making it difficult to get commit access won't do this.
>  Committing to bugzilla and relying on other devs is somewhat  
>  annoying and can stop people committing.
>
>  Ged.

[Ged]

Aleksey Bragin wrote:
>   Well, Ged - do you know that sometime ago we were
> discussing the 
>  process of one commiter, who would review ALL patches from ALL devs.
>  Certainly ReactOS wasn't ready for that decision back then, and IMHO 
>  this idea of 1 (one) committer per whole project is unacceptable. 
I agree. especially with this project, I just don't think it would work.

>  
>  But for example for WINE - it's not a big problem, everyone is 
>  commiting patches which Alexandre reviews and commits to the tree.
>  Linux - for years it was only Linus reviewing patches sent to him. I 
>  don't know if it's still this way, but anyway. 
I'm not convinced it's the right way for the Wine project, but I don't 
fully understand the in's and outs.
It makes sense for Linux though as it's used widespread commercially. 
Maybe if ReactOS becomes as successful then tighter control will have to 
be used on the main branch. However, that's in the future.

>  
>  But personally I'm not against giving more people commit-right. I'm 
>  for it, but candidate to devs should meet those minimum criterias. 
I agree. But I think all the devs who are getting commit access now 
deserve it, and are working within their capabilities to improve the 
project.

>   And, intention to give them commit access should be
> posted on 
>  mailing-list, and if no objections come within some days (week? a few 
>  days?) the person should be given commit-right. 
Yes, I think this is a good idea and should be adopted with any new devs.

Ged.