From: Rafaël Carré <rafael.carre_at_gmail.com>
Date: Fri, 9 Oct 2009 17:40:30 +0200

On Fri, 09 Oct 2009 17:22:57 +0200
Thomas Martitz <thomas.martitz_at_student.HTW-Berlin.de> wrote:

> On 09.10.2009 17:00, Rafaël Carré wrote:
> > On Fri, 9 Oct 2009 14:33:09 +0200
> > mailer_at_svn.rockbox.org wrote:
> >
> >
> >> Date: 2009-10-09 14:33:09 +0200 (Fri, 09 Oct 2009)
> >> New Revision: 23039
> >>
> >> Log Message:
> >> Remove PP bootloader v6 tag and mkamsboot 1.1 tags (going to be
> >> retagged).
> >>
> >> Removed:
> >> tags/bootloader-pp-v6_no-samsung/
> >> tags/mkamsboot_1.1/
> >>
> >
> > The reason I chose to use trunk is to not have to backport all the
> > changes that happened in trunk since then.
> >
> > That includes : removal of copyright notice and various changes to
> > Makefile.
> >
> > Since these changes do not critically affect mkamsboot, there's no
> > reason to tag mkamsboot from a branch.
> > And of course I tested mkamsboot myself before releasing it.
> >
>
> I saw no notice of that. I don't know what platforms you tested it
> on. (nor that you tested at all). I didn't know you already built
> binaries also.

I do not think I had to justify myself on those points.

> > Note that the binaries I built are already available on
> > http://download.rockbox.org/bootloader/sandisk-sansa/mkamsboot-1.1/
> > (even if the server still says 403)
> >
> > mkamsboot v1.1 really is what I built off this particular revision,
> > with APPVERSION defined to 1.1; a user who would build mkamsboot
> > from this checkout would read mkamsboot vrXXXXX, because it's not
> > what I built and labeled v1.1
> >
> >
>
> What did you do to make it read v1.1? Again, I saw no notice of that,
> I assumed it would show rXXXXX since that's what the trunk code does.

I edited Makefile

> > If we really need binaries built from trunk to be versioned, there
> > are other possibilities than using branches.
> >
> >
> > We won't keep a branch opened forever for future mkamsboot releases,
> > all the code committed to trunk will be tested and reviewed (by me
> > at least)
> >
>
> Well, my plan was to tag 1.1 off the branch, and then delete it. Then
> have 1.2 tagged from another branch (or trunk if we agree on).

I wasn't aware of a release plan for 1.1 after 1.1RC was made.

> 1.1RC is based off the branch too.

right

> > Please remove the mkamsboot tag you made from branch and re tag
> > mkamsboot v1.1 from trunk.
> >
> >
>
> It is already removed.

Will you tag mkamsboot_1.1 again from trunk ?

>
> >
> > Last but not least, please next time do bring the issue on the
> > mailing list before reverting a commit/tag without the opinion of
> > the committer.
> >
> > I feel very bad about you reverting my work without asking me first.
> >
> >
>
> Same feeling here. mkamsboot 1.1RC was based off the branch (which
> contains safe code), 1.1 is not.

trunk contains safe code as well.

> You didn't explain it anywhere. It
> rather looked like you would follow linuxstb who said " I never liked
> that branch, so would suggest just using trunk.".

What did I have to explain ?

In the irc logs just before this quote from linuxstb, I say:
"should I still use the bootloader_ams_pp branch for building new
mkamsboot binaries ? Then I would need to backport all commits made
since the branch was created"

Which is pretty explicit.

> You just made a release without waiting for opinions of other AMS
> port guys. Your lack of conversation confused me much.

I am the main author of mkamsboot and I didn't think opinions of other
people working on AMS port was needed at all.

> But I didn't behave much better, that's true, I was a bit bitter
> about it.

I don't think there is a reason for you to be bitter?


> We can avoid this if we're a bit more patient about releasing *very*
> critical tools.

I rather think the way to avoid bitter reactions if you wait and think
before acting.

The way I interpret your reaction, I think you got upset after reading
this quote from linuxstb on the irc log, interpreted this comment as
against you, and then thought I was against you as well.

Am I right?


> I don't want to "just release" tools that can brick
> hardware without having a talk with some involved guys about it.

It can not brick hardware, it was tested.

You didn't develop mkamsboot, I did, and I know it works.

-- 
Rafaël Carré

Received on 2009-10-09