[rear-devel] RFC: what should I do with my ideas for "rear 2.0"?

Johannes Meixner jsmeix at suse.de
Fri Oct 30 12:20:39 CET 2015


On Oct 30 12:51 Gratien D'haese wrote (excerpt):
> I have no problems with a rear-2, but currently I have no clue
> what you want to modify at all?
> Perhaps, prepare a design document first?
> So, we simple coders can follow you reasons and ideas.

Yes - of course - don't worry - usually I provide a lot
of documentation.

> Perhaps, it would be better to fork rear-2 from rear to avoid
> interruption for those who use rear in production?
> Afterwards, we could merge the branches again.
> Or not, if it deviates too much.
> If rear-2 becomes usable and successful we could leave
> the rear-1 branch as is.

I appreciate it very much!
This is exactly my plan.

And if the rear-2 fork becomes not usable or not successful
it would have at least shown what not to do and how to
continue usable and successful with the rear-1 branch.

Nothing can go really wrong with a separated rear-2 branch.
Things can only get better with it.

What I do not want is that a separated rear-2 branch
becomes a SUSE-only thingy.

I will try hard to make it working also for other Linux
distributions because I will implemenmt it as far as I can
in a generic way and keep special stuff sufficiently separated
from generic code (e.g. SUSE's 'snapper' default subvolume).

But I will not have the time to test other Linux distributions
sufficiently well which means I would need at least testing
from users of other Linux distributions.

> However, if it is a matter of *not enough* documentation
> in the code - why not deal with that and ask us what you
> do not understand?

I will do that of course but each question is an interruption
of what one actually likes to implement and then I like to be able
to just proceed with what I actually like to implement.

For this it is important for me that I can work on a separated
rear-2 branch so that I do not need to worry too much about
possible regressions.

Just as an example while fixing
I asked for such information but I have to wait until it comes
so that for now I just fixed it by minimal changes but I did not
clean up the current fragile code.

> Some parts not written by the core devs are difficult to read,
> but rewriting everything to your taste is very time consuming
> and as nobody wants to pay for it we leave it as it is (if the
> code works). We just fix what is broken.

Yes - exactly as I do currently - and currently I am not
really satisfied with my contribution to rear.

In contast on a separated rear-2 fork I could just go ahead
and clean up everything that I need (which is not everything
that exists but proably much of the core functionality).

Next week I will start (provided there is no urgent
customer issue that pervents that)!

I look forward to much more contribution to rear from me
and to a vivid discussions about this and that in rear.

Have a nice weekend!

Kind Regards
Johannes Meixner
SUSE LINUX GmbH - GF: Felix Imendoerffer, Jane Smithard,
Graham Norton - HRB 21284 (AG Nuernberg)

More information about the rear-devel mailing list