Re: [DPP-Devel] IntelliJ file system implemantation, v2?

[Zieris, F. <Fra...@fu...>]

Franz asked:

> Sounds like something bigger (?) that should have been discussed on the
> mailing list.
>
> @Stefan, @Tobias: Could you spare a few words on this?

Stefan (forgetting to include dpp-devel in CC) answered:

> http://saros-build.imp.fu-berlin.de/gerrit/#/c/3314/


For further reference, I copy the discussion from there:
(new content follows below the ===== line)

Stefan wrote:

> Just FYI. I am starting to integrate some parts of your big patch.
> It is not the original code. As I already stated, we do not need
> Unbound stuff if we are working with paths instead of virtual file
> objects.

Tobias wrote:

> I don't want to sound ungrateful (as I currently am somewhat low on
> time and therefore appreciate your efforts to integrate my patch),
> but this patch has somewhat put me in an awkward position.
>
> Firstly, it is kind of harsh to work on something for multiple weeks
> and then only see a tiny sliver of it make it into your integration
> (I am aware that this is a very streamlined version throwing out all
> the unneeded code written for future expandability; these were
> written before the prototype nature of the current implementation
> was made clear to me). I know this is more of a personal problem,
> but I hope you can see where I am coming from.
>
> Secondly (and more importantly), this somewhat complicates my further
> work as I am still set on fixing the functionality concerning the
> change of the session scope (SharedResorucesManger and
> FileSystemChangeListener).
> What am I supposed to base these fixes on? I have no frame of
> reference how long the integration is going to take and when it
> will be at a testable point.
> Should I just base it on my original big patch? And how am I
> supposed to push it in that case?
>
> Furthermore, I would like to do the integration work myself in the
> future as it is part of the work (and learning) experience and I am
> the one who is going to work with it afterwards, which is much
> easier if I have written and integrated the code myself.
>
> As I already said, I appreciate your work, especially since the
> scope of the patch got way out of hand and I am running low on
> time. I hope that all of this will work more smoothly in the future.
>
> I also don't know how I should proceed with the big patch. Should
> I still push the last patchset with the small changes I already noted?
> Should I complete the classes written in this patch? What was your
> thinking on how this is supposed to proceed? Are you going to do the
> entire integration and re-re-implementation yourself?

Stefan wrote: 

> Well, just look at the patches (the abandoned one) containing my name.
> Sometimes Franz already revert some of my features. It is
> frustrating but you have to deal with it.
>
> But I already told you that you do not need Unbound resources if
> you are working with paths instead of virtual files.
>
> Just because someone writes a big bunch of code does not give
> him/her the rights to insist that the code gets committed.
>
> As for the bug fixing stuff, just do it in your big patch.
>
> The splitting after you are finished shouldn't be that hard,
> believe me. It is basically a just renaming stuff after the V2
> files are ready.

=====================================================================

Here is what I take from this:
 * Tobias worked on the IntelliJ filesystem quite some time, discussing
   his patches with Stefan.
 * Stefan created an alternative implementation without discussing it with
   Tobias, and without proposing a procedure on how to integrate the efforts.
 * Tobias complained about this, asking on how to proceed.
 * Stefan did not answer with a proposal on how to proceed, but seems to have
   something in mind ("after the V2 files are ready")

OK, Stefan, here's is your chance: What do you propose?
Please share your conception on how to proceed in this manner.

On a different note: It sounds like you, Stefan, are in for "egoless
programming", but at the same time, you seem to hold some grudge against
me for "reverting some of your features".
I honestly have no idea what you're referring to.
The last time I git-reverted one of your commits was in 2012 [1,2].
In terms of Gerrit patches [3], I stopped looking when I reached 2014:
I did not abandon any of your patches.
So: What's the matter here?

Franz

[1] git log --author=zieris --grep=Revert"
[2] https://github.com/saros-project/saros/commit/a6ae0ef3567f
[3] http://saros-build.imp.fu-berlin.de/gerrit/#/q/owner:srossbach%2540arcor.de+status:abandoned