Re: GNOME Online Accounts 3.34 won't have documents support

[Debarshi Ray <rishi is lostca se>]

On Wed, Jan 23, 2019 at 05:39:44PM +0100, Bastien Nocera wrote:
> As Emmanuele mentioned, the problem isn't so much that services will
> disappear from under the applications (but it's a problem nonetheless),
> it's that there was no communication explaining that applications
> shouldn't have relied on GNOME Online Accounts in the first place, as
> the functionality could disappear for reasons not caused by those
> services, or applications.


You are talking as if the application maintainer and the GNOME Online
Account maintainer are two disjoint entities. As if an active
community of contributors have been jeopardized by the arbitrary will
of the mythical GOA maintainer.

That's false.


[rishi@kolache gnome-documents]$ git shortlog -ns | head
  1036 Cosimo Cecchi
   357 Debarshi Ray
    78 Alessandro Bono
    76 Daniel Mustieles
    68 Piotr Drag
    59 Bastien Nocera
    52 Kjartan Maraas
    39 Marek Cernocky
    36 Matej Urbancic
    36 William Jon McCann

Since everybody is concerned about the Online Accounts integration,
let's look at gnome-online-miners.git. That's where the said
integration lives.

[rishi@kolache gnome-online-miners]$ git shortlog -ns | head
   101 Debarshi Ray
     6 Pranav Kant

And I am just not going to bother digging up review statistics from
Bugzilla. ;)

I was also the only maintainer at least pretending to keep up with the
GNOME schedule.


There wasn't any active community.

We regularly released with glaring bugs that some of our downstreams
would consider blockers.  Fedora releases would have blocked, had
those bugs been known. They weren't known because very few people, if
any, ever used the application, so nobody ever reported them.

RHEL 7.x releases actually did block on those bugs. That's how those
eventually got noticed, fixed and backported.

Boy, did I spend hours diligently backporting all those fixes,
spinning tarballs, doing downstream builds. Sometimes the backports
went across three or four stable branches - that's how glaring and old
some of the bugs were. Not a soul cared.

These bugs were regressions introduced by the occasional patch that
would get merged, or by changes in our JavaScript stack, or something
else. The upshot being that the reviewers themselves weren't using the
application much, or didn't have enough time to diligently review the
patches; nor did it have any users in the wider GNOME community and
beyond.

It was also clear that the GNOME designers weren't that excited about
GNOME Documents any more.


Yes, I could have started by listing all the reasons behind why
Documents is considered a dead-end. I didn't do that, so I am very
sorry about that.

However, I did give a brief background in my very next email to this
thread. Cosimo, Michael, Jakub and Allan were all part of the
discussions about it's future.

But then, I haven't yet found anybody honestly asking why we gave up
on it. Instead people went ahead and drew whatever conclusions they
wanted to draw. I find that odd.

It's not like this is the first time we have dropped things from GNOME
Online Accounts. Back in 2017 [1] we had dropped Telepathy. I had
written a wall of text explaining that decision. Guess how many
replies that thread got.  Surely, Telepathy has had a lot more users
in its time than Documents.

So, I do find it strange that people are suddenly coming out of the
woodwork passionately fighting for the survival of GNOME Documents.


[1] https://mail.gnome.org/archives/desktop-devel-list/2017-October/msg00040.html