[hibernate-dev] Stride

Tue May 22 04:26:25 EDT 2018

I think the problem here is about how we're choosing to migrate to another
tool. We are basically moving to Stride because Atlassian is being
uncooperative with updating HipChat, even though Stride is still not fully
functional.
Atlassian made HipChat, and we're not happy with HipChat. Atlassian made
Stride too, so let's not trust them blindly? Sure, we can always move to
another platform after moving to Stride, but it'll be a waste of time for
everyone. Let's see what our options really are, first!

I don't think we have many needs, and frankly I don't care which tool we'll
be using. I'd rather it be open source, but I'll just use whatever we pick
in the end, as long as I can write messages and receive messages. But I
want people to be on this platform: both the team, and ideally community
members. So let's just make sure everyone is at least comfortable with our
choice.
Sure some features are not useful to everybody, but in the end it's a
matter of taste and personal workflow, and we can't argue about that
without enforcing a standardized workflow. I think it would be more useful
to just list what we want, then see which platforms fit best, then pick
one, compromising as needed.

I started a document to compare various solutions, feel free to add
anything you think is relevant:
https://docs.google.com/spreadsheets/d/1oiI_SO4O5OwTx7Fd8_lvhOvVcIIYLbnSY1MAqemz-JI/edit?usp=sharing

... But I think the one criteria that will make us pick Stride is free
hosting. Most other platforms either do not have a free plan, or do not
provide all of their features to free plan users. Zulip apparently removes
OAuth authentication in its free plan, for instance. The Infinispan team
has OAuth authentication enabled though... Do they pay for their Zulip
instance?

Of course if we could host it ourselves it would be another story, but I
doubt anyone wants to add that to the list of tools we currently maintain.

On Mon, 21 May 2018 at 14:02 Guillaume Smet <guillaume.smet at gmail.com>
wrote:

> Hi,
>
> On Fri, May 18, 2018 at 1:37 PM, Steve Ebersole <steve at hibernate.org>
> wrote:
>
> > You can't "do your job" without yet another way to get notified of CI job
> > status?  I'm confused - did Jenkins remove all of its other forms of
> > notification?  ;)  Seriously though,  I've never understood this desire
> to
> > have yet another communications "inbox" spammed by automated
> notifications
> > - and its even worse in Hip Cat because I can never hide them.  So it is
> > hard for me to incorporate this into the argument against moving.
> >
>
> Typically, for the websites builds, I like to have the notifications on
> HipChat. Not vital but it's something I appreciate.
>
>
> > Y'all really wanted to move to Hip Chat in the first place even though to
> > me it always felt (and feels) half-baked itself.
> >
>
> I wasn't part of the team at that point.
>
>
> > And if the web client is as good as the native client, I assume you use
> > the web client instead?
> >
>
> Yes. I have a pinned tab for it. Desktop notifications from the browser are
> well integrated now.
>
>
> > So to me it really comes down to what are the blockers to not making this
> > move now.  So far I hear:
> >
> >    1. No Jenkins notifications - see above
> >    2. Guest access - meh - If having to have an account to join the
> >    discussion is bad then we should immediately make our forums
> >    guest-accessible again as well ;)
> >
> > As mentioned in my email, I don't mind not having guest access if we have
> some sort of external auth integration (Github/Twitter/Google). This way,
> it avoids creating yet another account. That's what we have on our new
> forums.
>
> Maybe Stride has it, I don't know, but I would like our next chat system to
> have that.
>
> I took a look at several accounts we have on our forums and most of them
> are from GitHub or Google (I took seven randomly and 1 had an account
> specific to our forums - no better statistics sorry).
>
> >
> >    1. There may be better options out there - at some point can we just
> >    pick one and use it?  Is one "inbox" really that much better than
> another
> >    "inbox"?  And clearly I am not even tied to Hip Chat - I was one of
> the
> >    people wanting to not move there.  Radim, what makes Zulip so amazing?
> >    2. "Coordination tool"?  Not really sure what this one is about.  Is
> >    this back to Jenkins notifications?  If you mean a communications
> tool, of
> >    course it works.  They are largely the same.  Andrea, Sanne and I have
> >    played with it, so we in fact do have some idea if it will (spoiler:
> it
> >    does)
> >    3. We should go where WildFly goes (?).
> >
> > My main point is that we have no idea if Stride is stable at the moment.
> They definitely don't want to massively migrate the HipChat users and I
> found at least one comment of a person who has migrated there and is not
> happy at all (not saying it makes all the migrated users unhappy, just
> saying that it might be a bad move if we do it now).
>
> So again, what problem are we trying to solve by migrating to Stride while
> it's still a work in progress? Because you reversed the question but it is
> the correct question.
>
> - If it is the client issue, is the Stride client better integrated? And is
> it enough to trigger the migration?
> - If it's the access difficulty, does Stride allow external
> authentications? Because that's what external users want.
> - If it's the "it's a pain to be logged in to multiple instances", are we
> all moving to Stride and will it solve the issue - meaning is it easier to
> be connected to multiple instances?
> - Other issues I'm not aware of?
>
> When you want to migrate to something new, you ponder the risks with the
> benefits. For now, noone has stated any clear benefit, apart from "it's
> new" and "HipChat has issues but we don't know if they will be solved by
> Stride".
>
> Note that my position is not "we should stay on HipChat forever" because we
> can't as they will close the service at some point. My position is:
> - I'm not sure the risk of moving to a half-baked system is worth it if
> there are no clear benefits (I would wait a bit)
> - is Stride gonna solve our issues or do we need to move to another system?
> (because if it's the case, let's avoid the Stride square).
>
> --
> Guillaume
> _______________________________________________
> hibernate-dev mailing list
> hibernate-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/hibernate-dev
>
-- 
Yoann Rodiere
yoann at hibernate.org / yrodiere at redhat.com
Software Engineer
Hibernate NoORM team