16 Nov
2015
16 Nov
'15
8:43 a.m.
Ok. I propose that we simply agree on this best practice for now and we try it without any
automation and see how it goes:
Proposal text to add to our best practices on xwiki.org:
"
When someone reports or notices a blocker issue that will affect end users in a serious
way (the definition of “serious” is left to the discretion of everyone), he should add it
in the error block at the top of the release notes in which the issue exists.
The error block should have the following format:
{{error}}
The following important issues were found after this version was released. You should
verify if you’re using or need the affected features and decide whether to use a newer
version that has it fixed or continue with this release:
* <issue 1>
* …
* <issue N>
{{/error}}
"
Thanks
-Vincent
On 16 Nov 2015 at 02:08:00, Denis Gervalle (dgl(a)softec.lu) wrote:
On Sun, Nov 15, 2015 at 1:02 PM, vincent(a)massol.net <vincent(a)massol.net>
wrote:
Again, I am not sure this is so simple. Marking issue as blocker is
promising issues to be fixed for the release, and if this does not really
happen, be sure we care a lot about them, and not simply push them to the
next release without an agreement. This does not necessarily means that the
issue affect a large amount of users, and that it is so important that it
should prevent many users to try that release.
As I have explained earlier, if we do not change our current usage of
blocker (and I like this idea), and if I look at your auto-generated list
of blockers, many of them are unintelligible for the end users, and I do
not see what would be the benefit to list them, all it will do, is make
users confused or scared.
So, the manual decision is difficult to avoid. Maybe we could use a manual
tag to exclude (or include) some blocker from the list, meaning that we
would go for a show them by default (or explicitly), and we have a way to
explicitly make them not listed (or listed) while we still consider them
for the next release. But, I am not sure this is fine, since anyone (not
just committers) could create a blocker JIRA, and cause a change of the RN…
is this acceptable to list unconfirmed blockers ?
I do not know exactly what happened on Friday, but I would say that like
any manual process, we cannot be fully error safe, but I am sure all of us
will do their best to ensure we care about this in the future. Don’t you
trust we could do that ?
Hi Denis,
Thanks for your feedback, see below.
On 15 Nov 2015 at 12:30:13, Denis Gervalle (dgl(a)softec.lu(mailto:
dgl(a)softec.lu)) wrote:
nature
visible.
release,
* Proposal 2 (Denis, Edy): This is what we do now. Continue doing this
On Fri, Nov 13, 2015 at 9:48 PM,
vincent(a)massol.net
wrote:
>
>
> On 13 Nov 2015 at 19:35:56, Eduard Moraru (enygma2002(a)gmail.com
(mailto:
> enygma2002(a)gmail.com)) wrote:
>
> > I am +1 to keep doing what we have been doing so far, which is, on a
> > case-by-case basis and depending on the importance of the blocker, to
> > manually add a notice (/error macro) in the release notes, informing
> users.
> > If we want to document this as a best practice, I`m ok with it.
> >
> > I`m not a fan of increasing maintainance/work load in this aspect,
but
> > that`s just my opinion.
>
> I agree that low maintenance when possible is always better. That was
the
> point of the automated approach BTW. Even
though automated has
limitation
> don’t you think it’s better than what we’re
doing ATM (i.e. adding
stuff
> when we think about it, i.e. missing 80% of
what we should do)?
>
> About this, I don’t agree that what we’ve been doing is good because we
> almost never do it. AFAIR I’ve been almost the only one doing it (I
could
> be wrong please correct me if that’s the
case) and pushing others to
do it
and I
would prefer not to have to
I can only speak for myself, but I do as well. Now it depends of the of the blocker. All regression are blockers, but
the regression could
still
be minor and not so annoying for most users. In
that case, I do not. Up
to
now, what we put at the top in error box were
blocker that we think make
the release unrecommended for most users. And IMO, we should continue
doing
this manually at least.
> continue doing this all the time. An example was later today, we
created a
> very important blocker and fixed it and yet
we didn’t add it to the RN
of
> 7.3 until I raised it and added it myself.
I’d rather not be the one
having
> to think about this and pestering others
about it...
>
> I’d like a common agreement on whether we want to add blockers to RN or
> not. If we think it’s not a good idea then I’m fine and we shouldn't
add
them.
We already add know issue we know about to the RN, so adding blockers at
the top is not really adding them to the RN, but making them more This could be a good idea but not in replacement
to our current practice,
since it is less understandable, and more common then our current
practice,
making the information less meaningful for users.
This could even cause
more confusion, if those blocker are minor regression on specific areas.
>
> What I’m not fine about is saying that we think we should add blockers
and
> then not to do it, that doesn’t make sense.
>
> Regarding importance of blockers, for me a blocker is a blocker,
otherwise
it’s not
named correctly and we should fix this :)
This is not so true, blocker means we don’t want this in the next and as I said, it is generally true for any
regression.
This is what I was trying to say with “for me a blocker is a blocker”. I think
we’re saying the same thing here:
* We need a way to mark (in JIRA) very important issues that affect some
user-feature of xwiki (since RN are for users)
* I agree that regressions don’t have to be blockers (and they’re not all,
there are several regressions we don’t mark as blockers because they’re
not…). You should know since the Mail API has had a few regressions you
introduced and they’re marked as regressions but not blockers ;)
But the actual
blocker could have several level of importance for different kind of
user,
and in the end only affect very few of them but
so hardly that it is
still
blocker. If you look at your sample, I am even
afraid by what users will
understand of them. Either they will be afraid so much, that they will
not
upgrade, or they will not understand and consider
they are not affected.
See XWIKI-12502, XWIKI-12342, XWIKI-12251, what users will think of them
?
Yes that’s the risk and that’s the reason we have created manual release
notes indeed. Because just listing jira issues is not user-friendly enough.
And on the other side, XWIKI-12218, while we want
it in the next release,
was it really so blocker for end users ?
Again, there’s only a global process which is to separate real blockers
from a user POV (which we want highlighted in the RN) from important issues
that we want fixed but that are not really blockers for users.
The only question is when do we do this separation:
* Proposal 1 (Vincent): Do this sorting in JIRA and have a generic query
in the RN for listing those issues
sorting at random, by random people when they think
about it. In most cases
it’s forgotten and not added to RN.
I do not know for Edy, but for the above definition does not really
translate what I have said so far !
What I said, is that we cannot simply list blockers at the top to replace
the actual manual warnings because JIRA tickets are often unintelligible
for end users, and that all blockers (as we used to do them so far) are not
at the same level, and does not necessarily need to be enhanced at the same
level. That if we go that way, it means we need to rethink the way we use
blocker in JIRA, and that I do not believe we should.
I’m in favor of proposal 1, ie improve our JIRA tagging so that we
identify real user blockers and list them automatically in the RN. This
would be the same process as we have now, ie we review issues being created
every day
I’d be also ok for proposal 2 but only if we can define a process so that
we don’t forget about them (and that it’s not me doing it most of the time
and being the bad guy reminding everyone).
If this is really intelligible from the simple JIRA title, I would be +1,
but I doubt we can really achieve that.
In your example above, those that
could be interesting and understandable warnings for some users are
XWIKI-12612, XWIKI-12160 and XWIKI-11993.
So listing them in red at the very top may be too much, or confusing IMO.
This does not mean the the could not list them somewhere in the RN, maybe
in a {{warning}} macro, below the possible {{error}} we do manually. This
could also help us think about doing a red one, when looking at the RN.
I think listing in red at the top is ok for really serious issues (ie real
blockers from user POV), i.e. to tell users to NOT use this version if they
used the feature mentioned.
What afraid me in the end, is that we do less
blockers just because we
know
they will appear in RN and they are so technical
that nobody will get it.
That would be fine IMO because it would mean they’re not real blockers and
that releasing without them is ok. In the end what should be blocking us
from a release is the real end user blockers.
So the next
question is, does all of them represent what you wanted at
the
top, compare to other issues ? Is those blocker
more significant ? Should
we change the way we use blocker ?
Yep I think this is the real question.
I am not sure we should change anything in our
current usage of blocker,
we
may want to have them enhanced in RN, but it does
not mean all the
information will be so significant.
I’m ok to not change our current usage of blockers but I don’t understand
what you’re suggesting for RN. You’re not addressing the reason for this
thread in the first place, which is that I feel that we just add warning at
random and we forget a lot and that it’s mostly me doing it. How do we
prevent what happened Friday (ie. Marius not adding the blocker to the top
of the 7.3 RN)?
Thanks
-Vincent
--
Denis Gervalle
SOFTEC sa - CEO
_______________________________________________
devs mailing list
devs(a)xwiki.org
http://lists.xwiki.org/mailman/listinfo/devs
Thanks,
>
> Thanks for your feedback Edy. I’m curious to know what the others
think at
> this point! :)
>
> Thanks
> -Vincent
>
> > Thanks,
> > Eduard
> >
> > On Fri, Nov 13, 2015 at 8:21 PM, vincent(a)massol.net
> > wrote:
> >
> > >
> > >
> > >
> > >
> > >
> > > On 13 Nov 2015 at 19:14:11, Eduard Moraru (enygma2002(a)gmail.com
> (mailto:
> > > enygma2002(a)gmail.com)) wrote:
> > >
> > > > Hi,
> > > >
> > > > Another problem with automatic is that we don`t always tag
> > > affects-version
> > > > on the actual version when it was introduced, but the version we
> actually
> > > > tested with. This might lead to displaying blockers on the
release
> notes
> > > of
> > > > versions that were not responsible with introducing the actual
> problem.
> > > >
> > > > Until now, we haven`t had that many cases and the 7.2, 7.3
releases
> could
> > > > be considered (from my POV) exceptions which had multiple issues
due
> to
> > > the
> > > > nature/complexity of the changes they introduced.
> > > >
> > > > So, in the end, I believe I`m +1 to stick to manual.
> > >
> > > Whether manual or automatic, are you +1 that we do this as a best
> practice
> > > (ie always add blockers to RN, at the top and using an error
macro)?
> > >
> > > Thanks
> > > -Vincent
> > >
> > > >
> > > > Thanks,
> > > > Eduard
> > > >
> > > > On Fri, Nov 13, 2015 at 7:00 PM, vincent(a)massol.net
> > > > wrote:
> > > >
> > > > > To summarize:
> > > > >
> > > > > Pros and cons of automatic:
> > > > > + Not missing any
> > > > > - Technical descriptions
> > > > >
> > > > > Pros and cons of manual:
> > > > > + Nice hand-made messages
> > > > > - Usually missing some, unless we add a process (like have the
RM
> > > ensure
> > > > > that none are missing the previous release notes)
> > > > >
> > > > > WDYT?
> > > > >
> > > > > Thanks
> > > > > -Vincent
> > > > > On 13 Nov 2015 at 17:56:00, vincent(a)massol.net (
vincent(a)massol.net
> )
> > > wrote:
> > > > >
> > > > > OTOH if we do it automatically we won’t have as nice
description as
> > > what
> > > > > we can have when we hand-make them. For example on
> > > > >
> http://www.xwiki.org/xwiki/bin/view/ReleaseNotes/ReleaseNotesXWiki71
> > > we
> > > > > have:
> > > > >
> > > > > “
> > > > > This release introduces a bug concerning the upgrade of the
> subwikis
> > > > >
([[XWIKI-12208>>http://jira.xwiki.org/browse/XWIKI-12208]])208]]).
You
> can
> > > > > still upgrade them via Distribution Wizard but only at the same
> time
> > > you
> > > > > upgrade the main wiki. You can also upgrade your subwiki's
UI
like
> any
> > > > > other extensions, by clicking on "Check for updates".
> > > > >
> > > > > The best is to wait for the release of 7.1.1 which will fix
this
> > > problem
> > > > > and should be ready really soon.
> > > > > "
> > > > >
> > > > > And if we use the strategy below we get:
> > > > > https://www.evernote.com/l/AHfEQXsj4mNH4pjXgxuiL4uZiomR5kUpkpA
> > > > >
> > > > > As you can see on the screenshot if we do it manually we’re
> missing a
> > > lot
> > > > > of blocker issues.
> > > > >
> > > > > WDYT?
> > > > >
> > > > > Thanks
> > > > > -Vincent
> > > > >
> > > > >
> > > > > On 13 Nov 2015 at 17:51:50, vincent(a)massol.net (
vincent(a)massol.net
> )
> > > wrote:
> > > > >
> > > > > Hi devs,
> > > > >
> > > > > I’d like to suggest a new best practice which would consist in
> > > > > systematically adding blockers issues in the release notes
> > > corresponding to
> > > > > the affects versions, at the top, using an error macro. For
example
> > > I’ve
> > > > > updated the release notes for 7.2 and 7.3:
> > > > > -
> http://www.xwiki.org/xwiki/bin/view/ReleaseNotes/ReleaseNotesXWiki72
> > > > > -
> http://www.xwiki.org/xwiki/bin/view/ReleaseNotes/ReleaseNotesXWiki73
> > > > >
> > > > > The idea would be to update our Release Notes template to
> automatically
> > > > > list them using this template snippet (example for XWiki 7.2):
> > > > >
> > > > > "
> > > > > {{error}}
> > > > > The following blocking issues were found after this version was
> > > released.
> > > > > You should verify if you're using the affected features and
if
so,
> you
> > > can
> > > > > click on them to see in which version they are fixed):
> > > > >
> > > > > {{jira url="http://jira.xwiki.org"
style="list" source="jql"}}
> > > > > category = "Top Level Projects" and affectedVersion
in
> > > > > ("7.2-milestone-1", "7.2-milestone-2",
"7.2-milestone-3",
> "7.2-rc-1",
> > > > > "7.2") and fixVersion > "7.2" and
priority = Blocker and
> resolution =
> > > Fixed
> > > > > {{/jira}}
> > > > > {{/error}}
> > > > > “
> > > > >
> > > > > The goal is to make it easy for our end users to see blocking
> issues
> > > in a
> > > > > given release. The goal is also to make it visible to us how
many
> > > blockers
> > > > > we introduce in each release and work on reducing this number
as
> much
> > > as
> > > > > possible.
> > > > >
> > > > > WDYT?
> > > > >
> > > > > Thanks
> > > > > -Vincent