ECF participation in the annual simultaneous release

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

ECF participation in the annual simultaneous release

Scott Lewis-2
Hi Folks,

ECF 3.13.1 was released on-time as part of the Neon Simultaneous
Release.    ECF has participated in all 11 Simultaneous Releases and  
has been an Eclipse Platform dependency (filetransfer for p2) for > 5
years.  In that time we have met all SR requirements, met our all our
commitments to Platform, and never been late.  I think that's an
incredible accomplishment and record for a small all-volunteer team.

Unfortunately, I think we are at a point where ECF's SR participation
cannot continue.   As project and Foundation requirements have steadily
increased, the support for ECF (committer time) has decreased to the
point where I as project lead cannot sustain SR participation because of
my limited personal resources.   As well, I now think that ECF gets far
less from the SR than we put into it, and that inequity also can't
continue.   If the current resources situation doesn't somehow change
substantively over the next few months, my intention is to withdraw ECF
from the next SR.

Note this does not mean that ECF will slow down in our delivery to and
support of our community.   We have typically averaged > 4 releases per
year, and I expect that will continue...and perhaps even increase.   Non
participation in the SR will only mean that I won't have to do all the
work necessary for meeting the SR release requirements as well as doing
the actual releases, and so will be able to concentrate on adding value
in other ways...e.g. supporting the developer/Remote Services user
community, doing more examples/tutorial/documentation, creating more
distribution and discovery providers, recruiting, mentoring, and
supporting ECF committers, and working on additional Eclipse-based
tooling to make Remote Services/RSA development easier.

I wanted to bring this up early in the cycle so that this could be
discussed in public via this list.   If you have comments please feel
free to make them either in response to this posting in public
(preferred) or to me directly if you deem appropriate.

Thanks,

Scott



_______________________________________________
ecf-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/ecf-dev
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ECF participation in the annual simultaneous release

Wim Jongman-2

Hi Scott,

I don't think that this is a good idea. Every time I open the release train repo I see ECF. This must almost be on a weekly basis. So will thousands other people.

I think it is valuable 'free' publicity and with the whole microservice and IoT surge I think giving up your front seat row is a bad idea.

I don't think the overhead is all that much especially since you already plan to release 4 times.

I would rather advice to release 3 times but still keep riding the train.

Cheers,

Wim

On Jul 6, 2016, at 22:54, Scott Lewis <[hidden email]> wrote:
Hi Folks,

ECF 3.13.1 was released on-time as part of the Neon Simultaneous
Release. ECF has participated in all 11 Simultaneous Releases and
has been an Eclipse Platform dependency (filetransfer for p2) for > 5
years. In that time we have met all SR requirements, met our all our
commitments to Platform, and never been late. I think that's an
incredible accomplishment and record for a small all-volunteer team.

Unfortunately, I think we are at a point where ECF's SR participation
cannot continue. As project and Foundation requirements have steadily
increased, the support for ECF (committer time) has decreased to the
point where I as project lead cannot sustain SR participation because of
my limited personal resources. As well, I now think that ECF gets far
less from the SR than we put into it, and that inequity also can't
continue. If the current resources situation doesn't somehow change
substantively over the next few months, my intention is to withdraw ECF
from the next SR.

Note this does not mean that ECF will slow down in our delivery to and
support of our community. We have typically averaged > 4 releases per
year, and I expect that will continue...and perhaps even increase. Non
participation in the SR will only mean that I won't have to do all the
work necessary for meeting the SR release requirements as well as doing
the actual releases, and so will be able to concentrate on adding value
in other ways...e.g. supporting the developer/Remote Services user
community, doing more examples/tutorial/documentation, creating more
distribution and discovery providers, recruiting, mentoring, and
supporting ECF committers, and working on additional Eclipse-based
tooling to make Remote Services/RSA development easier.

I wanted to bring this up early in the cycle so that this could be
discussed in public via this list. If you have comments please feel
free to make them either in response to this posting in public
(preferred) or to me directly if you deem appropriate.

Thanks,

Scott





ecf-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/ecf-dev

_______________________________________________
ecf-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/ecf-dev
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: ECF participation in the annual simultaneous release

Scott Lewis-2
Hi Wim,

On 7/6/2016 2:18 PM, Wim Jongman wrote:

>
> Hi Scott,
>
> I don't think that this is a good idea. Every time I open the release
> train repo I see ECF. This must almost be on a weekly basis. So will
> thousands other people.
>
> I think it is valuable 'free' publicity and with the whole
> microservice and IoT surge I think giving up your front seat row is a
> bad idea.
>

I think characterizing it as 'free' publicity and 'giving up a front row
seat' is not accurate.   Truth is, according to Foundation we are not in
IoT and don't receive any publicity...either as a Runtime Project or as
an IoT project (logically we are both RT and IoT, but don't get any of
the promotional/marketing/visibility benefits for either of them).  For
a timely example of what I'm talking about, please see discussion on
this bug [1] and subsequently on the runtime pmc mailing list [2].

[1] https://bugs.eclipse.org/bugs/show_bug.cgi?id=493809#c124
[2] https://dev.eclipse.org/mhonarc/lists/rt-pmc/msg04666.html

I'm not arguing that leaving the SR is necessarily a 'good idea', I'm
just saying that under the current conditions, I can't meet the
necessary commitments and so since I'm currently responsible for those
commitments as project lead, I won't make them without having a way to
meet them.

> I don't think the overhead is all that much especially since you
> already plan to release 4 times.
>

There is significant overhead of the SR that I experience as project
lead and primary active committer....meeting requirements,
communication/bug fixing with Platform and SR projects, regular releng,
doing specialized work under continually more difficult SR releng
conditions (for example last year see [3]), and other things.

All of [3] was/is imposed by the SR and would not even be of concern
with ECF-only releases.    In my experience, something akin to [3] has
occurred nearly every SR (e.g. with p2/Filetransfer).   Please trust
me...it is significant overhead, even if it doesn't seem so on this
mailing list.

[3] https://bugs.eclipse.org/bugs/show_bug.cgi?id=487857

> I would rather advice to release 3 times but still keep riding the train.
>

Here's my bottom line:   To stay on the train we need a commitment to do
the necessary work.   I'm not able to continue being the only jockey
actually riding the train when it starts moving (i.e. see the part that
requires significant overhead) with no visible means of support.   I've
asked previously for that commitment from Platform, EF, member companies
that depend upon us, and others, and it hasn't come through in previous
cycles, so I'm not at all confident that it will during this cycle.

Therefore, the only way I will agree to have ECF continue riding the SR
is when the request comes with necessary resources.  Not because I think
it's a 'good idea', but because that's currently all I can do.

Scott


_______________________________________________
ecf-dev mailing list
[hidden email]
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/ecf-dev
Loading...