incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ross Gardler <Ross.Gard...@microsoft.com>
Subject RE: Preliminary NetBeans cost findings (was: [DISCUSS] Apache NetBeans Incubator Proposal)
Date Sun, 25 Sep 2016 15:45:48 GMT
You seem to have taken my comment as an indication that I have concerns one way or the other.
That is not the case. What I'm saying is that to make a case for extra budget there needs
to be solid justification that  a move to ASF will help the community grow. The ASF is not
a magic bullet, there needs to be a plan coming from the incoming project. The costings here
are more than we usually get when a new podling is considered. This is a very good start.

The data I refer to is only one data point. If you have data that contradicts it then provide
it in your request for funds (yes this has been discussed to some extent across the main discuss
thread, but it needs to be packaged up nicely for VP Infra, Prez and finally Board to consider.

My one data point is http://pages.zeroturnaround.com/RebelLabs-Developer-Productivity-Report-2016.html?utm_source=rebellabs_allreports&utm_medium=rebellabs&utm_campaign=rebellabs
(requires sign in). That reports shows a decline from 14% in 2012 to 10% today. To be fair
that has been steady since 2014.

The reason for my explicit request is that the foundation is currently running at a significant
deficit. That's not a problem since we have many years of cash in the bank at the current
deficit. However, we do need to plan for the future. So any new budget requests need to be
fully justified. That's all I'm asking for. A "just because" is not sufficient. Like you and
others have said there needs to be evidence to back up claims, simply adopting the apache
way does not mean that NetBeans will be successful as an Apache project. If my data (limited
to the above single data point) is inaccurate/invalid/not representative then you should have
no problem providing evidence to the contrary when you ask for this budget.

One final note, back in Jan 2015 the board approved a limited experiment with directed sponsorship
to help alleviate issues like this. Maybe this would be useful to the NetBeans community.
See presidents report here: http://apache.org/foundation/records/minutes/2015/board_minutes_2015_01_21.txt

Ross

> -----Original Message-----
> From: me@wadechandler.com [mailto:me@wadechandler.com] On Behalf Of
> Wade Chandler
> Sent: Saturday, September 24, 2016 8:04 PM
> To: general@incubator.apache.org
> Subject: Re: Preliminary NetBeans cost findings (was: [DISCUSS] Apache
> NetBeans Incubator Proposal)
> 
> First, I think we need to see the data you are referring to. Anecdotally the NB
> community seems to be growing. We are certainly competing with more
> projects such as VS Code and others in recent years. However, given reviews
> over the past many years of Java IDEs, NB has consistently been in the top 3.
> IntelliJ IDEA Ultimate is not an open source project by the way, so I suggest any
> comparisons to it, especially in the context of an organization such as Apache,
> is not relevant. Money being one thing, and everything else another, including
> OSS versus sort of OSS, I think it a fair question, but I hope not a subjective and
> biased one.
> 
> Has moving to Apache ever reversed trends which you are referring? For
> instance, does Apache champion it's own model over others? Why should a
> project move to the Apache way? Us in the NB community have pushed Oracle
> to move to a more open and community focused model for years. This
> sounded like it was about to happen, and many were excited to hear Apache,
> but I don't know what goal post this is, and if realistic, and if this email is to be
> viewed negatively or not.
> 
> It doesn't seem oriented towards analyzing statements of cost to be applied in
> support of other projects, or a way forward based on cost reduction or code
> sharing given the initial estimate, but instead focuses on a seemingly nebulous
> decline of NetBeans which is the first news I have seen of this.
> 
> Are there ways to cut the cost estimates? GoDaddy (surely others) has some
> nice plans with unlimited storage and bandwidth, and some rewrites of some
> systems with PHP, could make some things more viable. What about cost share
> across projects with similar needs? Do no other Apache projects have plugins
> or distribution needs? Other than build servers, what can't be consolidated?
> What about monetary donations to projects or specific Apache line items? Has
> there been any such talk?
> 
> How many other OSS Java IDEs are their? Seem only 2 at the Eclipse and
> NetBeans level. Having them both exist makes the entire ecosystem healthier
> in my opinion. It would be a shame to not have one of the real open source
> Java IDEs exist as an Apache project IMO.
> 
> Thanks
> 
> Wade
> 
> On Sep 24, 2016 7:16 PM, "Ross Gardler" <Ross.Gardler@microsoft.com>
> wrote:
> 
> > The ASF need to justify spending an extra $10k per year in this one
> > project at the expense of that $10k going to other projects.
> >
> > Don't make the request until the IPMC can present an argument that a
> > move of NetBeans to the ASF will reverse the decline in interest that
> > NetBeans is seeing.
> >
> > It may sound trivial, but we can support three "traditional" ASF
> > projects for NetBeans budget. As a charity we need to think carefully
> > about how we spend our money. A solid argument that this would reverse
> > the downward trend for NetBeans will go a long way to reassuring me
> > (as one member, but also as the person ultimately responsible for
> > paying such a budget request to the board).
> >
> > Ross
> >
> > ---
> > Twitter: @rgardler
> >
> > ________________________________
> > From: Ted Dunning <ted.dunning@gmail.com>
> > Sent: Saturday, September 24, 2016 4:04:34 PM
> > To: general@incubator.apache.org
> > Subject: Re: Preliminary NetBeans cost findings (was: [DISCUSS] Apache
> > NetBeans Incubator Proposal)
> >
> > Should this request come from IPMC? Seems like it should be at least a
> > coop request between infra (who get the budget and the operational
> > onus) and incubator (who cause the problem).
> >
> > Certainly the budget shouldn't come to the IPMC if approved.
> >
> > I will work with the board to determine the best form.
> >
> >
> > On Sat, Sep 24, 2016 at 7:57 PM, Chris Mattmann <mattmann@apache.org>
> > wrote:
> >
> > > Daniel this is great work. Thank you for outlining this. Wow!
> > >
> > > Chris
> > >
> > >
> > > On 9/24/16, 3:17 AM, "Daniel Gruno" <humbedooh@apache.org> wrote:
> > >
> > >     Hi folks,
> > >
> > >     I've been going over the requirements for NetBeans
> > > infrastructure,
> > it's
> > >     ballpark costs, bandwidth, machines needed and so forth, and the
> > cliff
> > >     notes are as follows:
> > >
> > >     - 40-50TB/month in traffic required (mostly downloads+plugins)
> > >     - 8-13 machines/VMS are required
> > >     - Ballpark hardware costs are between $3k and $10k per year,
> > depending
> > >       on how much we can move to existing infrastructure and how close we
> > >       come to the original setup. The most likely figure we are
> > > working with
> > >       is $4.9k, but we should be prepared for a larger cost, just in
> > case.
> > >     - The maintenance will be split between infra (downloads, web
> > > site,
> > CI,
> > >       new build machines) and the project (services, plugins,
> > statistics),
> > >       which will undoubtedly incur additional costs in terms of
> > > infra
> > time
> > >       spent on this, possibly to the tune of $10-20k in the initial
> > phase.
> > >
> > >     Certain services like the plugins hosting will rely on Legal
> > > giving
> > the
> > >     go-ahead for it, otherwise we'll have to find other people willing to
> > >     host this.
> > >
> > >     Other items like downloads may be offset by CDN providers
> > > offering their
> > >     assistance, but we should be prepared for this not being the
> > > case
> > from
> > >     the beginning, thus the 40-50TB/month. Likewise, some machine costs
> > >     may be offset by cloud providers offering services for free.
> > >
> > >     Thus, I would submit to the IPMC that they consider asking the
> > > board for
> > >     a budget of roughly $10k per year for the NetBeans project, as
> > > well
> > as
> > >     the additional time required of Infrastructure to implement this into
> > >     the existing ASF infra. As we may be able to pool resources and
> > utilize
> > >     the new hardware for multiple projects, the cost may go down in the
> > >     coming years, but this is the baseline I suggest we consider when
> > >     approving NetBeans as a new podling.
> > >
> > >     With regards,
> > >     Daniel.
> > >
> > >
> > >     ------------------------------------------------------------
> > ---------
> > >     To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> > >     For additional commands, e-mail:
> > > general-help@incubator.apache.org
> > >
> > >
> > >
> > >
> > >
> > >
> > > --------------------------------------------------------------------
> > > - To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
> > > For additional commands, e-mail: general-help@incubator.apache.org
> > >
> > >
> >

---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org
Mime
View raw message