Hi Bart,

Really, thank you for even trying to help at all.  You are remarkable.

The docker idea could help, but that would still involve someone with time to 
create it.  Side note, I know Richard has put some effort into the build and 
ensuring it can work in Intellij, so things might be better for you now.  Even 
with the build perfect, we'd still need someone with the time to provide 
development guidance and merge PRs.

I had an idea on the sponsoring concept.  We do have 34 committers that have 
been added over the last 24 years.  Maybe we get sponsor one of them to help 
onboard new contributors like yourself in their evening hours for a few weeks.  
We could use the Github Sponsor path to get them the funding and I'd match 
whatever you put up.  We put some formality on it, like specific dates where 
online sessions would happen (say one hour in the evening per week, for 
specific date range) and I can use my platform to advertise it.  Then we go 
through the list of people who have showed interest in helping in the last few 
years and try to get them to attend and see what magic we can make happen.

Then we'd have someone who knows the codebase a bit, can fix build issues, 
review and merge PRs, etc. and perhaps we can turn some hopeful contributors 
into committers.

The project has more or less "died" three times over the last 24 years and the 
challenge each time was getting enough people enabled to help others.  The 
previous times I basically did it myself with brute force and huge investments 
of time and it'd get things running again for a few years.  The last time was 
2007.  We now have Github Sponsors, Twitter and way more eyes on the project; 
advantages that weren't there before.

If it works we could maybe do it a few times and keep pumping up the project 
and related projects.  The real truth behind this is our dependent projects are 
in the same boat: CXF, Johnzon, OpenWebBeans, BatchEE, BVal, etc.  All the 
projects mentioned are actually the only other implementation of the spec in 
the industry and sadly, TomEE is the only Jakarta EE implementation that ships 
them.  All other vendors combined pool their resources on just the Eclipse 
implementations of those specs, which means they set a pace that doesn't 
reflect actual diversity and is why we are always behind; if you actually do 
what standards are for (implementation choice, innovation, competition) you 
have a disadvantage in this market as you're the only one doing it which is 
just backwards.  We could potentially have a big impact if we can get a pattern 
that works to create committers.

I've got a couple people in mind who might like to do such a thing.  Could be 
worth a try at least once.

Thoughts?


-David

> On Mar 20, 2024, at 3:38 AM, Bart van Leeuwen <bart_van_leeu...@netage.nl> 
> wrote:
> 
> Hi David,
> 
> thank you again for the extensive email, I see the 'deadlock' or 'race 
> condition' in supporting financially, and one could indeed argue that 
> might not be the way to go.
> 
> I would certainly be interested in having 'Tomee development for dummies' 
> session or something.
> I haven't looked at the development environment in a while so I might say 
> something that is already covered. 
> My first attempt at compiling Tomee stranded in the build environment, 
> would a docker image with a development environment be feasible?
> 
> 
> Met Vriendelijke Groet / With Kind Regards
> Bart van Leeuwen
> 
> 
> 
> From:   "David Blevins" <david.blev...@gmail.com>
> To:     users@tomee.apache.org
> Date:   20-03-2024 01:20
> Subject:        Re: Question: Can I donate to an Apache project? (was Re: 
> Will Tomee be discontinued ?)
> 
> 
> 
> Hi Bart,
> 
> If you have the ability to contribute at work we should make that plan A 
> and do our best with that despite the challenge it will be for everyone. 
> It has the most potential benefit to the project and you long term.
> 
> The trick with financial support is nothing short of enough to cover a 
> salary really helps.  My experience is no one has that kind of budget to 
> spare and get nothing but open source in return.  You usually need a few 
> supporters to cover one person.  They typically get something in return 
> (like 24x7 support) and that often means the person who gets brought on 
> now has a huge learning curve like you would have, but also a list of 
> tickets to solve before they can spend time on oss, so they still end up 
> with the "I have day job" problem.  Only now it's a day and night job 
> (24x7 support), so the evenings you had to contribute are now gone.  That 
> is pretty much the current situation with the original contributors like 
> myself.
> 
> Contributing time directly has its own challenges.  A big one is that once 
> a project gets so low in people with time, there is no one available to 
> help new potential contributors.  We have 34 committers, but we're lucky 
> to see 3.  It's a big project with dependences that are quite large 
> themselves and the time to teach all of that is nearly equivalent to the 
> time for the other person to learn it.  By the time people learn it, they 
> often move on to other jobs and often the value the project got is low for 
> the time spent.  All that said, if you or I had unlimited resources this 
> would still be the case.  We'd have a bunch of newly hired people who 
> don't know what to do and only a small number of (very overworked) people 
> to enable them.  It would still be hard for everyone.
> 
> The big win would be you could be a positive example: we use it, so we 
> contribute.  People need to see this.  We need to encourage people to 
> follow that example.
> 
> It is possible to start small and slowly work towards bigger things.  It 
> does take constant guidance and someone with time to help and that is the 
> big trick.  People frequently don't get too far and that's another 
> challenge  and a real opportunity cost if it doesn't work.  We can greatly 
> lower that risk if we can get a few people like you together and teach a 
> handful of people how to contribute at once.
> 
> Maybe if we can get a few others signed up, we can setup some kind of 
> semi-regular zoom meetings to help people find a way in.  You could lean 
> on each other as you go.  This greatly helps in feeling safe asking 
> "stupid" questions as you know others who need the answer.  Then there are 
> more people to help others get in and we start to get back on our feet 
> again.
> 
> Are there others in Bert's position who have employer support to 
> contribute but aren't sure where to start?
> 
> 
> -David
> 
> 
>> On Mar 19, 2024, at 3:38 AM, Bart van Leeuwen 
> <bart_van_leeu...@netage.nl> wrote:
>> 
>> Hi David,
>> 
>> Thank you for your extensive and personal email. It is good to hear the 
>> back story of these projects now and then.
>> 
>> I have to admit I'm in the same boat as Vicente, I would like to 
>> contribute, I can commit work time, but the sheer volume of the project 
> is 
>> intimidating at least and makes it hard to get started.
>> So I would be willing to support development financially, preferably in 
> a 
>> way that gets most of the money to the people who do the work, or the 
> ones 
>> that pay their paycheck.
>> 
>> That however is not solving the real problem, the lack of commiters to 
> the 
>> project. Ideally there would be a way for lesser gods, like myself, to 
>> take small bites of the project, instead of eating the whole elephant.
>> The question obviously is, is that even possible?
>> 
>> Met Vriendelijke Groet / With Kind Regards
>> Bart van Leeuwen
>> 
>> 
>> 
>> From:   "David Blevins" <david.blev...@gmail.com>
>> To:     users@tomee.apache.org
>> Date:   19-03-2024 03:20
>> Subject:        Question: Can I donate to an Apache project? (was Re: 
> Will 
>> Tomee be discontinued ?)
>> 
>> 
>> 
>>> On Mar 16, 2024, at 9:02 AM, Vicente Rossello <cocorosse...@gmail.com> 
>> wrote:
>>> 
>>> Hi,
>>> 
>>> I've tried a few times to do some contributions to the project, but 
>> testing
>>> the TCK or solving almost any issue is really hard, and very far from 
>> what
>>> I'm used to do in my daily work. And now I don't have much time... 
>> family
>>> and work consumes almost all my time.
>>> 
>>> I really find this project relevant in the jakarta EE and I would love 
>> to
>>> see it keep going. What I can do is to make some donations. I guess 
> that
>>> the donations should go to apache, my question is can I fund this 
>> specific
>>> project? Also I see that donations are tax deductible in the US, does
>>> anyone know if this is possible in Spain (or even any country in 
>> Europe)?
>> 
>> First, I just want to say on a personal level, I find your email 
> touching. 
>> Most people only ask what can I get and not what can I give.  The world 
>> needs more people like you.
>> 
>> Donations to Apache aren't used to fund development of Apache projects. 
>> The foundation in terms of being a corporation is actually incredibly 
>> small; less than 10 employees and contractors combined.  The funding 
>> Apache gets goes to that very tiny crew and covers infrastructure, 
> legal, 
>> the conferences Apache coordinates and some limited marketing.
>> 
>> Everything else including the board of directors are all volunteers.
>> 
>> What that means is there is no way for you to sponsor "the project", you 
> 
>> would have to single out individuals and sponsor them directly.  I've 
> used 
>> Github sponsors to sponsor a few of the people I saw contributing, such 
> as 
>> Daniel Dias, Richard Zowalla and Thomas Andraschko.  They take 10% and 
> and 
>> handle tax.
>> 
>> I agree with your perspective on not wanting to sacrifice family time 
> for 
>> open source.  Unfortunately it is the main source of contribution for 
> most 
>> Apache projects and the main reason people burn out and stop 
> contributing.
>> 
>> I used to encourage people to contribute in their spare time and did so 
>> myself.  TomEE 1.0 to 1.5 were created and shipped by people working in 
>> their spare time.  We would frequently use vacation time to hack on open 
> 
>> source together, cut releases, etc.  The 1.5 release was actually cut 
>> while Jean-Louis was in the hospital while his wife was giving birth to 
>> their second kid and I was on vacation helping.  On my side I ended up 
>> having to quit my job in order to get permission to work on TomEE in my 
>> spare time after having gotten in some hot water for taking a week off 
> to 
>> cut the 1.0.  I later learned Jonathan Gallimore had to apply similar 
>> pressure to his employer to get the permission to also work in his spare 
> 
>> time.
>> 
>> There was some occasional employer support.  Atos/Worldline was 
> supportive 
>> as they used OpenEJB and had a smart manager, Jean-Francois James, that 
>> saw benefit in allowing some contribution on company time when they had 
> a 
>> specific need (this is where Jean-Louis Monteiro, Romain Manni-Bucau 
>> worked).  IBM was very supportive of me in the 2005 - 2010 range when 
>> Geronimo was active as long as it benefit Geronimo and my contributions 
>> did not compete with Geronimo (which of course they did and that 
>> ultimately meant I had to work on my spare time most of the time).
>> 
>> That's the very delicate balance that built TomEE.
>> 
>> I no longer encourage individuals to sacrifice personal/family time to 
>> work on Open Source projects, I don't feel it is ethical anymore.  I 
> admit 
>> that I also do not find it ethical for you as an individual to sponsor 
>> other individuals.  It's that the majority of contribution comes from 
>> individuals contributing in their family time (not going to call it 
>> "spare"  time), while the majority of consumers are for profit 
> companies.
>> 
>> I can't advise people to use their remaining time after work to 
> contribute 
>> to Open Source.  This primarily benefits the employer using the software 
> 
>> and comes at the expense to your family.  Nor can I advise people to use 
> 
>> the money they earned working for their employer to sponsor an 
> individual 
>> contributor.  Yes, the contributor benefits, but there's no denying 
> you're 
>> essentially helping cover the cost of the open source software your 
>> employer uses and using money your family needs to do it.
>> 
>> When a company moves onto an open source project to save money and that 
>> project is only possible because of individual contributors, it is 
>> essentially the families of those contributors who enabled that savings. 
> 
>> Essentially cost has been shifted from the employer to employees and 
> their 
>> families.
>> 
>> Unless of course the companies and for-profit consumers also contribute. 
> 
>> Then I no longer have any issue.  Then it's as open source is mean to to 
> 
>> be: everyone who uses also contributes. 
>> 
>> Open source is like stone soup.  It's a shared cost model.  Everyone 
>> shares the cost by contributing a little and everyone eats.  Without 
> that, 
>> however, it isn't a beautiful story where everyone shares and everyone 
>> eats.  It becomes a story where the townfolk all give their last carrots 
> 
>> and potatoes to make soup for the wealthy.
>> 
>> None of this is because people are evil, just that open source is very 
> new 
>> and our relationship to it is very young.
>> 
>> The true benefit of open source is you can shift from high-cost 
>> proprietary software while not having to do everything in house either. 
>> You can share that cost with others, which will be a fraction of what 
> you 
>> all paid collectively.  The true benefit of an Apache open source 
> project 
>> is that as a contributor, your relationship to other contributors is 
> equal 
>> and fair and all matters are voted and have board oversight to correct 
> if 
>> things become unfair.
>> 
>> Those of you using TomEE at work should have a conversation with your 
>> manager so you can all talk and see what you can bring to the project. 
> If 
>> you are unsure how to have that conversation, we can talk about it.  I'm 
> 
>> ok to help people offline as long as you're willing to pay it forward 
> and 
>> help others with this like I helped you.
>> 
>> I don't recommend taking time/money away from family.
>> 
>> 
>> -David
>> 
>> 
>> 
>> 
>> 
> 
> 
> 
> 
> 

Reply via email to