Something I've talked about recently with a few folks is that TPF has "too much" money. Specifically, TPF has reported an increasing balance at the end of its last couple fiscal year. Nonprofits are not supposed to consistently make a profit (no kidding), and this has a bad smell. I don't think there's anything fishy going on, mind you, it just doesn't look right.
Part of the problem comes from a large $35k grant paid to TPF by NLNet for Parrot. TPF got the money from NLNet back in 2005 (IIRC) and it took a while before that money started flowing out of TPF.
In discussions with a couple folks in TPF, they've said that they'd really like to spend the money, but they don't have avenues to do so. As most folks know, conferences and workshops are generally profitable for the organization (Frozen Perl netted around $1k) so that's not a way to spend money.
Then there's the grants program. I'm on the grants committee list and I've seen all the grants that've come through for the past couple years. The problem with the grants system is that there's not nearly enough grant applications coming in. Then of the applications that do come in many are simply unrealistic. Either they are too vague, too niche, or too hard, and so don't get approved.
I've been thinking about this recently and I think that this failure mode is basically built-in to the current grants system. First, TPF has a sort of unwritten rule that it won't fund travel, because there are so many Perl folks who'd like to go to so many Perl events that it would be hard to handle. I think there's definitely some truth to this, though I could see a use in funding travel specifically for project hacking (which TPF has done from time to time).
Another unwritten rule is that individual grants will not be more than $10k. This also makes sense, as $10k is a lot of money to give to one proposal. So what's the problem?
I for am unlikely to ever apply for a grant under the current scheme (ignoring the fact that I'm ineligible because I'm a grant manager), even though I could probably come up with something TPF would be willing to fund.
The problem is that I just can't see how a grant could be an incentive for me. I already put a fair bit of my time into FS/OSS projects just because I want to do so. I'd love to put in much more time, but I have things like a mortgage and family to consider. Realistically, the only way I'm going to put more time into my projects is to take a sabbatical from work, or at least work part-time for a while.
But if you look at the work/money ratio for past grants there's no way that could happen. Even if I aimed for something like %60-80 of my current FT income, a grant could not come close. It'd be more like 20-30%. So the grant provides no incentive for me. I suppose I could still apply for one anyway, but I don't feel right about that because it would just be funding work I would do anyway!
I'm sure many other developers have the same issue. So what is the point of the grants program, if not to make work happen that wouldn't otherwise happen? The acknowledgement of one's work is nice, but I already get that in many other ways, and if I'm looking for acknowledgement in the form of cash, I'd expect a heck of a lot more than a couple thousand dollars (Amazon and others, contact me privately for an address to which you can mail a big fat check).
Personally, I'd be perfectly happy to see TPF fund three months of full time work on any Parrot, Perl 5, or some other project likely to be of great benefit. Of course, the number of people eligible for this sort of grant are few in number, but it might achieve more in the long run.
Re:Also, is there the issue of "hogging" grants?
autarch on 2008-02-22T04:07:46
But I don't see how this even remotely leaves TPF in a situation of having "too much money".By "too much money" I simply mean that TPF has been increasing it's net bank balance over the last few fiscal years (according to its filings with the IRS).
I don't mean they have too much, as in more it should, or more than it deserves, simply that it is taking in more money than it spends, but as a nonprofit it should aim towards breaking even, which for a while will mean spending down some of its outstanding balance.I would be overjoyed if TPF had Mozilla's 50m river of hold and could hire 50 or 100 full time P5P/CPAN/Parrot/P6 people.Me too! OTOH, it did have nearly $100k in the bank as of its FY2006 filing.
Re:As a fellow committee member
autarch on 2008-02-22T04:13:26
The grants committee has done a good job of picking grant recipients with the submissions it receives. As you imply, there's a lot of not-so-good submissions which don't get funded.
I think that there could be more submissions if the nature of the grants were different. Specifically, I'd like to see grants aimed at funding a sabbatical from work for the recipients. Right now, the grants are more like little bonuses for doing a lot of work, as opposed to enough pay to live on while doing a lot of work.
Of course, if TPF were to regularly do those sorts of grants, there'd be plenty of flack over that too.
I'm less keen on the random award type of idea. I'm not convinced that this sort of thing would do much to further TPF's mission.
Re:As a fellow committee member
Alias on 2008-02-22T07:45:01
I'd happily apply for something like that...
I'm currently trying to arrange a sabbatical to do the entire June to August conference season, with city visits/talks in between, and hacking on various useful and interesting stuff as I go.
It means a pricey around the world ticket, no income for 3 months, and relying on being "hosted" in the various cities in order to try and keep expenses down.
If any of those fall through and I start incurring hotel bills, or extra flights, or what have you, it gets ugly fast.
And it's a massive undertaking I'm pretty sure I will only be able to pull off once.
If something was in place to help support that sort of trip, it would be awesome.
But currently it is outside the scope of the grant system, which is highly code-based and goal oriented.
I know TPF gave a bunch of money to Damian a few years ago, and I'm not sure how good value money they considered they got in the end...
Since the grants are generally relatively small, doing a grant of that sort again (the $50k annual thing) seems hard to justify.
But if there was an option on something in the 5-10 range that I could apply for to just hack on useful (to other Perl coders) thing, travel the conferences and mongers, and generally do Perl stuff for 3 months, I'd apply in an instant.Re:As a fellow committee member
Alias on 2008-02-22T07:45:43
But of course, then we get down to favouritism again... since it would be my third grant...Re:As a fellow committee member
btilly on 2008-02-22T08:27:29
I was suggesting the awards as a way of reducing he bank balance and not of doing great things for the mission. (Though I do think it would likely help the mission.)
For several months now, I've been tempted to kite a grant proposal for Thread::Sociable, which (sadly) languishes while I focus on revenue generators. I've written up the proposal and everything.
So why haven't I submitted it ?
My sense is (please correct me if I'm wrong):
So it sits unfinished. And everytime I see a threads question popup on perlmonks or elsewhere, I have to shake my head and bite my tongue (I've already pimped too much).
(If any org out there has an interest in sponsoring faster concurrency, STM, and a foundation for transactional distributed memory for Perl5, please don't be a stranger!)
Re:Timely post
btilly on 2008-02-22T08:48:54
Actually the Perl 6 comment is exactly backwards. Most of the grant managers have far more Perl 5 experience than Perl 6, and it would be easier for us to evaluate Perl 5 proposals. But we get more good Perl 6 proposals than Perl 5 proposals. Indeed two have just voluntarily stepped down because they felt uncomfortable deciding so many Perl 6 grants when they didn't know that much about what was happening Perl 6.
Threads related depends on the pitch that is made. We're aware that there are people trying to use threads, and we're aware that Perl's threading implementation leaves a lot to be desired. Having read your module description, it is something that I, at least, could vote for. Particularly if it was backed up with evidence that it was doable and you're the right person to do it. (Good evidence could include testimonials from someone who knows threads, a description of your threading background, or a description of what the part you've already built does.)
The dollar amount is a real concern. However $500-$3000 isn't set in stone. Somewhat larger dollar amounts can be approved if a good enough case is made for them. For instance it would help if your proposal broke down how you intended to spend the money. For instance $x towards the following equipment, etc. But whatever amount you select it is very important that your proposal says how much you want. We're not in a position to bargain about the figure, we just vote yes or no on what was requested. So if the project is only doable with $6000 then be up front about that. The worst that can happen is that you'll get a "no", and you're no worse off than before. (Actually you're better off because you'll have a detailed proposal drawn up to submit elsewhere, and you'll get feedback on how others reacted to it.)
In general if this is something you would do regardless, it can't hurt to submit it to TPF. If you can come up with someone who will offer you support if you can come up with support from someone else, submit it to TPF. If you would do it if you got a little help (for instance to buy equipment), then by all means submit it. But if you're looking for a reasonable wage while you solve this problem, TPF won't be of much assistance.
As discussed, the grants aren't much of a carrot, but they're an excellent stick. Let's see what happens inside a developer's head both with and without grants.
I don't feel particularly guilty when one of my own, unfunded volunteer projects doesn't receive the attention it deserves. Everyone knows that volunteers never have enough time for everyone, and I can hardly be blamed for taking a break now and again, especially if that's to earn a crust.
Funding, on the other hand, means the whole world knows that I'm getting paid for my project. I'm expected to blog about it. I'm expected to give talks at conferences. I'm accountable, and I don't have a choice but to work on the project, because now my reputation now hangs in the balance. Screwing up with grant money means I face the possibility of public humiliation and ridicule. That means that grants are an excellent motivator.
The trick, therefore, is to identify people who will respond well to fear, and "motivate" them with a grant, even if they haven't actually put in a proposal. The risk (as Adam has mentioned) is that this may be seen as favouritism; I don't have a good solution to that.
Re:Grants and motivation
autarch on 2008-02-22T15:40:03
I think this is true, but it doesn't address the problem I see of not getting enough good grant proposals! I for one am not going to be motivated to write a grant proposal because it's a good stick. Maybe we need more skilled hackers who are also masochists in the Perl community
Funding, on the other hand, means the whole world knows that I'm getting paid for my project. I'm expected to blog about it. I'm expected to give talks at conferences. I'm accountable, and I don't have a choice but to work on the project, because now my reputation now hangs in the balance. Screwing up with grant money means I face the possibility of public humiliation and ridicule. That means that grants are an excellent motivator.;)
Maybe TPF should do some outreach to the BDSM community.Re:Grants and motivation
pjf on 2008-02-22T22:47:25
Ah! I seem to have missed properly communicating my point. Rather than wait exclusively for grant proposals to arrive, we can find people or projects that are doing good things, and motivate them to continue that excellent work. We take the stick to them, as it were.
This may take the form of introducing a new class of grants (the TPF motivation grants), which have the same expectations as a standard grant (you have to deliver the goods), but are instead offered by the TPF to particular individuals or groups, rather than requring a grant proposal to be submitted.
Of course, that potentially pushes even more work onto our grant committee in terms of identifying worthy projects, and can potentially be seen as controversial if accusations of favouritism are made. The half-way mark would be to still identify projects or ideas that may benefit from funding, and informally encourage them to submit for a grant. I expect (or at least hope) that this happens already.
On this note, I'm regularly speaking to groups of dozens to hundreds of developers at both training courses and conferences. I'm very happy to advertise the TPF grant process as part of this, especcially if there are starting materials I can work from.
All the best,
PaulMy posts to use.perl in this thread are being written entirely on trains, planes, and trams, where I have no Internet connectivity. I apologise for any lack of situational awareness this may cause.
Re:Grants and motivation
autarch on 2008-02-22T22:55:58
This may take the form of introducing a new class of grants (the TPF motivation grants), which have the same expectations as a standard grant (you have to deliver the goods), but are instead offered by the TPF to particular individuals or groups, rather than requring a grant proposal to be submitted.I believe the grants committee is already doing exactly this, and I think that's a good thing.
But the recipient may not want to accept the grant because it just adds responsibility without making the work easier to do. This is the exact same problem that would prevent them from submitting a grant in the first place.
I'll come back to the example of myself. I would be a good candidate for getting a grant (useful modules, well known in community, good track record, showers daily). I don't want one. It's not enough money for the hassle, and it would just cause me more stress to take the grant. That would suck the fun out of doing the work.
That actually brings up another question, which is whether or not having money involved (even enough for a sabbatical) is actually an incentive. Maybe it just ruins the fun for people. OTOH, I'm sure there are people out there who'd rather work on Parrot or Perl 5 full time for a few months than whatever other work they might do.
Re:Grants and motivation
pjf on 2008-02-22T23:33:33
That actually brings up another question, which is whether or not having money involved (even enough for a sabbatical) is actually an incentive. Maybe it just ruins the fun for people.
I suspect the obvious solution here is to say, "We would like to offer you a grant (if you'd like it), or an award/certificate/recognition/XP for your work. A listing in the TPF Hall of Fame has considerable cool-value, without the problems associated with money.
If desired, this could even be changed into a game; people receive points for completing TPF tasks, and community members can track their progress with a high score table. Kwalitee turned CPAN into a game, and I think it had a noticeable improvement on CPAN quality. Perl is fun, after all.
;) Paul
Re:Grants and motivation
nicholas on 2008-02-24T09:45:58
But the recipient may not want to accept the grant because it just adds responsibility without making the work easier to do. This is the exact same problem that would prevent them from submitting a grant in the first place.
I'll come back to the example of myself. I would be a good candidate for getting a grant (useful modules, well known in community, good track record, showers daily). I don't want one. It's not enough money for the hassle, and it would just cause me more stress to take the grant. That would suck the fun out of doing the work.
<aol>Me too!</aol>This is exactly what I found when doing Improving Perl 5
Also, things I found which may be personal to me:
- I don't like working at home
- I don't like working on something alone (which, as the only "full" time person is effectively what it was - I had questions no-one else had time to dig into to answer). If one is in a full time job with colleagues, one is likely to have people one can ask.
So, whilst I guess I am about to be looking for a job, which would seem to make me an appropriate person to apply, I know that it's not for me.
Re:Grants and motivation
autarch on 2008-02-24T14:40:54
I should say that if I could get a grant sufficient to substitute for real work for a time, I'd probably like that. I do like working from home, and I don't mind working alone for a while, though it's not always ideal.
My big reason for not wanting a grant as they stand is that it would provide pressure to do work without providing any extra time for that work.Re:Grants and motivation
chromatic on 2008-03-01T00:22:07
My big reason for not wanting a grant as they stand is that it would provide pressure to do work without providing any extra time for that work.I hadn't realized I felt this way until I read it. I agree.
Re:Grants and motivation
zby on 2008-04-30T20:46:01
You could have a look at Jelly. Looks like a solution to both problems.
Has there ever been consideration of posting bounties on specific Perl TODO list items ?
I suppose there are logistic issues to handle when 2 contributors show up with different solutions to the same problem. But at least it moves much of the bureaucratic burden from the developer to the TPF.
Frankly, as currently worded, the TPF grant application guidelines tend to discourage, rather than encourage, applicants.
What I see when I consider the guidelines:
Spend a lot of time on your proposal. We'll probably throw it out anyway, but you need to spend a lot of time on it. And make sure its a big important project, cuz we've got to make sure we spend money on important things. But not too big, cuz we don't really think you can finish it. And if you manage to finish, and blog it all over, and travel around making presentations, we'll give you a few hundred bucks.
Yes, I realize thats not how its intended...but thats how it reads to me, and I suspect many other potential applicants.
Re:Bounties vs. Grants ?
jmcada on 2008-02-23T01:59:11
bounties++ # they take the guesswork and challenge out for granteesRe:Bounties vs. Grants ?
nicholas on 2008-05-03T14:40:21
Has there ever been consideration of posting bounties on specific Perl TODO list items ?Yes. Vienna.pm is intending to offer bounties on TODOs, as part of how it distributes the surplus from YAPC::EU 2007. domm posted the announcement here, mailed it to perl5-porters and has probably publicised it elsewhere. To date, zero comments, and the only TODOs submitted so far are two from me for the Perl 5 core.
Why is there no interest?
Re:Bounties vs. Grants ?
autarch on 2008-05-03T19:21:24
I think bounties are worse than grants in terms of motivating people.
You don't get the money til after you do the work, and you don't even get to participate in determining how much money you get for that work.
I think when the bounties are really large (think X-Prize) they can be motivating, but I doubt they work well at the small scale.
It seems like it all comes down to money.
Re:Bounties vs. Grants ?
nicholas on 2008-05-04T18:31:33
I think bounties are worse than grants in terms of motivating people.In the general case, I'd agree with you completely for the reasons you give - I certainly don't see small bounties as a viable way to sustain regular interest and make steady progress.
My hope with them, is that for tasks that can be done in an evening or so, they'd be a teaser to draw in new people, at least to do that task, people who might not otherwise have considered contributing. What is depressing me is the level of interest from project maintainers, who as I see it have little to lose by adding bounties paid for by someone else on existing bugs in their projects. Does it really cost that much effort to identify tasks to add? If they never get done, the loss is the cost of adding the task. If they get done, woohoo! Free work.
Re:Balance from year to year
autarch on 2008-02-22T21:12:40
But keeping right around that balance should be the goal, no? I don't think you're suggesting we spend down to zero each year. But you don't like the 100k number. So I guess we should have something in between.It's not the specific number that concerns me, it's seeing a year-to-year increase in the bank balance without a clear sense of why that's happening. That's all. Maybe 100k is the right amount of cushion, in which case that's great, TPF has reached that goal, and should make sure that in the future it spends enough to keep the balance right around there.
Personally, I suspect that's a bit high, and that a balance of maybe $30-50k is probably best. That should be more than enough to cover down payments on a couple conferences, and gives us $50k to go to town on;) We do spend money on things other than grants, but we haven't done a good job of explaining to people where it goes. We're working on that so people know better what they are supporting when they support TPF.I think this is the key. I wouldn't really worry about any bank balance, regardless of amount, as long as it was clear why that balance should be maintained.
It seems like every time I think about things I dislike about TPF, it mostly comes back to communication issues.
The reason for my journal entry wasn't to point the finger and say "TPF is hogging up our money", it was really to raise the question of what do we do with the money, and if we are accumulating more than is needed in the bank, can we think of other ways to spend it that would further TPF's mission and be acceptable to the donor base. I deliberately phrase this as "we", despite not being a TPF officer, because I don't expect the TPF steering committee to have all the answers.
Personally, I like the idea of sabbatical-funded grants, mostly cause it's the only way I can see myself applying for one, but I'm sure people can come up with lots of other interesting ideas too.
Re:Balance from year to year
Alias on 2008-03-01T11:41:02
I wouldn't worry too much about the amount of money in the bank.
At the very worst, this money is earning extra interest (it is in a cash-generating fund right?) that strengthens the position of the TPF, because it's created more of the all-important cash flow.
If I was Mozilla, I'd be sitting 50 or 100m in the bank generating the operating cash.
Further, with that money available, at least you HAVE the option of spending it if something comes up down the track that deserves it.
If you feel too much pressure to spend it now, it won't be available later.
Each 50k in the bank means one more grant funded per year.
That can't be a bad thing.Re:Balance from year to year
autarch on 2008-03-01T15:57:29
I wouldn't worry too much about the amount of money in the bank.I wouldn't care much if TPF didn't do things like solicit people to "rush the stage" at YAPC to donate money, and things like that. Basically, why is an organization that is accumulating profits year to year begging for money in that manner?
Mostly, it's a question of transparency, like I've already said. If I knew there was a long-term plan that required $200k in the bank, then that'd be fine. As it is, there is this weird pressure to donate at YAPC and similar events, without any reason why TPF needs money right now.
Also, more and more, I'm convinced that TPF should be focused primarily on seeking large donations from businesses. As an individual developer, my main contribution to Perl's health should be through using it and improving it, not cash donations. And of course, individuals also do things like organize YAPCs
and workshops.If I was Mozilla, I'd be sitting 50 or 100m in the bank generating the operating cash.AFAICT, most of Mozilla's income goes to the Mozilla Corporation, which is a real business, and can act like one. The money that Mozilla has is from business activity, not individual donations. When your funding comes from individual donations, that also comes with an obligation to those donors.Each 50k in the bank means one more grant funded per year.If that's the plan for the money, great, say so. I wouldn't object to TPF trying to accumulate an endowment for the sake of funding things off the interest.
That can't be a bad thing.
good idea. TPF should advertise
Re:sabbatical work..
autarch on 2008-02-22T21:16:35
since you are the grant manager, can't you do something about it?I am not the grant manager, I am a grant manager for the NLNet-funded Parrot work. That mostly means I pay attention to Parrot's progress and authorize the release of funds as milestones are met.
Note that this money has already been earmarked for Parrot. I do not have any say on additional TPF spending on Parrot or any other project. That is up to the TPF steering committee and grants committee. The grants committee reviews grants submitted according to the ongoing grants program. The steering committee runs the organization as a whole, and could authorize other sorts of spending.
However, I know many of them are reading this discussion, so it's getting heard by the right people.
Re:Submit a proposal
zby on 2008-03-12T21:51:37
Unfortunately I am a techie too - but I'll keep that in mind and I'll look around.Re:Submit a proposal
zby on 2008-03-21T08:03:09
Another idea could be to sponsor web hosting. It is easy these days to have code repository and a static web site for free - but if you want to showcase an application or (as was the case of CSS testing application for HTML::FormFu) when you need a non standard collaborative application then you have to buy some hosting.
