Projections based on assumptions, incomplete knowledge, misunderstandings, and ignorance. Mostly bad assumptions, in fact.
Just to make my position clear: I’ve been on the PSF Board of Directors ~9 years since its creation, including 2017-2019 and 2020-2023. I was the interim General Manager (unpaid! So don’t worry about that expenditure!) between Ewa’s departure as Executive Director in 2021 and the Board’s hiring of Deb in 2022. I’ve been a PyCon US volunteer for many, many years. I know a lot of the things I’m explaining here first-hand, but I don’t just rely on my personal knowledge here. I don’t have special insights in what the Board or the staff have been doing over the last two years. These things are knowable and discoverable for anyone.
Wow, that was a very quick departure of “only facts”. Sure, expenditures are speeding up, because the PSF is spending money. That’s what it’s for. The last couple of years the PSF has gotten grants for specific things, which it spends on those things. Things like the Developers-in-Residence and the PyPI Engineers. Those expenditures do not come from general funds, and it’s not simply a matter of not spending the money. The funds are earmarked, and can’t just be used for something else. Moreover, the cost of these programs is balanced with the sponsorship income from those programs. Lowering the cost would lower the sponsorship. You can’t look at a single number below the line and extrapolate.
You claim to be concerned because the PSF set a firm limit on outgoing money (the PSF grants “drying up”) – even though the pause on grants is because the PSF spent the budgeted amount – and this leads you to conclude it’s “bleeding money” and “likely to default”? The PSF’s sponsorship income is always uncertain, because it’s largely reliant on corporate sponsors and some of those are surprisingly unwilling to fund the ecosystem and community.
But the PSF is in a position to reasonably balance its expenses to compensate, which it is doing by setting a budget for the grants program. Other programs are directly reliant on corporate sponsorship (like the Developers-in-Residence), and those programs will only exist as long as (corporate) sponsorship is available. The financial and operational stewardship is done with care and deliberation, so that if it turns out money is no longer available, the operations can be ramped down responsibly.
Let’s dig into those numbers, shall we!
The PSF has always been frugal with salaries, and it’s long had the policy of providing solid benefits and a pleasant work environment rather than paying top dollar. It’s also largely relied on volunteer efforts for a lot of the important things. It’s always been hiring staff with great care and deliberation, only taking on new people when it’s absolutely necessary. New people being hired over the last couple of years is because the work has grown significantly over the last ten years.
Despite being frugal, the PSF does have to pay fair wages, even if only to be able to attract and retain people. And the PSF has some of the best people working for it despite not offering top salaries. With salaries (salaries, not total compensation!) in tech jobs going as high as $400k and higher, pardon me for not clutching my pearls at $170k. The same goes for the other “high” salaries: they may seem high for Europe or other markets, but they are definitely not high in the markets the PSF is hiring in. You get what you pay for.
PyCon US costs a lot of money, yes. Large venues cost a lot. The services in large venues cost a lot. Economy of scale does not apply to large events, as anyone who has tried to grow a small-to-medium-sized event will tell you. But rest assured! PyCon US did not lose $1.8M in 2023 or $2.5M in 2024. Those were the costs, not the losses. The PyCon staff (who are mostly PSF staff) are always trying to manage the cost, considering what trade-offs are possible without negatively affecting the conference experience too much. It’s very difficult! Things like whether there should be an information booth, or breakfast, or whether the coffee break should be a half hour or an hour, or whether coffee should be available all through the day. (The difference in cost is staggering.) The only realistic options for significantly reducing PyCon US’s costs would be to scale it down significantly, or cut significant portions of the event. Those options would also significantly reduce sponsorship income, of course, even if the ticket price wouldn’t be changed. And even though I think those are all pretty bad options, those things are still carefully considered and weighed.
As for “costly consulting contracts”: Altitude/C are the A/V folks PyCon US has used most years since PyCon US was in Montreal (2014, I believe, although I think the company had a different name at the time). I say “most years” because the contract is always put up for bids each year. For 2022, the first in-person year after the start of the COVID-19 pandemic, Altitude/C wasn’t an option (I believe it was pandemic concerns, but I don’t remember exactly how it played out), and the A/V contract went to the venue’s in-house supplier. As I recall it ended up significantly costlier, but they were certainly significantly less flexible, and the quality was significantly worse. PSF staff had to touch up and re-edit so many videos, which is not normally their work or their core competency. People eager for 2022’s videos may remember how long it took for them to become available on YouTube. Every penny spent on Altitude/C has been worth it, for the community and for PyCon US’s financial situation.
You can’t predict trends or even determine the current pace of losses based on two data points from last year. But rest assured, the PSF has already made it clear that they are working on this. For example, by holding the PSF Grants WG to the budget limits they set.
It’s not minuscule, it is in fact a significant part of the PSF’s general funds spending. Funds earmarked for specific projects cannot be spent on grants, other projects, or general expenditures. Pay close attention to the term “donor restrictions” in the annual report.
The main driver are expenditures that are matched with income – restricted funds. Since 2020, the PSF has hired five highly technical, well-regarded professionals using targeted donations (the Developers-in-Residence, the Security Developer-in-Residence and the PyPI Safety & Security Engineer.) The PSF has also run a variety of projects to produce improvements to PyPI with grants from different institutions. Disregarding for the moment what effect trying to skimp on people would’ve had, saving money on these people or projects would mean less grant and sponsorship income.
The incorrect, blind, and willfully ignorant assumption that “number goes up, therefor they get paid too much”, is frankly insulting.
Just because you didn’t get the information you wanted over the Labor Day weekend doesn’t mean there’s anything here that the IRS would be concerned about. The PSF has been independently audited for the last couple of years, which is a requirement for government grants. I haven’t been on the Board for a few years, but the audit reviews I’ve been in made it clear the PSF’s record at the time was squeaky clean.
As I recall, in 2022 the PSF had 8 members of staff, plus our first Developer-in-Residence as contractor. Deb started some time in April that year.
As was already pointed out, Deb’s salary didn’t increase by 50%, because she wasn’t paid for a full year’s work in 2022. Nor did other salaries grow nearly as much as you imagine. Instead, the PSF hired more people. I believe over 2023 the PSF hired the new Communications Manager role, as well as the Security Developer in Residence and the PyPI Safety & Security Engineer. I believe the PSF hired one or two more roles in 2022, but I don’t remember the details. (Also two new Developers-in-Residence, but as I recall they started early/late December, so they won’t have made much of a dent in the number you are so concerned about.) Two of those roles, I will point out, were funded by restricted donations. Total salaries only going up by 40% given those new roles is, frankly, quite frugal.
That is correct. All three Developers-in-Residence are contractors, because they are not US citizens or residents. The PSF cannot hire them as employees. That means they do not get the secondary benefits that employees get, and are compensated accordingly. If you have concerns about the salary ranges and selection process for Developers-in-Residence, that’s actually been the Steering Council’s authority, since it comes from funds allocated specifically to CPython.
These assumptions are so laughable I don’t even know where to begin. Linear extrapolation is absurd, even if the PSF staff and Board weren’t actively paying attention. None of this is even remotely plausible. How about instead of assuming ridiculous things, you wait for actual numbers? Sure, it may take more than a weekend, even an extra long holiday weekend, but it’s worth the wait.
I hope expenditures aren’t slowing! The PSF is spending money on things that need money spent on them. It would be sad if the PSF had to stop. But I do know the PSF will stop if the money isn’t going to be available, because as I pointed out, a lot of the expenditures are from targeted donations, which are locked down before spending the money.
Salaries aren’t rising any more than market demands (and I think significantly less, actually). PyCon US expenses are constantly re-evaluated (for example, PyCon US dropped the ‘online’ component in 2025. In 2026 it’ll have fewer sprint days, and it might find an external venue for those in the future). The Altitude/C contract isn’t a new expenditures, and just part of PyCon US expenses – and one of the best ways PyCon US spends money; just ask anyone who has dealt with A/V at PyCon US, or was waiting for PyCon US videos in 2022.
I hope that alleviates any concerns readers of your post may have had.