This year, it blew up in everybody's faces. Whose fault is it? Defcon, for having unrealistic budgets? Contractors, for taking on an unrealistic project for the prestige?
IMO there's plenty of blame to go around.
My reaction was "only 60% over budget"? This is a low volume custom computer. The way Defcon pushes promotion and recognition, I do not think they paid full commercial price.
EE claims, they stayed in Budget while also claiming they did not receive the agreed upon payment.
DC claims, EE went 60% over budget
A run with 30k badges is not that small and eight months is not that short of a timeframe.
Somewhere I read about a $20 target price all in which I find realistic for PCB parts and assembly, but not if the cost include injection molding ($10k / mold and $1/pcs I guess) and especially not if it contains hardware and software development cost
If (like me) you hadn't seen this one, I think it is https://news.ycombinator.com/item?id=41172531
(Some of the top-level comments do indeed seem a bit oddly negative to me)
My priors align with the client having unreasonable expectations and then squabbling over the inconsequential bill. That is totally a thing that clients sometimes do.
When your project blows up, the professional thing to do is to resolve the problem with the client before billing another hour over the SOW. The common, crazy thing to do instead is optimistic invoicing: the client must share our priors, we're all reasonable people, so we'll just implicitly revise the SOW to match our learnings on this project and proceed, prioritizing what we believe would be a successful delivery of the project over everything else. That rarely works.
Serious consultancies routinely eat billable weeks of time in order to meet client success criteria and retain relationships.
Why reach for a stop work order if the whole thing is a rounding error. Entropic seems like they were able to finish, except that cost was an issue
My interpretation is that the project was at risk of not being delivered. No doubt Entropic had made and broken many promises leading to the stop work, and at some point DEF CON needs to take ownership of the project in order to ensure delivery.
> We were clear as early as our first conversation in January that the risk in trying to push to mass production of this size and on this timeline was immense, even advocating for a DEFCON 2025 release of this particular badge. DEFCON’s Badge Team remained confident that they could meet and mitigate this risk.
> Once a month, we billed for our work and submitted an updated estimated per badge final cost - committing as costs built to discount our work as necessary in order to hit DEFCON’s per unit cost targets.
> In June, after 5 months of late night work, badges were fully designed, prototypes were working, and mass production was ongoing with the manufacturers we contracted on behalf of DEFCON. We billed DEFCON for our most recent work, discounting our labor by 25% in order to meet the agreed upon targets. Unfortunately, we were instead met with a work stoppage request and informed we would no longer be paid for services already rendered.
https://www.entropicengineering.com/defcon-32-statement
It feels to me like DEFCON is relying on being able to say "well, we're DEFCON" when defending themselves, and people like you are just blindly trusting their word. How many times have big organizations like this screwed their suppliers? And yet DEFCON is "clearly" innocent? You must be joking.
At least wait until we get a better picture of everything before deciding on a judgment of any of the parties involved. It's going to take time before we find out what actually happened.
This doesn't really add up to me. Ultimately we'll never know for sure. Personally I'm inclined to believe Entropic more on this one (though I'm sure neither party is blame-free, regardless of where the bulk of fault lies), but I can see why you'd be more inclined to trust DEFCON's response.
(I just don't consider "DEFCON made a falsifiable claim and Entropic didn't" to be strong evidence either way.)
It’s perfectly possible for Entropic to be calculating to a $100k total cost (for the board), while Defcon is calculating the cost for everything (case/lanyard, etc.), and gets to $160k for the entire project, then says they’re 60% over budget.
I imagine Entropic was discounting their bills to say within the 100K, while Defcon already felt they were over.
(Numbers made up)
A fairly simple mold costs 5-10k, a complex mold 10-20k. Since this is a limited production run, I would argue that they didn't need the highest tolerances.
Thus my guess would be ~$50k for the molds and ~$5 for the parts.
https://old.reddit.com/r/Defcon/comments/1en7jxz/
EE: It will cost us $300K to make those
Def CON: Ok too bad, see you around.
EE: No no, what I meant to say is 'it would cost us $300K, but we will gladly do for $200K, we will "discount" our work.
Def CON: Dont care about your internal accounting, can you do it for $200K or not?
EE: Where do I sign up?
...time passes...
EE: keeps trying to send invoices totaling $300K
Where are you getting that, though? EE says:
> Once a month, we billed for our work and submitted an updated estimated per badge final cost - committing as costs built to discount our work as necessary in order to hit DEFCON’s per unit cost targets.
And:
> We billed DEFCON for our most recent work, discounting our labor by 25% in order to meet the agreed upon targets.
So it sounds like they were billing with an expected total of $200k, not anything higher. Perhaps they were listing in their invoice all the actual costs, but then had a "discount" line item to bring the cost back down, but if so... so what?
Assuming this is true it’s simple: you walk away. If you’re being contracted to do something you don’t think you can do, you don’t sign the contract. Anything else is a recipe for pain.
Arguably it still is amazing publicity, at least for the people who will end up believing Entropic's version of events. Maybe not a good look when it comes to people who side with DEFCON on this.
And that’s not even mentioning the Easter Egg! Good grief!
I read about the $20 target price, which I find quite hard to reach, but that is something you know before you agree to deliver X by date Y costing Z
My alarm bells go off loud when people invoke tropes indirectly and lazily, hoping it'll influence my perception of a situation
DEFCON isn't the pointy-haired boss stereotype that needs you to deliver exactly $X, this quarter, with duct tape and glue. They know tech and wouldn't have talked a team saying they can't do it until 2025 into 2024.
Well, what if they really wanted it for publicity?
DEFCON had 0 stake in a new Raspberry Pi release, and Entropic self-reports they were the ones with early access to an unreleased product and decided it was the right vehicle. [2]
[1] We were clear as early as our first conversation in January that the risk in trying to push to mass production of this size and on this timeline was immense, even advocating for a DEFCON 2025 release of this particular badge. DEFCON’s Badge Team remained confident that they could meet and mitigate this risk.
[2] The specifics of what they requested in January were extremely difficult / almost impossible, but we had been working with Raspberry Pi as a Design Partner and had early access to the unreleased Raspberry Pi RP 2350, a chip that would enable exactly the kind of device DEFCON was requesting.