How to Survive - without reducing SNO compensation

@Ambifacient

You spent a lot of words accusing my post of being long, overconfident and “numerically decorated.”

Yet after all the phrases about “spreadsheet fan fiction,” “deadline cosplay,” a “product fever dream” and a “technically literate ultimatum,” you provide not one alternative customer price, one alternative payout schedule, one operating-cost ceiling, one financing requirement or one prioritized product plan.

Your central criticism is that a public proposal does not contain private data that only Storj possesses.

That limitation was stated openly in the post. It is not the discovery you present it as, and it does not invalidate the calculations that can be made from the filed numbers.

The title is How to survive without reducing SNO compensation.

So yes, the plan starts with that condition. It is neither buried nor disguised. It is literally in the title, the opening paragraph and the conclusion.

The purpose is to answer a specific question:

Can Storj become viable without reducing SNO compensation?

Every serious plan has constraints. Service continuity is a constraint. Legal obligations are constraints. Customer contracts are constraints. The availability of capital is a constraint. The minimum price at which a critical supplier base remains available can also be a constraint.

A restructuring does not require management to cut every cost category merely to prove its neutrality.

It requires management to determine which changes produce the best net result without destroying the business it is trying to preserve.

You use the phrase “metaphysically exempt” as though I claimed a law of nature. I did not. I stated a commercial position:

I will not support or contribute to a restructuring that reduces SNO compensation.

Creditors state the treatment they will accept. Investors state the return and protection they require. Employees decide whether compensation is sufficient to remain. Customers decide what price and risk they will accept.

SNOs are also allowed to state their conditions.

Correct—and insolvency does not prove that the supplier price is excessive.

The court filings identify the failure of the enterprise-sales expansion, the unsuccessful Object Mount investment, disappointing results from Petagene and Valdi, and payroll and technology costs greatly exceeding income.

They do not identify SNO compensation as the cause of the filing.

That does not make SNO rates immune from discussion. It does place the burden on anyone proposing a reduction to show that it creates a positive net result.

“Supplier prices are renegotiated constantly” is not an analysis. Suppliers also reject new terms, withdraw capacity, reduce quality or leave the relationship entirely.

The buyer being insolvent does not magically make the buyer’s preferred supplier price sustainable.

The comparison exists to show scale.

A 20% storage-rate cut would generate approximately $22,800 in gross monthly savings against an estimated monthly operating gap of roughly $137,000.

That does not mean every smaller saving is worthless. It means this particular saving does not remotely remove the need for customer repricing, central cost reductions, claim restructuring and new capital.

More importantly, $22,800 is the gross saving.

It is not demonstrated net value.

The net result also depends on operator departures, the amount of customer data affected, repair traffic, repair payouts, central repair costs, reduced diversity and the future cost of rebuilding capacity.

You demand operator supply curves, elasticity estimates and internal concentration data before I may reject a cut.

Fine.

But then those same missing data also prevent you from treating a cut as a useful component of the restructuring.

You cannot require a complete operator-level model from the person opposing a cut while requiring no such model from the people proposing one.

That is the actual asymmetry here.

No. That is your logic, not mine.

A cost-saving measure should be judged by its net effect.

A payroll reduction that saves $300,000 but removes the people required to operate the network could be a bad reduction.

A software-contract reduction that saves $100,000 but creates $200,000 in migration and operational costs could also be a bad reduction.

The word “cumulative” does not transform every gross saving into a good decision.

And SNO compensation has one particularly relevant feature: Storj already pays nothing for unused storage.

If the network has 100 PB of unnecessary empty capacity, the direct storage payout for that empty capacity is still $0.

A lower storage rate therefore does not reduce the cost of the free space repeatedly offered as justification. It reduces compensation for occupied storage containing customer data.

If Storj wants less excess capacity, it can slow onboarding, slow vetting in oversupplied placements and recruit only where additional capacity is required.

Those actions address excess capacity.

Cutting payment for existing customer data does not.

Of course they are different scenarios. That is why the post calculates several different reductions instead of pretending they are identical.

The table shows the spectrum:

  • small reductions produce limited gross savings;
  • larger reductions produce larger savings but increase operator risk;
  • total abolition removes the supplier network.

Showing the endpoints is not claiming that a 10% reduction and total abolition are the same event.

It shows why there is no obvious painless point where the saving becomes transformative while the supplier consequences remain irrelevant.

You have not identified such a point either.

It is a sensitivity analysis.

A sensitivity analysis is supposed to be arithmetic.

The filed budget shows that cash inflows would need to increase by approximately 40.7% merely to cover the same outflows with zero customer loss and no additional reserve.

The proposed move from $7 to $12 is approximately 71%. The table then asks how much weighted usage could disappear before the higher price no longer covers the filed outflows.

It does not claim that customer loss will actually be 5%, 10%, 17% or 25%.

The post says explicitly that no outsider can predict churn without Storj’s customer-level data.

You attack the table for failing to predict customer behavior when the text directly states that it is not a prediction.

It is the exact algebraic break-even point under the stated assumptions.

It is not an empirical churn forecast.

Those are different things.

Writing “approximately 17%” may be easier to read, but it does not change the calculation. Numerical precision in an equation is not a claim that customer behavior can be forecast to two decimal places.

You list contract restrictions, discounts, collection failures, procurement rules, bankruptcy risk and customer-specific behavior.

Those are precisely why the post demands a customer-cohort review before applying the proposal blindly.

The absence of private customer data limits every outsider. It does not make the current $7 price sustainable, and it does not make a transparent sensitivity table “fan fiction.”

The plan does not assume that every fixed contract can be repriced immediately. It explicitly says the opposite.

New customers move immediately. Month-to-month customers receive notice. Fixed contracts move at the earliest lawful opportunity. Material customers require individual treatment.

The Day-30 number is a turnaround target, not a promise that every customer has already completed several billing cycles.

And the company does not have the luxury of waiting a year before discovering whether its customer economics work. The filed budget provides roughly twelve weeks of runway.

An aggressive deadline may be missed. That is why the plan contains failure triggers.

Calling it “deadline cosplay” does not supply a more credible timetable.

How long do you propose Storj should wait before establishing whether its post-restructuring revenue can cover its recurring costs?

It is derived from what the business can afford.

At $500,000 in monthly collected revenue and 40% direct COGS, $300,000 remains for fixed costs.

That is a solvency boundary.

A bottom-up department budget is still required, but no outsider can produce it without payroll records, contracts, staffing plans and infrastructure details.

A turnaround requires both directions:

  • a bottom-up determination of what operations require;
  • and a top-down determination of what available revenue can support.

You dismiss the second because the first is not public.

That does not make the affordability limit arbitrary.

It means management must decide whether Storj can operate below it. If not, then $500,000 is not enough revenue and the revenue or financing requirement rises.

It is not “convenient.” It is accurate terminology.

A forecast claims what management expects to happen.

A target states what must happen for the strategy to remain viable.

The Year 2 and Year 3 numbers are not evidence that Storj will achieve them. They are performance gates. If Storj cannot build a credible acquisition model capable of supporting those targets, it should not expand spending on the assumption that growth will somehow arrive.

Pretending the numbers are forecasts would be dishonest.

Calling them targets is the opposite.

You took items spread across a 36-month roadmap, stacked them into one list and presented them as though all of them were promised as finished production products within ninety days.

That is not what the plan says.

The core idea is one reusable provisioning and integration layer. WordPress, NAS, backup, media and AI tools are not each supposed to reinvent account creation, credentials, buckets, billing and health monitoring.

They reuse the same layer.

Storj also does not begin from zero. It already has S3 compatibility, native Uplink, Object Mount technology, Rclone support and manual guides for several of these systems.

The proposal is to turn existing technical capability into a usable product instead of forcing every customer to reconstruct the integration manually.

Yes. That is what software products do.

They move complexity away from every individual customer and implement it once in reusable, tested software.

The fact that an automatic installer must handle authentication, credentials, caching, updates and recovery is not an argument for making every customer handle those things manually.

It is an argument for proper engineering, limited scope and staged releases.

The plan also requires named ownership, security review, code review, testing, support tooling and a real release path. You repeat those requirements as though they were omitted.

Nobody claimed otherwise.

That is why the plan requires paid pilots, paying design partners, contribution-margin measurement and failure triggers.

An integration is a hypothesis about reducing customer-acquisition friction. It is tested by whether customers activate it, pay for it, retain it and require an acceptable level of support.

If it fails those tests, it is stopped.

That is considerably more disciplined than Storj spending substantial resources on products before establishing sufficient traction—which is exactly what the court filings say happened with Object Mount.

That is a straw man.

Stating the amount of capital a turnaround requires is not “ordering” anyone to provide it.

Every financing process begins by estimating the funding requirement and use of funds.

Only then can potential sources, valuation, security, dilution and return be negotiated.

The post identifies possible instruments: equity, preferred equity or a deeply subordinated convertible instrument. It proposes conversion or extension of major existing claims and milestone-based release of capital.

The exact valuation and investor terms cannot responsibly be invented without the cap table, asset valuation, final claim treatment and actual investor discussions.

Doing so would be the pseudo-precision you accuse me of elsewhere.

Whether anyone is willing to invest remains a market test. If the capital cannot be raised, the plan contains a strategic-sale trigger.

That is not an assumption that money will materialize. It is a stated financing requirement followed by a fallback if the requirement cannot be met.

You offer no alternative estimate and no alternative funding structure.

The comparison includes Backblaze B2 and Cloudflare R2 and expressly states that Storj will not be the cheapest solution for every workload.

That is the opposite of hiding inconvenient competitors.

AWS and Azure demonstrate that the proposed prices remain below important premium alternatives. B2 and R2 demonstrate that Storj cannot compete on price alone.

The proposed response is not merely a “list of desirable nouns.” It includes specific prices, bundles, self-service checkout, direct integrations, agency and MSP channels, partner compensation based on collected revenue, paid pilots and sales compensation based on collected gross profit.

You may disagree with the strategy.

Pretending it consists only of the words “security” and “performance” ignores most of the actual proposal.

The calculation is explicitly introduced as a simplified uniform-distribution illustration.

It does not claim to describe Storj’s actual placement-weighted network.

It answers one narrow hypothetical:

If one of three independent host domains failed at each of three large operators, under uniform placement, what would the piece-loss distribution look like?

Under those assumptions, the calculation is mathematically correct.

Scientific notation is not “aesthetic authority.” It is simply a practical way of writing a very small number.

You are free to reject the assumptions as unsuitable for estimating the real network. The post itself says real concentration must be tested using Storj’s internal operator, host, subnet, ASN and placement data.

The purpose of the illustration was not to certify every large farm as risk-free. It was to show that a partial host-domain failure is not automatically equivalent to the simultaneous complete disappearance of every node operated by those companies.

That distinction remains valid.

Of course I have an interest.

So do the creditors, management, employees, investors, customers and every other participant in a restructuring.

The relevant distinction is not between interested and interest-free stakeholders. There are no interest-free stakeholders.

The relevant distinction is between a hidden interest and a disclosed one.

Mine is in the title, the premise and the conditions of my offer.

You quote those conditions as though you uncovered them.

I offered development work with a normal commercial value in the five-figure range.

I did not claim that this contribution alone would finance the entire roadmap, maintain every product or replace exit capital.

It is an additional contribution toward a defined part of the work.

The post also states that donated code requires product ownership, specifications, sandbox access, code review, security review and a real release path.

Your warning that “software is not free merely because the initial commits are unpaid” attacks a claim I never made.

That is not a contradiction.

It is the premise being tested.

A plan entitled How to survive without reducing SNO compensation naturally tests whether all other available levers are sufficient to preserve the current rate.

The calculations show that this is at least financially possible under the stated assumptions.

They do not prove that management will execute it, that customers will accept every price, or that investors will provide the capital.

That is why the proposal contains tests, milestones and failure triggers.

Your preferred methodology appears to be that no external proposal may be called a plan until it contains all internal data, completed customer research, final investor commitments, a department-level budget and validated product-market fit.

At that point it would no longer be a proposal.

It would be the completed restructuring.

The central problem with your criticism

You repeatedly demand evidence for every part of my proposal while supplying none for the alternative you preserve.

You say I cannot prove an SNO cut is harmful without operator elasticity and concentration data.

Then you cannot prove that it is a useful net saving without those same data.

You say a price increase is unvalidated.

Correct—that is why the proposal includes a sensitivity analysis, cohort review and failure triggers rather than pretending the outcome is known.

You say integrations are not automatically demand.

Correct—that is why the plan requires paying design partners, margins and staged releases rather than assuming every integration succeeds.

You say $5 million is not automatically available.

Correct—that is why it is described as a capital requirement, with a strategic transaction if it cannot be raised.

Those are uncertainties to be tested.

They are not contradictions.

After more than two hundred lines of criticism, you have not shown:

  • what Storj should charge customers;
  • what it should pay SNOs;
  • how much a cut would save net of its consequences;
  • what recurring fixed cost it can afford;
  • what product should be built instead;
  • how much capital is actually required;
  • or what happens when your preferred assumptions fail.

You have shown that a public proposal lacks Storj’s private data.

Everybody already knew that—including the person who wrote the proposal and stated it explicitly.

If you want to replace the plan, provide a better one.

Put numbers on the customer price, SNO rates, fixed costs, runway, financing, product priorities and expected consequences.

Until then, this is not a demolition of the proposal.

It is a very long objection to the fact that an external stakeholder cannot already possess the internal information required to complete it.

Your response does not rehabilitate the proposal. It merely restates the proposal’s assumptions with greater indignation, more semantic cushioning, and an even more elaborate apparatus of self-exoneration.

The fundamental criticism was never that an external stakeholder lacks omniscience. The criticism was that you repeatedly transform that admitted lack of knowledge into categorical conclusions, numerically ornamented declarations, aggressive deadlines, and purported solvency boundaries—and then retreat to “this is only a target,” “this is only a sensitivity analysis,” or “management must supply the real data” whenever the evidentiary vacuum beneath those declarations is exposed.

That is not rigor. It is epistemic arbitrage: claiming the authority of precision while disclaiming the obligations that precision ordinarily entails.

Declaring your bias openly does not convert it into analysis

You repeatedly insist that the prohibition against reducing SNO compensation was disclosed in the title, as though conspicuous disclosure somehow resolves the methodological problem.

It does not.

Nobody accused you of hiding your preferred conclusion particularly well. The criticism is that you selected the conclusion first and constructed the analytical machinery around it afterward. Announcing the predetermined destination before drawing the map does not make the route objective. It merely makes the circularity easier to observe.

Yes, stakeholders may state conditions. Creditors may demand repayment. Employees may demand salaries. Suppliers may reject lower prices. Investors may demand returns. All of that is commercially ordinary.

But a stakeholder demand is not automatically a comprehensive restructuring plan merely because it is accompanied by tables, algebra and an architecture diagram.

“I will not accept a reduction” is a negotiating position.

“The company can viably avoid any reduction” is an analytical proposition.

You continually slide between those two statements as though they were interchangeable. They are not. The first requires only personal conviction. The second requires evidence about the company’s economics, customers, suppliers, operations and financing—the very evidence you repeatedly acknowledge you do not possess.

Disclosure eliminates concealment. It does not eliminate bias, validate assumptions or convert advocacy into neutral analysis.

Your burden-of-proof argument is backwards

The most conspicuous rhetorical maneuver in your reply is the demand that a critic provide a complete alternative customer-pricing system, payout schedule, operating model, financing structure and product roadmap before being permitted to identify defects in yours.

That is nonsense.

A person does not need to design a new bridge before demonstrating that your bridge calculations omit wind load. A reviewer does not need to write an alternative operating system before identifying a security vulnerability. A restructuring proposal does not become sound merely because the critic has not supplied a rival restructuring proposal of equal length.

You made the affirmative claims. You proposed the prices. You selected the deadlines. You declared the fixed-cost ceiling. You asserted that there is no economically useful SNO-cut range. You proposed the capital requirement. You selected the product roadmap. You claimed the combination was financially possible.

The burden therefore rests with you to support those propositions.

“Where is your complete alternative?” is not a substantive answer. It is an attempt to replace scrutiny with an admission fee: unless the critic first writes another enormous speculative document, your own document is supposedly entitled to provisional credibility.

It is not.

A critique can successfully show that a proposal is unsupported, internally inconsistent or operationally implausible without simultaneously solving the entire company. Demolition inspectors are not required to construct a replacement building before identifying that the existing one lacks foundations.

You cannot use missing data as both shield and sword

Your treatment of unavailable information is especially revealing.

When challenged to demonstrate that maintaining current SNO rates is economically optimal, you say the necessary operator elasticity, concentration and supply data are private.

When challenged on customer churn, you say the necessary customer-cohort information is private.

When challenged on the fixed-cost structure, you say payroll, contracts and staffing information are private.

When challenged on financing, you say the cap table, valuation and investor discussions are private.

When challenged on product demand, you say the integrations would be tested through pilots after implementation begins.

Yet despite this pervasive absence of decisive information, you still announce:

  • that no useful SNO reduction exists;

  • that a 70% weighted price increase is the appropriate opening move;

  • that $300,000 is the relevant fixed-cost boundary;

  • that $5 million is the appropriate capital requirement;

  • that multiple integrations constitute the correct product direction;

  • and that the resulting configuration is financially possible.

Apparently the missing data are fatal whenever someone challenges your preferred conclusion, but merely administrative details whenever you assert it.

That asymmetry is not cured by accusing the critic of asymmetry.

The intellectually defensible conclusion from missing operator data is not “therefore SNO cuts are beneficial.” But neither is it “therefore current rates must remain untouched.” The defensible conclusion is that the net effect is unknown and must be modeled before either categorical policy is adopted.

Your proposal, however, does not preserve that uncertainty. It converts your preferred side of the uncertainty into a covenant and assigns the unresolved evidentiary burden entirely to everyone else.

A partial saving does not need to solve the whole company

You continue to attack an argument nobody made: that reducing SNO compensation alone would rescue Storj.

The point was not that a 10%, 20% or 50% adjustment closes the entire funding gap. The point was that restructuring packages are cumulative. Payroll savings, contract renegotiations, product closures, debt treatment, supplier adjustments, financing and repricing can each contribute without individually constituting a complete solution.

Your response now says that each measure should be evaluated by its net effect.

Precisely.

But you do not know the net effect of an SNO adjustment. You have no operator supply curve, no empirically grounded departure model, no measurement of which operators are marginal, no quantified repair-cost response, no analysis of geographic substitution and no reliable estimate of capacity reacquisition costs.

You therefore cannot legitimately move from “the gross saving is $22,800 per month” to “the saving is not demonstrated net value” and then onward to the practical conclusion that the category should be contractually immunized.

An uncertain net value is not the same thing as a negative net value.

You demand proof of benefit before considering a reduction, while requiring no comparable proof that preserving every component of the existing payout structure is optimal. The status quo receives a presumption of perfection merely because it is the status quo you prefer.

That is not commercial neutrality. It is selective evidentiary maximalism.

Your discussion of unused capacity is similarly beside the point. Yes, Storj pays for occupied storage rather than empty disk space. Nobody remotely competent disputes that. But the fact that unused capacity has a direct payout of zero does not prove that the rate paid on occupied capacity is economically optimal. It merely describes the billing unit.

A company can possess excess available supply while still paying more than necessary for the units it actually consumes. Whether that is true here requires evidence. Repeating that empty space costs nothing does not answer the question.

Exact arithmetic can still communicate false confidence

Your defense of the 17.21% break-even figure is formally correct and substantively evasive.

Yes, the equation produces an exact algebraic result under the assumptions entered into it. Nobody suggested that your calculator malfunctioned.

The criticism concerns the inferential weight placed upon the output.

A model may calculate a break-even threshold to ten decimal places while remaining commercially useless because its inputs do not approximate customer behavior. Mathematical exactness conditional on speculative premises is not empirical precision. It is merely deterministic arithmetic performed on uncertain assumptions.

Calling the table a sensitivity analysis does not immunize it from criticism. A sensitivity analysis is useful only insofar as the selected variables, ranges and relationships illuminate the system being examined. Your table largely assumes a uniform price multiplier, applies stylized usage reductions and compares the resulting arithmetic with filed outflows.

It does not meaningfully model customer segmentation, delayed migrations, negotiated pricing, contractual timing, collection lag, workload elasticity, expansion freezes, bankruptcy-related procurement concerns, competitive counteroffers or the possibility that revenue deterioration occurs nonlinearly.

You acknowledge these omissions and then behave as though acknowledging them converts the output into a reliable turnaround instrument.

It does not.

A caveat is not a substitute for a model.

Nor does calling the Day-30 figure a “target” solve the temporal incoherence. A target is not rendered credible merely because failure is permitted. One can target a Moon landing by Friday and include a trigger requiring reassessment on Saturday. The existence of the trigger does not make the schedule operationally serious.

If fixed contracts cannot move immediately, customer responses cannot be observed immediately and collected cash necessarily lags announced pricing, then a Day-30 collected-revenue target derived substantially from repricing requires a concrete cohort-level migration schedule. Without that, it is not an execution target. It is a wish expressed in the imperative mood.

An affordability ceiling is not an operating design

You say the $300,000 fixed-cost cap is derived from what the business can afford.

No. It is derived from what your chosen revenue and gross-margin assumptions leave available.

That distinction is the entire criticism.

A top-down ceiling may reveal that the company must operate below a certain level to survive. It does not demonstrate that the company can actually provide the required service, compliance, support, security, engineering, billing and administration below that level.

Suppose the minimum viable operating structure costs $450,000 per month. Your equation does not magically compress it to $300,000. It demonstrates that your assumed revenue configuration is insufficient.

The bottom-up model is therefore not some clerical appendix management can fill in later. It is essential to determining whether the central operating premise is feasible at all.

Calling $300,000 a “solvency boundary” gives the figure a gratifying air of inevitability, but it remains a residual generated by assumptions you selected. Arithmetic can define an affordability constraint. It cannot establish operational viability.

Targets are not forecasts—and labels are not evidence

You devote substantial prose to distinguishing targets from forecasts.

The distinction is valid and largely irrelevant.

Nobody needs a seminar on the dictionary difference between what management hopes or requires and what management predicts. The objection is that your targets are presented as the backbone of a proposed turnaround without a credible causal account of how they will be reached.

A target unsupported by a customer-acquisition model, staffing plan, delivery capacity, sales funnel, product validation and financing pathway is not fraudulent merely because it is aspirational. But neither is it useful merely because it has been correctly labeled.

You cannot transubstantiate conjecture into discipline by renaming every unsupported number a “performance gate.”

A failure trigger tells the company when the imagined plan has failed. It does not make success more likely. An emergency exit is valuable, but painting one onto a wall does not create a door.

A shared platform does not abolish product complexity

Your defense of the product roadmap relies heavily on the claim that multiple integrations reuse one provisioning and control layer.

That may reduce duplication. It does not collapse the integrations into one product.

WordPress media management, WordPress backups, NAS deployment, FUSE behavior, media-server streaming, cache management, migration orchestration, hosting-panel integration, backup ecosystems and AI-data workflows each contain distinct compatibility surfaces, failure modes, user expectations, support burdens, release cadences and security implications.

A reusable credential and provisioning layer does not make WooCommerce compatibility equivalent to Plex range-read behavior. It does not make filesystem semantics identical to WordPress URL rewriting. It does not make NAS packaging equivalent to Kubernetes templates. It does not make customer support reusable merely because each product ultimately stores objects.

“Yes, software hides complexity” is not a response to the scope objection. It is a slogan about abstraction.

The engineering question is how much complexity can be reliably centralized, how much remains domain-specific, who will maintain each integration, what minimum viable subset creates paying demand and which one deserves scarce organizational attention first.

Your roadmap answers this by naming nearly every plausible adjacency and assigning overlapping alpha, beta and production milestones. That is not prioritization. It is a catalogue of ambitions.

Paid pilots and failure triggers are sensible mechanisms. They do not justify building toward numerous hypotheses simultaneously during insolvency. Validation should narrow scope before large commitments, not serve as a ceremonial disclaimer attached to a sprawling roadmap.

A capital requirement is not a financing strategy

You object that stating a $5 million requirement is not the same as ordering investors to provide it.

Of course not. The original criticism was rhetorical, not a claim that your forum post possesses legal authority over capital markets.

The substantive point remains untouched: identifying how much money you would like to spend is not the difficult part of financing.

The difficult part is articulating why a rational investor would provide that money, on what terms, against what assets, with what expected return, under what governance, at what valuation and with what credible route to liquidity or profitability.

Your answer is that these details cannot be invented without investor discussions and internal information.

Correct—which is why the financing section is a capital requirement and use-of-funds proposal, not an exit-financing strategy.

Once again, you defend an undeveloped component by explaining why it is undeveloped, then object when someone accurately describes it as undeveloped.

A fallback sale trigger does not solve capital formation either. “Raise the money or sell the company” is a contingency tree, not evidence that either branch is executable on acceptable terms.

Transparency about self-interest does not make self-interest irrelevant

You state that every participant has an interest and that yours is disclosed.

Again, true and beside the point.

The criticism was not that you secretly benefit from preserving SNO compensation and protecting professional multi-node deployment models. The criticism was that these personal commercial requirements are elevated into inviolable design constraints while nearly every other stakeholder is instructed to accept repricing, reductions, restructuring, delay, conversion, risk or dilution.

Your contribution may be valuable. Your willingness to provide development labor may be sincere. Neither fact entitles your preferred supplier terms to analytical immunity.

A disclosed conflict remains a conflict. Disclosure allows readers to evaluate it; it does not neutralize it.

Indeed, the explicit conditions strengthen the original characterization of the document as a stakeholder negotiating proposal. There is nothing shameful about that. What is misleading is continuing to present the resulting document as though it were a comprehensive examination of the available restructuring space.

It is not. It is a restructuring scenario constructed within boundaries selected to protect your own economic position.

The central problem with your response

Your response is an enormous exercise in terminological relabeling.

Unsupported forecasts become targets.

Speculative product bets become hypotheses.

An arbitrary residual becomes a solvency boundary.

An unproven price increase becomes a sensitivity analysis.

A funding wish becomes a capital requirement.

A predetermined supplier protection becomes a constraint.

Missing evidence becomes private information.

Personal commercial conditions become stakeholder rights.

And every criticism becomes invalid unless the critic supplies a complete competing plan.

This creates a self-sealing argumentative system. When a number looks authoritative, it is offered as evidence that the plan is concrete. When the number is challenged, it becomes merely illustrative. When the timeline appears decisive, it is an execution target. When the timeline appears implausible, it is only an aspiration with failure triggers. When the proposal is called biased, the bias is defended as disclosed. When it is called incomplete, the incompleteness is blamed on unavailable data. When its conclusions are questioned because of that unavailable data, the critic is challenged to prove the opposite using the same unavailable data.

Nothing can falsify the proposal because every component changes epistemic category the moment pressure is applied.

That is not robustness. It is rhetorical liquidity.

You conclude by demanding that critics “provide a better one.”

No.

First establish that this one is more than a highly elaborate statement of what you would like Storj to do while leaving your own compensation and deployment model untouched.

The critic is not obligated to produce another speculative 36-month corporate reconstruction merely to point out that yours substitutes arithmetic for market evidence, targets for execution capacity, integrations for validated demand, capital allocation for capital availability and disclosed self-interest for analytical neutrality.

Your reply contains more words, but it does not contain the missing evidence.

It contains a sophisticated taxonomy of excuses for why the missing evidence should not count.

The original proposal looked like a plan because it possessed the visual and linguistic furniture of one: tables, formulas, milestones, capital buckets, APIs, gates, covenants and scenarios.

Your response now adds another layer of verbal architecture around that furniture.

But a larger rhetorical building erected on the same empirical vacancy is not a stronger structure.

It is simply a more ornate way of standing on air.

That’s it from me.

Any usage of LLMs in a human-focused forum should be met with further LLM response. If you are using it for “translation” just write it in German.

Congratulations, that is quite an impressive wall of text. No, I am not going to write in German on an English-language forum; my last reply took me more than an hour, while you treated it as a joke and used it to discredit me.

Schade um die Zeit. Ich glaube kaum, dass Shuja Ali hier mitliest.

I won’t write a 20-page essay here, but to sum it all up: cutting SNO payouts which might save STORJ around $25k–$30k a month (at best scenario) won’t solve their problems in the slightest.

You can see the amounts involved here for yourselves - the court will either decide to shut STORJ down or give it a chance, but changing SNO payouts won’t make a difference. There was simply no need to put dozens of people through all that stress.

What stress? Surely there ae not a large number of SNOs (if any) making a living out of Storj who would be stressed about whether Storj lives or dies?
Emotionally invested, maybe. Disappointed, even. But “stressed” I find hard to believe…

That’s why the forum has been buzzing all week, right?

Nah, I’d call it interest.
But who knows? :man_shrugging:t2:

Forum is buzzing because there is announced that need to do reduction but no decisions how much it will be or will be something at all. Unknow future is that drive out people the most. We need that storj tell us, we made calculations and it will be like THIS from this time.
Then people can make own decision. Also need information well it be paid at all. But it is up to the court. We all know that government institutions work very Slow, but there is ton of papers.

Most people here expressing there emotions, in than pervious situation it is understandeble.

This is a pretty well-structured analysis. I don’t have the attention capacity to verify it in more details, so I’ll assume that e.g. the numbers are correct. I also haven’t looked at the court documents, didn’t have time for that, sorry.

I’d like to comment on that I don’t think that Storj with the limited crew has capacity to develop any new features, like the one-click installers; besides, it has been painfully obvious though a number of Windows-only issues even with storage nodes that Storj has never had good Windows programming capabilities—it is truly a different thing from developing for UNIX-like servers. In this sense I agree with @Ambifacent. These integrations must therefore be done through partnerships with entities experienced in delivering integrations like these, likely externally funded too. I know that some folks running Wordpress or smaller stores would benefit a lot from CDN-like features of Storj to optimize media delivery latency, but Storj’s egress is not priced competitively for these customers. I’d therefore assume these partnership deals would have to have custom pricing. And partnerships would be better structured to handle support of many small users.

We should not forget about some non-web, or non-public network scenarios, like integrating into non-web software, and selling private cloud services. Both were done by Storj in the past, and, as was discussed on today’s office hours, at least the latter is still very viable—even if not really interesting from a node operator’s perspective. They offer a different model of earning money by Storj that might be able to leverage skills of engineers who had to leave Storj.

There is a way to structure content so that it is easy to follow for humans. Modern off-the-shelf LLMs don’t do that by default. Most humans can’t do this without training either. But the failure modes are different: where LLM will simply follow the instruction and prepare an overly academic or explanatory writing which takes ages to digest, humans will take shortcuts assuming context or background knowledge, and will tend to wait for questions to learn which shortcuts did not work. So it’s less of a pain and at least the intent is clear.

If you value your readers, you should learn how to get your LLM to write arguments in a way that is easy to read. I’m not going to read the exchange between the two of you here. I would read a summary though.

Storjers,

I usually don’t like getting into internet discussions, regardless of the subject, but this one deserves recognition. What GfTmbH wrote was impressive. This was not just a discussion — it was a master class.

I may not agree with everything that was said, but I have to show my respect for that post. I don’t know if it was written with AI assistance, but if it was, please let me know which AI was used, because I need that one.

I’ve been part of the Storj community and a node operator for a few years. I currently have 45 nodes running, with 500TB+ of total capacity.

I fully agree with the point that node operators are not doing this for free. Maybe some are, but I believe the vast majority are not. I can understand not being paid specifically for traffic, but for me, the storage capacity being used absolutely needs to be compensated.

Of course it does.

Making it an Open Source Community project, cuts mosts costs to 0$.

Cut S3, that brings down satellite costs. We have to depate on how the “keep track of things” satellites are financed, but still, that is small part of satellite costs.

You are falling multiple times for the “false dilemma” fallacy.

If we go back to the petrol station example, it is not that there arent any other petrol stations that can offer gas at a price were customers don’t refuse to pay. The gas station next to us sells gas for 4$ a gallon and buys it for 3.8$ per gallon. The customer does not care that we buy our for 4.1$ and want to sell it to him/her for 4.2$. Even thoug we have a smaller margin, it is still 20cent more expensive. Why should the customer come to us? He/she doesn’t. Just like with Storj, customers did not come.

Sure, but on all other S3 commodity metrics, STORJ is also not great.
Like discussed a million times, I either need high performance (aka in datacenter, something which is impossible for STORJ or any external S3 to provide) or I want cheaper. That is why there are no paying customers. Tell me a single customer group, that is not niche, that choses Storj over Backblaze or datacenter storage? You won’t. You can’t. You and Storj talked FOR YEARS now about how customers love the georedudancy and the speed and blablablabla. Nothing happened.

Yep, and this is why the thing that I told you for years would happen, actually happened; chapter 11.

It was fun while it lasted. There was VC money, there was hype, but you can’t outrun economics forever. For many years, even decades. But not forever.

Ok, let us look at costs.
Backblaze 7$ per TB, free download two times, all over S3.

S3 is out of the window for obvious reasons. In case some people are ultra dense, I will still repeat it: S3 needs satellites, satellites need to pay for traffic, because traffic is not free. Roughly 5$ per TB. So that one is isn’t even worth discussion. S3 flies directly out of the window.
We already start with a HUGE, like fucking HUGE disadvantage over Backblaze. Remember that we need to undercut cut pricing drastically because of that.

Free download we also can not offer. Otherwise is gets complicated again on how you pay nodes. And you do need to pay nodes somehow for traffic, otherwise everyone would just limit upload to 10mbit or something. But that is not as big of a problem, if STORJ is just a backup thing. Most users assume they never need to make use of it, and if it happens they are willing to pay. So I don’t know, lets say 5$ per TB? 4$ go to nodes, 1$ to the project?

Now to the storage pricing. We remember that we have to drastically undercut the 7$ from backblaze. Because otherwise nobody will bother to start a docker or whatever to upload with storj native. Remember, it is still a huge issue that we don’t have any native TrueNAS or Synology implementation. The only “mass appeal” implementation we have is FileZilla.

So I would say 50% discount. That means 3.5$ for customers. For nodes, this means 3.5 / 2.2 = 1.59$ per TB stored.

If that is too little for you, you should never have entered storj to begin with.

You don’t need to make this complicated. A customer pays $x/TB stored and gets y TB/month free download. Node operators get paid some amount for storage and some for egress.
It’s like when a seller offers free shipping. You know the delivery company gets paid anyway, so it means that the shipping cost is just part of the item price.
Similar here. A customer pays a bit more for storage and the egress is included there, so you can offer “free download”. Then, not all customers will use up their free download, so you have some money left over.
Having free download may work differently psychologically, similar to having free shipping. People may not want to have to pay to download a file they just uploaded, having free download once or twice may attract more people.

The issue with that is that the pricing will be less fair. Current pricing is $7 per TB for used space and egress. I believe we would have to change that to at least $7.7 per TB with 1x free egress. Now you have the situation that customers with low egress pay more while customers with high egress end up with a lower invoice. We have customers that don’t like that idea on both sides. Customers with high egress are worried that we are going to increase costs for them the closer they get to 1x free egress because in this example we would make a loss on customers with more than 10% egress. To maximize profit it makes sense to get rid of these customers… Thats the wrong incentive. On the other end of the scale are also customers with low egress that prefer to pay for what they consumed without cross financing other customers.

So pay for that you used is fair and logic for all.

Such plans work, though it depends on the customer and how he thinks. I guess a large company, that is the intended customer, would calculate the cost and choose whatever is cheaper compared to a regular person.

Make a simple example.
I made mine.

Product Customer Node Storj profit
Stored TB per month 3.5$ 1.6$ 0$
Download per TB per month 5$ 4$ 1$

Now, IMHO we could also change the download price, since backup users don’t care that much.

Product Customer Node Storj profit
Stored TB per month 3.5$ 1.6$ 0$
Download per TB per month 10$ 5$ 5$

both IMHO fine.

Still needed in both cases:

  • ditch loss leader S3
  • satellite has to be cheap to operate. If that means minimum chunks of 4MB or something, so be it
  • offer native integrations in appliances

Now the I can hear the “I need more for stored” crowd scream in the background. The thing is, it is simply not possible. Play the numbers, it does not work. Lets assume your 5$ per TB fantasy:

Product Customer Node Storj profit
Stored TB per month 3.5$ 5$ -7.5$
Download per TB per month 80$ 5$ 75$

That model would only work if:

  • users are actually fine with paying the insane amount of 80$ per TB download
  • backup users download the insane amount of 10% of their total stored storage per month

However you change it, it is not possible. AKA you can only use unused resources. AKA your node has to run for free.

This one is missing the expansion factor. A customer pricing of $3.5 is a loss for storj. (Maybe one day I can share the actual unit economics.)

No it is not. If I the customer pays 3.5$ per TB, and you pay nodes 1.6$ per TB, that is with the expansion factor 2.2:
3.5 - (1.6 * 2.2) = 0$ profit for storj.
Just like in the table.