[ranger_danger]

You're not wrong, but it's funny that this same topic was posted just earlier today with a very different sentiment in the comments.

https://news.ycombinator.com/item?id=44606646

But if you dig hard enough, there's actually links to more evidence of why it is that complicated... so I don't think it was necessarily intentionally done as a method of lock-in, but where's the outrage in that? /s

"Complicated file format has legitimate reasons for being complicated" just doesn't have the same ring to it as a sensationalized accusation with no proof.

[mjevans]

Weren't those reasons effectively...?

'special case everything we ever used to do in office so everything renders exactly the same'

Instead of offering some suitable placebo for properly rendering into a new format ONCE with those specific quirks fixed in place?

[lozenge]

What would that look like?

"You have opened your Word 97 document in Office 2003. The quirks have been removed, so it might look different now. Check every page before saving as docx."

"You have pasted from a Word 97 document into an Office 2003 OOXML document. Some things will not work."

[ranger_danger]

If you want a look at how this (doesn't) work in practice, just look at Libreoffice. I once made 3 hours worth of comments on a Word doc, and as soon as I saved it, they all vanished into thin air.

And for the pedantic, yes it warns you when saving as a .docx that "not all features are supported", but it does that every time, for every document, so nobody pays attention to it or has any idea what it even means. To me the way it handles this is just completely unacceptable.

[mjevans]

Office could freely continue to support their old proprietary formats if they wanted.

In an ideal world a converter would generate E.G. a 1200 PPI render of each page, then compare it to a similar render as provided in the nearest rendition in the allowed simple new format. Those could be diffed to produce a highlight of areas that changed.

The software could then ask if the transcription from one format to the other was close enough, or if there were some corner case that wasn't good enough.

Bonus points, collect feedback if the end user is willing to submit examples.

[dathinab]

you can have both

having a lot of intend to keep it complicated and cause vendor locking and comply in bad faith

and this being very easy to archive just by not trying to improve on a status quo and creating a standard where you are the only one to decide what goes in where. Or other simple things like intentionally putting a senior engineer you know tends to painfully overweening things but keep it in a working state, etc. etc. Just by management decisions done in a higher level then the project you can pretty reliable mess up things in various ways as needed pretty reliable as long as you have enough people to choose from.