Very surprised to see just few comments mentioning EPUB, which is IMO is much more suitable for document-centric approach. An open standard with freely available[1] specification and never had any problems with EPUBs on PC, tablets and phones.
Not only simple browser plugins per the other reply (and a plethora of non-crashing mobile apps, whereas mobile PDF reading apps crash on me all the time) - the ePub format is just a zip file in disguise with plain text (HTML) inside and maybe some images/etc.
In a manner of speaking, ePub as a design has an inherent built-in fallback mechanism to manually obtain the internal content in case of failure - including ability to try and repair a broken zip format (zip -F/-FF) and grep it in place (zipgrep).