[DNSB]

Originally Posted by hobnail

I suspect that if you use a "reset" that ereaders that don't know epub3 will at least render the unknown tags reasonably. For example, I use

article, aside, figcaption, figure, footer, header, main, nav, section {
  display: block;
}

address, article, aside, blockquote, canvas, dd, details, div, dl, dt, figcaption, figure, footer, h1, h2, h3, h4, h5, h6, header, hr, li, main, nav, ol, p, pre, section, summary, ul {
  margin: 0;
  padding: 0;
  line-height: inherit;
  font-style: normal;
  font-weight: normal;
}

a, abbr, b, bdi, bdo, cite, code, data, del, dfn, em, i, ins, kbd, mark, q, rp, rt, rtc, ruby, s, samp, small, span, strong, sub, sup, time, var {
  font-size: inherit;
  vertical-align: baseline;
  font-style: inherit;
  font-weight: inherit;
  color: inherit;
  text-decoration: none;
}

Which I copied from the blitz site.

Personally, I find the reset code almost useless and unneeded.

If I recall correctly, an epub2 renderer is supposed to disregard any elements and attributes it doesn't understand. The problem there is that if you are using the epub3 element for a purpose, now it doesn't happen and that can lead to an ugly rendering of the page. To fix this, you can end up with workaround code which is a pain to maintain since it may not render even close to the same on multiple devices/apps.