[rfc-i] draft-hildebrand-html-rfc

Iljitsch van Beijnum iljitsch at muada.com
Fri Jul 13 03:01:00 PDT 2012


On 13 Jul 2012, at 11:46 , Martin J. Dürst wrote:

>> I don't know about other systems, but in Safari SVG is extremely problematic.

> Can you say exactly what's problematic about it? I have had experiences where Safari was better than any other browser for some SVG.

It used to be that stuff like resizing and moving images didn't work very well. But it looks like this is much better now. Some examples:

http://commons.wikimedia.org/wiki/File:Aquitani_tribes_map-la.svg
http://commons.wikimedia.org/wiki/File:Map_of_the_Imperial_Circles_(1560)-LA.svg

>> And AFAIK none of my image editing programs handle them, either.

> SVG is a vector graphics format, not an image format.

I would still want to be able to read the files into a bitmap image editor.

> For the record, I think that SVG is the way to go for diagrams. It's a bit early to use in a context such as RFCs, but then we don't exactly want every RFC to have a ton of graphics, so it should be okay.

Let me repeat what I wrote a while ago: I'm ok with having SVG or some other vector format as the highest quality alternative, but there are so many pitfalls with vector formats, such as the fact that you can't hack together your own viewer because those viewers are too complex, unbounded file size regardless of image size, possible ambiguities when displaying in different viewers / sizes, ability to use excessively small text and other elements, etc. So I think for every image there should be a relatively low resolution (think ~ 640 pixels wide) version, which would be the authoritative version of the image. And RFCs should be "fully functional" even when their images are removed.

>> For those who want "real" equations: please explain the consequences that has for greek letters and italics in the text of RFCs.

> What consequences would you expect?

Both of these are commonly used in equations, would be interesting to say the least if they can't be referenced in text. Although I guess real mathematicians let the equations speak for themselves, the way real programmers let the code speak for itself. But allowing greek and italics in the text would make it very hard to read RFCs on traditional text terminals and/or converted to the current "legacy" format.


More information about the rfc-interest mailing list