Text reflow woes (or: I want bullets back!)y

Aaron Janse aaron at ajanse.me
Mon Dec 16 02:35:19 GMT 2019
On Sun, Dec 15, 2019, at 5:53 PM, Brian Evans wrote:
> I believe, once again, we are wading into territory where wrapping
> and reflowing are being conflated. 

Wow. I absolutely did make this mistake. Thank you for clarifying.

The speculative spec states:
> In order to facilitate a comfortable user experience with simple
> clients which do not implement reflowing, authors SHOULD limit the
> width of lines to 78 characters, excluding CRLF pairs.

I got so caught up in the false dichotomy of either reflowing or 80-column
text that I never questioned why source code is hard-wrapped in the first
place. 

> At present, wrapping is the default that is happening for the vast majority
> of clients

I agree. The spec's justification for suggesting that source code is
hard-wrapped is the idea that it makes life easier for simple client.

And yet:
1. It's much easier for simple clients to wrap than to reflow
2. Afaik, most simple clients won't even need to worry about wrapping because
   it will already be handled by their output medium (e.g. terminal, browser).

>From a while ago:
> Text reflow IS A MUST. We live in a world of wildly varying screen sizes
> these days. The trend is clearly headed towards polarization: huge 22'+
> desktop screens in one corner, small less than 7' smartphone screens in the
> other. Nothing in the middle. There's no way you're going to get decent
> presentation of non reflown text on both kinds of displays. No matter how
> many kludgy rules you instate on default width, max before forced wrap, etc.

All of this is completely possible without reflow. All we need to do is to
state in the spec that text MUST NOT be hard-wrapped at 80 columns unless it's
supposed to be displayed that way.

tl;dr If the spec is changed to discourage hard-wrapping source code, there'
would be no need for reflow in order to have text fit the width of a client's
screen. This would make life much easier for both authors and readers using
the gemini protocol.

Pardon my ignorance if the above is what everyone else was already thinking.


Archive home | About this archive