187093 - Optimize font size choices (+17, +22, +30, +34, +44; -minimum: 6, 7, 8)

Status: VERIFIED FIXED

(Core :: Layout: Text and Fonts, enhancement, P4)

Description

[Felix Miata]

Trunks 2002122905 Linux, 2002122812 OS/2, 2002122908 W32.

Current pixel choices are 8-16, 18, 20, 24, 26, 28, 32, 36, 40, 48, 56, 64, 72.

17 and 22 should be added to the existing choices, and possibly 30 & 34 also.

The reason for 22 should be obvious, but in case anyone needs it spelled out, it
is the choice I would most likely make if I was using 1280x960 or 1280x1024
resolution. I should think that many of those that do use either of those
resolutions cringe at the 17% gap between 20 and 24 that doesn't exist between
24 & 28, which is filled by 26.

The reason for adding 17 is less obvious, but quite real. Many common fonts go
from normal stroke to bold stroke at 18 pixels. Arial, helvetica and various
Gothic fonts are examples. I use 1024x768 and 16px Arial for my default. I'd
like a bigger default font without making my default font bold, which is what
happens if I change from 16px to 18px, and would not happen if I could choose 17px.

Compared to W32 and OS/2, in Linux there is an irritating and considerably
longer delay upon selecting fonts in prefs before the panel changes from the
current selection to the fonts panel. I suspect this might be improved by
removing the gigantic sizes. Is it really possible anyone would ever select 48px
or larger as their default size? Maybe someone using 3200x2400 on a 35" screen?

Seems like an update to the bug 52080 fix could do this.

Comment 1

[Boris Zbarsky [:bzbarsky]]

Layout.

The linux issue is due to the fact that getting a font family list from the X 
server is very slow, not due to the sizes listed....

Comment 2

[Felix Miata]

Since 9px as a legibility floor seems to be consensus, it seems odd that both
8px & 9px are provided as default options. If it were up to me, 8px would
certainly be removed, and 9px probably as well. As a check, I set a 19" display
to 640x480 and Mozilla to 8px and went to http://www.microsoft.com, which sets
TD=xx-small in CSS. Nothing but the headings was anything but unintelligible
scribble. At 1600x1200, making sense of 8px requires a strong magnifying glass
and good imagination.

Comment 3

[Felix Miata]

Attachment: Full boat patch

My first ever patch submission.

This adds to proportional and monospaced choices 17, 22, 30, 34, 44; and
removes 56, 64, 72. It also removes the sub-legible minimum sizes 6, 7, 8; and
adds 17.

Comment 4

[Felix Miata]

Taking, since I already submitted a patch.

Comment 5

[timeless]

Comment on attachment 123556 [details] [diff] [review]
Full boat patch

i'm sold, i just had to be reminded that these are *pixel* values.

Comment 6

[jag (Peter Annema)]

Comment on attachment 123556 [details] [diff] [review]
Full boat patch

I'd say for now leave the 56, 64 and 72 values in. Safari allows you to go up
to 288px, MacIE allows you to pick up to 24px, or enter a value yourself.

Go ahead and make the rest of these changes though.

Comment 7

[Felix Miata]

Attachment: initial patch modified as per comment 6 sr (retain 56, 64, 72)

FWIW, here are the approximate characters per maximized viewport line using
arial @ UXGA (1600x1200):

56 @ 48px
46 @ 56px
39 @ 64px
34 @ 72px

To get 80 56px characters on one line would require in excess of 2780
horizontal resolution. To get 80 72px characters on one line would require in
excess of 3760 horizontal resolution.

Comment 8

[Mike Kaply [:mkaply]]

Comment on attachment 124031 [details] [diff] [review]
initial patch modified as per comment 6 sr (retain 56, 64, 72)

changes made, carrying over r, requesting sr again

Comment 9

[jag (Peter Annema)]

Comment on attachment 124031 [details] [diff] [review]
initial patch modified as per comment 6 sr (retain 56, 64, 72)

sr=jag

Comment 10

[Asa Dotzler [:asa]]

Comment on attachment 124031 [details] [diff] [review]
initial patch modified as per comment 6 sr (retain 56, 64, 72)

a=asa (on behalf of drivers) for checkin to the 1.4 branch.

Comment 11

[Mike Kaply [:mkaply]]

Checked into trunk and branch

Thanks!

Comment 12

[Felix Miata]

Conforming summary to the final patch.

Verified fixed in 1.4RC1 2003052912 Linux, 2003052912 OS/2, & 2003052908 W32.

Comment 13

[Felix Miata]

*** Bug 134991 has been marked as a duplicate of this bug. ***