[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: firefox/minefield screen resolution

Le vendredi 15 février 2008 à 20:29 -0800, darrell pfeifer a écrit :

> Thanks for the information, it really helped.
> xrdb reports:
> Xft.dpi:	120
> Setting that value in about:config gives a decent display. If I go
> higher (say the 143) then portions of the window on the right and
> bottom aren't visible.
> I'm assuming that the value from xdpyinfo is probably incorrect,
> though I don't know at this point where to look.

You can easily check if xdyinfo is correct by dividing your screen
dimensions (in pixels) by their actual physical value (in inches). If
you find significantly different values from xdpyinfo, x is wrong and
you need to open a bug to get it fixed (@rh is you use FLOSS gfx
drivers, @someone else otherwise).

If xdpyinfo is right that's a pure firefox problem. Two possibilities:

1. you've been lying to firefox about your dpi all along to get small
fonts → now that autodetection is working the correct fix is to change
the default font sizes (in pixels) in firefox (and vote/comment for
https://bugzilla.mozilla.org/show_bug.cgi?id=414427 so you don't have to
enter font sizes in two places with two different units in the future).
I guess the ultimate test would be to create a page with font sizes
specified in inches or millimeters, and then check than when it's
rendered by firefox the physical sizes are right.

2. something else is broken → open a firefox bug

As others wrote the dpi situation is in a lot of flux right now. Not
because someone broke something but because our previous handling was
massively wrong (with various apps abusing dpi or even hardcoding it to
fixed values, and downstream apps adding bandaids over bandaids to
compensate), so fixing it involves changing apps behaviour and replacing
font defaults when they've been set up for false scaling factors values.
Also the Xorg people underestimated the importance of getting xdpi
computation right when the latest xrandr landed, so many xrandrified
drivers still get this wrong.

Nicolas Mailhot

Attachment: signature.asc
Description: Ceci est une partie de message numériquement signée

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]