[|it|Nvb]

tried it for a bit and have to say it runs very nice and smooth.

seems like i found a bug though (and quite an annoying one :<), hopefully someone can reproduce.

i use a config based on nexu's with fovswitch/zoomswitch bound on one key. once that particular button is pressed and released mouse sensitivity is like 40% lower than before, though unchanged if you check in console.

steps to reproduce:
1)
/cg_fov 110
2)
/sensitivity // check sensitivity
// do 180 turn and check mouse travel
3)
/cg_fov 90
// here it breaks already
// do 180 turn and check mouse travel
/sensitivity // make sure its unchanged
/sensitivity 0.0001 // just to reset broken sensitivity
/sensitivity <xy> // put your sens for step 2 in here
// do 180 turn, check mouse travel and notice the difference

same thing happens if you switch back to fov 110, and sadly also with zoomin / out

[mitsubishi]

beast, on 04 February 2010 - 09:55 PM, said:

use .i386

it's for simplicity.


nvb, this may be a bug with /zoomsensitivity (something this added) if you try to do the same without it with scripts; I'll try to see if it's that, and if it is, for a way to let them both work..

ok i may have a solution soon..

This post has been edited by mitsubishi: 05 February 2010 - 01:19 AM

[mitsubishi]

ok, check the update(d file), it fixes an issue with /zoomsensitivity; it may be fixing your issue:

The way this previously worked had the initial sensitivity in memory for use on zooming for the duration of the client run; this may understandably produce quirky behavior, especially if one uses scripts that set sensitivity on the fly for zooming (without /zoomsensitivity).


thx for that feedback, that was a windows issue too.

This post has been edited by mitsubishi: 05 February 2010 - 11:38 AM

[Scan (old)]

Not a big deal but I can't alt-tab the game and when I alt-enter it just gets into a loop. Other than that it seems to work great. (still using your first build) Thanks again

[mitsubishi]

that's probably windows manager related; here i tried it on an ubuntu with gnome and it went normally to window mode with alt+enter. then with the console down, desktop could be accessed.

i suppose it may be related to sdl too.


similar behavior on kde.


hm, i guess it can be done explicitly by sdl..


ah we may have something..

ah, meh.. the minimize function of SDL appears to do anything only on windows; i could double check but i doubt it.

what they previously did and potentially worked might be interesting.

This post has been edited by mitsubishi: 05 February 2010 - 03:07 PM

[|it|Nvb]

mitsubishi, on 05 February 2010 - 01:29 AM, said:

ok, check the update(d file), it fixes an issue with /zoomsensitivity; it may be fixing your issue:

yes indeed, that fixed it, cheers :>

[mitsubishi]

ok, 1) i will have a 64bit binary soon,
2) the sdl error on first reply may be solved on 64bit linux with a ia32-libs package, but that may not be needed with an 64bin.

[mitsubishi]

ok, first of all, we have an 64 bit binary; spectacular fact is that this is faster than windows here at least (windows is only on 32bit binaries btw).

second, there's now /minimize which minimizes the client (can also be bound to a key, e.g. bind x minimize)
but also alt-tab explicitly tells SDL to minimize the window.

hence alt-tab can minimize it now on most window managers.

This post has been edited by mitsubishi: 05 February 2010 - 07:04 PM

[Scan (old)]

mitsubishi, on 05 February 2010 - 05:42 PM, said:

ok, 1) i will have a 64bit binary soon,
2) the sdl error on first reply may be solved on 64bit linux with a ia32-libs package, but that may not be needed with an 64bin.

64bit would be great, can you build it with r_smp enabled? not really sure that it would help anything but can't see that it would hurt to have the option to r_smp 1. Thanks

[mitsubishi]

64bit is here; check the last reply

mitsubishi, on 05 February 2010 - 07:04 PM, said:

ok, first of all, we have an 64 bit binary; spectacular fact is that this is faster than windows here at least (windows is only on 32bit binaries btw).

second, there's now /minimize which minimizes the client (can also be bound to a key, e.g. bind x minimize)
but also alt-tab explicitly tells SDL to minimize the window.

hence alt-tab can minimize it now on most window managers.

r_smp is buggy and i think it's enabled (available to use); it usually gives irritating artifacts on screen and ioq3 developers believe it doesn't do anything good on new hardware (in that form at least).

This post has been edited by mitsubishi: 05 February 2010 - 07:14 PM