Re: [AD] X11 unresponsiveness

[Chris <kcat@xxxxxxxxxx>]

Evert Glebbeek wrote:
> In effect, it would change behavior for a supporte platform. Which is 
> bad.

While I'll concede the issue at hand, I can't say I agree with this 
statement. A change in behavior for a suported platform should not be an 
issue as long as it continues to behave as Allegro allows, and that the 
benefits of changing it are worth it.

> The threaded version is already the default version build if pthreads is 
> installed. I don't think we need to take more active steps to encourage 
> its use over the sigalrm version.

I don't think including a note in the Unix install docs saying that the 
sigalrm version shouldn't be used unless necessary would be a bad thing. 
Just because pthreads is the default version doesn't mean Unix users 
won't want to try and use the sigalrm version for various reasons.

> In which case the needs_poll() family of functions should be changed to act 
> sensibly instead of turning on polling for platforms that don't need it.

I don't think the *_needs_poll() functions change anything.. they just 
report what the driver wants. It's when you actually call the poll 
function itself that modifies behavior. The docs say that when you first 
poll the device, you must continue to poll it for it to continue working.

- Kitty Cat