Popout and popups - Feature requests

[panos_k75]

i hope that's not 4 seconds!

[Dean Roddey]

The timing on the popouts is about 2 seconds. If it takes longer than that, then your machine isn't up to the task of doing it at optimal speed. It can take a little longer because it's waiting for the vertical retrace period before blitting each new round. If you are at 60hz, it might have to wait up to a 60th of a second (worst case) for the next vertical retrace after it has another one ready to go.

So, worst case, if it was big enough to require 60 rounds, and it hit a fairly worst case timing relative to the vertical retrace each time, that could add another second to it. Though as a rule it would generally only add half a second on average. If your retrace is higher, then there are more of them and the average delay for the retrace will be lower. And if your machine is fast enough, it'll always have data available at the next available time for it to spit out another image (it times them so that it's not dependent on machine speed.)

[Wynn]

Just so happens that I have a stop watch here at work. This isn't what I would call scientific or anything, but it looks like my 4 count equates to about 2 real seconds (actually 1.81 on the stop watch). So, its in the ballpark of what Dean says here.

[Dean Roddey]

Actually, you are correct. The internal cycle timing is 1 second, not 2. So add in a little overhead plus the average waiting for the vertical retrace on each round, and 1.5'ish seconds (a little more or less based on timing and machine speed and size of the popout in the direction of travael) is about what would be expected.

[panos_k75]

2 seconds for the Q1 is not too bad...Wynn could you please create a bigger transparent popout - something like 480x200 for example - and see how the Q1 handles this? I would really appreciate that.
Thanks,
Panos

[Wynn]

Well, that was weird. I made a 480x200 translucent image, same number of buttons, and tested it again. This time there was no noticeable difference in the Q1 to my main work machine. While the first test yesterday had the slightest hint of jitter on the Q1, this one was silky smooth. Not sure if my computers made better connections to the wireless network today or what (I turned all of them off last night).

[Dean Roddey]

It will depend sometimes on what else is going on on the machine. And making it bigger along the transverse direction (i.e. the axis it's not sliding on) won't have nearly as much impact as making it bigger along the axis that it has to slide out along.

[panos_k75]

Originally Posted by Wynn

Well, that was weird. I made a 480x200 translucent image, same number of buttons, and tested it again. This time there was no noticeable difference in the Q1 to my main work machine. While the first test yesterday had the slightest hint of jitter on the Q1, this one was silky smooth. Not sure if my computers made better connections to the wireless network today or what (I turned all of them off last night).

Wynn, do you have the Celeron version of the Q1? There is also a Pentium and a VIA one too.