"Komag" wrote:
Finally got around to implementing the speed test, EnTerr. My friend's 2050X gets 40-42ms, not 74. So I've set the cut-off point at 30ms, seems to work well, thanks. 🙂 I'll of course monitor/adjust this as I test across more platforms.
Oh, crud! (To quote Philip J. Fry)
You are right, it is 40ms on 2xxx models!
Thanks for mentioning, when i saw your numbers i wondered whence the difference and got an idea - that i made a mistake: screensaver was running on it while i tested and was slowing things down. So i woke up said Roku and re-tested - it is ~40ms indeed.
With that in mind, the cut-off margin should be ~27. Now onto the entertaining section of this post: the previous margin of 36 i suggested survived a single mistake of 2x of me measuring! I.e. it still correctly distinguished between slow and fast models, albeit barely. Amazing. Now if i were double-sloppy and mis-measured 30xx in screensaver too, then i would've gotten cut-off of 52 or so, which would have screwed up the pooch finally.
PS. and the reason i did not mis-measure twice might have been that screen saver does not get activated on 30xx when roScreen was in use (instead a black screen is shown)... but same screensaver runs on 2xxx models in roScreen mode? How about them apples.
