dial down update frequency for progress display to 0.2s

0.1s seemed a bit too jumpy. But kept 0.1s for --json-progress in case
some consumer depends on it.

BTW, the way rsync handles it is one progress update after every chunk
the rolling checksum algo identifies. So it's not a fixed delay. Weird.
This commit is contained in:
Joey Hess 2018-03-13 18:32:18 -04:00
parent 050ada746f
commit d8702c5259
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38

View file

@ -39,7 +39,7 @@ metered othermeter key getsrcfile a = withMessageState $ \st ->
showOutput
meter <- liftIO $ mkMeter msize $
displayMeterHandle stdout bandwidthMeter
m <- liftIO $ rateLimitMeterUpdate 0.1 meter $
m <- liftIO $ rateLimitMeterUpdate 0.2 meter $
updateMeter meter
r <- a meter (combinemeter m)
liftIO $ clearMeterHandle meter stdout
@ -50,7 +50,7 @@ metered othermeter key getsrcfile a = withMessageState $ \st ->
meter <- liftIO $ mkMeter msize $ \_ msize' old new ->
let s = bandwidthMeter msize' old new
in Regions.setConsoleRegion r ('\n' : s)
m <- liftIO $ rateLimitMeterUpdate 0.1 meter $
m <- liftIO $ rateLimitMeterUpdate 0.2 meter $
updateMeter meter
a meter (combinemeter m)
#else