Instead of the stupid time estimates, which we cannot grasp, have the following:

a. A square with area depicting amount of bytes to download. On top of box it says: so and so GB or MB etc
A square under the area, showing how fast current download is going: At Max, At Min, for longest period (or maybe some other statistic), and Average.

Each addition in time has a border around its size, so we understand how progress is going. The area could be a stripe like now, but if the units are too small to be shown, perhaps it can dynamically widen. The time units can vary accroding to the time: If its 3 hours to download, we could make 10 minute blocks, and put a smaller 10 minute entry devided into 1 minute sections, so that you see something happening...
-- pashute, Jun 28 2008

So, like a grid of blocks that fill in as parts are downloaded? Most BitTorrent clients do this. I agree with having min/max/average speed displayed though.
-- Bukkakinator, Jun 28 2008


Windows' defrag used to work like this too.
-- phoenix, Jun 28 2008


The same errors that confound a numeric representation would confound the graphic representation.
-- 4whom, Jun 29 2008


How about a scrolling line graph showing progress vs time? To allow for processes that might take awhile, there could be a 1x line and a 10x line (possibly 100x as well). During the operation, the 1x line would go from the bottom to the top once; the 10x line would go from bottom to top, wrap, bottom to top, wrap, etc. a total of ten times. If it's taking one minute for the 10x line to make each bottom-to-top trip and the 1x line is halfway up, that would suggest that the operation will take about five minutes more.

To allow for operations that might take widely divergent lengths of time, there could be separate or overlaid plots for 0.1 minute, 1 minute, and 10 minutes, or the time scale could be logarithmic, or there could be manual sizing controls. The fundamental aspect of the design would be the use of 1x/10x/100x curves representing fractional completion.
-- supercat, Jun 30 2008


no the same errors would NOT confound the graphic rep, because when you see the "signal strenth", what has actually been accomplished and how it is being accomplished along time, you get the correct idea of what is happening... ie. if at the beginning you get to big chunks of data, (2/10ths of the info) and then go down to getting 2/100ths in the same time, you realize that very little progress is happening, and that you have a long wait. When the signal picks up again, you'll immediately realize whats happening.
-- pashute, Jul 06 2008


Have you ever thought about downloading smaller things, closer to you?

The number one way of ridding oneself of frustration, is to avoid the frustrating.
-- mylodon, Jul 07 2008



random, halfbakery