Talk:Progress bar

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

What is progress?[edit]

This stub of an article says that a progress bar conveys the progress of a task. But it does not define progress. If we can't define progress, then we can't define what the progress bar displays or why it is there. This is not a trivial point. For certain cases such as downloads, a progress bar could display the percentage of the file that the computer has received. It's simple math. Assuming that the transfer rate is relatively consistent, the progress indicator could be generally relevant.
But many other activities display progress bars. For example, an install program might give you a progress bar that displays the percent done. But what does that mean? Suppose the installer has to perform five tasks. One of them involves copying files and the total bytes are a known factor. One involves generating data, and another downloads and applies updates. The first two create directories and other trivial tasks. The total time it takes would depend on many factors, including processor speed, download speed, and hard disk throughput. So it might be possible to compute the expected total time, and extrapolate an estimate for each step based on the anticipated percent of time and the monitoring of ongoing activity. But a developer could just as easily decide that each step is 20%, the copying step advances based on the percentage of files copied, and the download will use similar logic. Then users see that the install is 40% finished within a few seconds, and the rest takes 45 minutes. In what sense should that 40% done have meant anything to the user? It tells nothing relevant, gives the user no sense of how much time is remaining, or even what tasks are remaining.
The bottom line is that if progress is not defined, then the progress bar shows nothing more than the fact that something has started, or is at some random point between start and finish. If it's truly 100% done, the progress bar would have gone away, even though it's not uncommon to see progress bars get to 100% and linger for extended periods. So without defining progress, there's no theoretical difference between a progress bar and an hour glass in terms of conveying information. There are progress bars that do not move uniformly, and even stop for extended periods. So a progress bar is not necessarily an indication that something is happening or not happening.
Is there any attempt within the world of computer science to define the topic of measuring progress of an activity in terms of percent completed? —Preceding unsigned comment added by Hagrinas (talkcontribs) 21:01, 14 August 2008 (UTC)[reply]

Define progress as the amount of elapsed time divided by the expected time of the entire task. Of course the estimated time required for the entire task is unknown to the user. It may even change dynamically as the task is being completed. None of this renders the progress bar meaningless. Alpha Omicron (talk) 00:39, 1 December 2008 (UTC)[reply]
You are raising valid points, but only with relevance to programming. i.e. "How should a programmer use progress bars properly?" I believe that a styleguide for programmers is beyond the scope of this article. --BjKa (talk) 09:58, 27 April 2018 (UTC)[reply]

External Links[edit]

All of the external links are old sites that are now domain-squatted. —Preceding unsigned comment added by Danomagnum (talkcontribs) 18:22, 18 January 2009 (UTC)[reply]

mixed up[edit]

I think the article is mixed up. A progress bar shows how much is complete by a percentage, not a throbber. The article says that a progress bar shows an action is taking place. the template "{cleanup-reorganize}" should be placed. -Jasonxu98 —Preceding unsigned comment added by 69.136.72.16 (talk) 22:27, 12 December 2009 (UTC)[reply]

It seems that this has been resolved by now. --BjKa (talk) 09:58, 27 April 2018 (UTC)[reply]

History[edit]

In 1984 I was working on the Beta-phase circuit boards for Sequent Computer Systems. The SCED board tended to hang during the self-tests and it was very hard to know what was happening. I talked to the firmware engineer and suggested that he have it print a dot on the monitor at some regular interval. We ended up with a dot at the end of each sub-test and occasional dots at other specific times (spread out through the longer tests, for example.) The idea was to let us know that the board was still working, but we soon learned to count the dots and know where the tests were failing, too. I had never heard of or seen such a thing before, but it seems unlikely that I am the first one to think of this. I would be interested to see when the first actual use of a progress display was made on computers. (In the *really* old days you knew the progress by seeing how far it was through the stack of punch cards, or how far around the paper tape loop the boot loader got.  :-) User:FatBear1, 17 May 2010‎

It would be great if we could add something to the "History" section like "Even before graphical user interfaces became widespread, some programmers were using things like a printer printing dots at certain intervals to judge the progress of a computing task." Only someone might ask for a source as proof... --BjKa (talk) 09:58, 27 April 2018 (UTC)[reply]