|
Progress bars that reach 100% and then stay there for ages while the application completes whatever operations have to happen after all the operations the devs actually accounted for in the progress percentage calculation logic are complete
|
# ? Aug 19, 2019 14:44 |
|
|
# ? Jun 8, 2024 04:00 |
|
At one point the standard was that the number on the progress bar should be what percentage would be done after the current step. Some people never moved on.
|
# ? Aug 19, 2019 14:54 |
|
You might say they never progressed
|
# ? Aug 19, 2019 15:08 |
|
ultrafilter posted:At one point the standard was that the number on the progress bar should be what percentage would be done after the current step. What? You're pulling my leg, there's no way anyone could possibly think that's a sensible thing for a progress display to mean.
|
# ? Aug 19, 2019 15:30 |
|
imagine you were displaying progress through discrete named steps. wouldn't you tell the user that you were reticulating splines while doing so, not after? now imagine you also included the number of the step in the display. now remove the step name. it's wrong to display percentages that way, but i can see why someone wrongly thought it was sensible. e: use word more good Doc Hawkins fucked around with this message at 16:07 on Aug 19, 2019 |
# ? Aug 19, 2019 16:04 |
|
Unless you can reliably estimate when the task will finish, possibly using some method to learn based off previous attempts, progress bars are always going to be unsatisfying.
|
# ? Aug 19, 2019 19:36 |
|
The point of progress bars is so the user can see that progress is being made. This is also why the “progress bars” that just animate back and forward are worthless, because the only information they convey is that the animation is working
|
# ? Aug 19, 2019 19:50 |
|
The animated gif progress bars make people think that things are getting done faster despite the complete lack of information.
|
# ? Aug 19, 2019 20:31 |
|
My favorite case of progress bars: A long, long while back I was working for a company and we went for a government RFP. We lost it to a competitor, but part of the spec was a requirement that the software offer an open API for third-party systems to exchange certain bits of regulatory data. Thus, we got access to a testbed system, via their web frontend. They had a login window and once you'd entered your credentials it would change and display a lot of status messages and a progress bar. Encrypting Login Details... Securely Transmitting Login... Setting Up Secure Interface... That sort of nonsense that you'd use if you wanted to convince a bunch of bureaucrats that some sort of high-tech security magic was being cast. F12 revealed a Javascript array of messages that was iterated over with a 10 second pause in between each one, before the form was finally submitted. They had added something like 70 seconds of delay to the process - I suspect in part to artificially throttle usage, but also in part to do the classic 'oh poo poo we need a winning release' trick where they have prebuilt headspace to magically 'speed things up'
|
# ? Aug 19, 2019 20:38 |
|
A long time ago I encountered a web API for a supercomputing portal with a box asking "Number of Processes?" but figured out that its actual meaning was "Number of cores to assign per process?" So everyone using this interface was being way over-charged, e.g. someone would enter "16" to use up all 16 cores of a node with 16 processes, but instead would wind up with just 1 process running on each node and 16x as many submissions overall. And despite all of this stuff just sitting on top of PBS, for whatever stupid reason the web API would only allow you to request a number of nodes, not a number of cores. I can see and download the submission scripts that you're generating you incompetent idiots, I know what you're doing! Since SSH access was forbidden, I came up with a simple workaround: submit a shell script that submits new PBS jobs with whatever options that you actually want. QuarkJets fucked around with this message at 20:50 on Aug 19, 2019 |
# ? Aug 19, 2019 20:46 |
|
QuarkJets posted:A long time ago I encountered a web API for a supercomputing portal with a box asking "Number of Processes?" but figured out that its actual meaning was "Number of cores to assign per process?" And despite all of this stuff just sitting on top of PBS, for whatever stupid reason the web API would only allow you to request a number of nodes, not a number of cores. I can see and download the submission scripts that you're generating you incompetent idiots, I know what you're doing! Wait... the web API allowed shell scripts to be submitted? That seems pretty bonkers to me.
|
# ? Aug 19, 2019 20:55 |
|
1337JiveTurkey posted:Unless you can reliably estimate when the task will finish, possibly using some method to learn based off previous attempts, progress bars are always going to be unsatisfying. But my point is that it's infuriating to see a progress bar sit there saying it's "100% complete" for 10 minutes. If it was 100% complete you'd have hosed off already! Don't lie to me.
|
# ? Aug 19, 2019 21:48 |
|
How hard is it to make the progress bar top out at 99? Amateurs.
|
# ? Aug 19, 2019 21:50 |
|
taqueso posted:How hard is it to make the progress bar top out at 99? Amateurs. We have our new Fizzbuzz.
|
# ? Aug 19, 2019 23:23 |
|
bigmandan posted:Wait... the web API allowed shell scripts to be submitted? That seems pretty bonkers to me. It's an HPC portal, so the user has an account on the system already
|
# ? Aug 20, 2019 00:38 |
|
you hate to see it, folks
|
# ? Aug 20, 2019 02:56 |
|
taqueso posted:How hard is it to make the progress bar top out at 99? Amateurs. Years ago before our software actually tracked progress, we would start at 0, then go to 1/2, then 2/3, then 3/4 and so on, every few seconds until the operation was complete. Whenever it actually completed, it would jump to 100%. Only one client called us out on it.
|
# ? Aug 20, 2019 03:36 |
|
rarbatrol posted:Years ago before our software actually tracked progress, we would start at 0, then go to 1/2, then 2/3, then 3/4 and so on, every few seconds until the operation was complete. Whenever it actually completed, it would jump to 100%. Only one client called us out on it. The real power move is to have an estimated time to completion and just adjust it according to speed.
|
# ? Aug 20, 2019 03:41 |
|
Zeno's progress bar
|
# ? Aug 20, 2019 03:41 |
|
It can be extremely hard to predict when splines have finished reticulating.
|
# ? Aug 20, 2019 07:07 |
|
i loving hate those cutesy status messages. both slack and discord have sombre infuriatingly meaningless gibberish when they load for no fuckin reason
|
# ? Aug 20, 2019 10:28 |
|
redleader posted:i loving hate those cutesy status messages. both slack and discord have sombre infuriatingly meaningless gibberish when they load for no fuckin reason Those are annoying but nothing compared to cutesy error messages. "Oopsy woopsy! A random elephant ran into our server room while we were processing your payment! Don't worry our hilariously random pop culture reference crew are holding sporks! oWo!"
|
# ? Aug 20, 2019 10:58 |
|
everyone that touched the code, it is the only way to be sure.
|
# ? Aug 20, 2019 11:13 |
|
redleader posted:i loving hate those cutesy status messages. both slack and discord have sombre infuriatingly meaningless gibberish when they load for no fuckin reason Slack's is user settable, so your admin can let you see every time you join just how disappointed they are in you.
|
# ? Aug 20, 2019 14:10 |
|
redleader posted:i loving hate those cutesy status messages. both slack and discord have sombre infuriatingly meaningless gibberish when they load for no fuckin reason Of all the things to hate...
|
# ? Aug 20, 2019 15:17 |
|
"Shall we try to use less than 50% of the machine's available resources for our chat application?" "No, let's add some more nonsense instead! This is fun and more agile and will boost ~engagement~"
|
# ? Aug 20, 2019 15:23 |
|
Sagacity posted:"Shall we try to use less than 50% of the machine's available resources for our chat application?" "No, let's add some more nonsense instead! This is fun and more agile and will boost ~engagement~" And you know what? They're right. Nobody gives a poo poo about the consumed resources. They're just looking for a reason to buy the latest and greatest gadget. All they want is cutesy messages and cat pictures. Developers, technical people, the very folks on this forum, are guilty of this as well. Many moons ago I've ranted about Slack and how awful it is. I was very quickly reminded how important is the multimedia integration in a conversation with another person. The ability to embed emoticons, pictures, videos or web pages is of paramount importance. So no, "shall we try to optimize crap" is not in the vocabulary. Therefore, since we do have to work on something, to justify our waiting to vest time, we add nonsense.
|
# ? Aug 20, 2019 15:31 |
|
Volguus posted:And you know what? They're right. Nobody gives a poo poo about the consumed resources. They're just looking for a reason to buy the latest and greatest gadget. All they want is cutesy messages and cat pictures. Developers, technical people, the very folks on this forum, are guilty of this as well. Your last sentence is in conflict with the rest of your rant. You're adding stuff because consumers like it, not to give yourself work to do. I'm constantly amazed by how much disdain many developers have for the end users of their products.
|
# ? Aug 20, 2019 15:41 |
|
Why does a glorified IRC application even have load time, anyway? *fires up BitchX*
|
# ? Aug 20, 2019 15:45 |
|
Thermopyle posted:Your last sentence is in conflict with the rest of your rant. I don't think "Therefore, since we do have to work on something, to justify our waiting to vest time, we add nonsense." contradicts anything. They work on the product. They do not optimize it though, they add crap. Nonsense. Silly texts. Cute pictures. Emoticons. You know, the important poo poo.
|
# ? Aug 20, 2019 15:49 |
|
Ya’ll know they just did a massive optimization release?
|
# ? Aug 20, 2019 16:08 |
|
It is still poo poo. The optimization mostly applies to people who have tons of workspaces. My favourite "Slack consumes too many resources" story is that I was once running 2 Linux VMs on my notebook at once, running chrome with ~40 tabs and a full-hd stream in the foreground and Slack was still the most memory-consuming application
|
# ? Aug 20, 2019 16:55 |
|
lobsterminator posted:Those are annoying but nothing compared to cutesy error messages.
|
# ? Aug 20, 2019 17:13 |
|
Just one of the many reasons these forums suck
|
# ? Aug 20, 2019 17:24 |
|
Xarn posted:everyone that touched Harsh but it's also a fair cop.
|
# ? Aug 20, 2019 17:52 |
|
Riven posted:Ya’ll know they just did a massive optimization release? It is actually much much better, but it used to be so loving bad. That equals out to approximately .
|
# ? Aug 20, 2019 18:21 |
|
Riven posted:Ya’ll know they just did a massive optimization release? Is this the fourth rewrite of a three-platform "multi-platform" app?
|
# ? Aug 20, 2019 18:40 |
|
pokeyman posted:Is this the fourth rewrite of a three-platform "multi-platform" app? It's a webapp on each one
|
# ? Aug 20, 2019 19:08 |
|
did Lowtax ever get his trees back?
|
# ? Aug 20, 2019 19:39 |
|
|
# ? Jun 8, 2024 04:00 |
|
Volguus posted:And you know what? They're right. Nobody gives a poo poo about the consumed resources. They're just looking for a reason to buy the latest and greatest gadget. All they want is cutesy messages and cat pictures. Developers, technical people, the very folks on this forum, are guilty of this as well. Slack's most important feature for driving user adoption is the fact that any user can add new emoticons, it isn't locked down to admins only. I'm 100% serious. To be fair Slack also syncs conversation state across multiple devices and seems to be able to scale to tens/hundreds of thousands of users without falling down, two critical features neither IRC nor other chat clients can seem to manage. None of that would matter if people hated it, hence the emoticons.
|
# ? Aug 21, 2019 01:10 |