Bug 304510 - Battery plasmoid does not show remaining time
Summary: Battery plasmoid does not show remaining time
Status: RESOLVED INTENTIONAL
Alias: None
Product: plasma4
Classification: Plasma
Component: widget-battery (show other bugs)
Version: 4.9.0
Platform: Ubuntu Linux
: NOR normal
Target Milestone: ---
Assignee: Plasma Bugs List
URL: http://i.imgur.com/6bZ5S.png
Keywords:
Depends on:
Blocks:
 
Reported: 2012-08-03 17:31 UTC by railmaniac
Modified: 2013-06-30 15:08 UTC (History)
19 users (show)

See Also:
Latest Commit:
Version Fixed In:


Attachments
Plasmoid battery with alignment lines on the right side (45.73 KB, image/jpeg)
2012-10-18 16:47 UTC, Igor Strelnikoff
Details
I think the battery plasmoid should look like (28.20 KB, image/jpeg)
2012-10-18 17:11 UTC, Igor Strelnikoff
Details
And even so, it greatly extended (74.91 KB, image/jpeg)
2012-10-19 14:11 UTC, Igor Strelnikoff
Details

Note You need to log in before you can comment on or make changes to this bug.
Description railmaniac 2012-08-03 17:31:23 UTC
The battery plasmoid shows charge information correctly, but does not show remaining time (refer screenshot). If I run `acpi` on konsole it shows remaining time. In previous version (4.8.4) battery plasmoid showed remaining time. Problem only started after installing 4.9.00.

Reproducible: Always

Steps to Reproduce:
1. Start KDE
2. Click on battery symbol in system tray
3. Battery plasmoid with battery info slides out
Actual Results:  
Battery plasmoid does not show remaining time

Expected Results:  
Battery plasmoid should have shown remaining time

uname -a

Linux core 3.2.0-27-generic #43-Ubuntu SMP Fri Jul 6 14:25:57 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
Comment 1 Sebastian Wessalowski 2012-08-05 12:52:04 UTC
I can confirm this.

Linux BlackBeauty 3.4.7 #1 SMP Tue Jul 31 04:18:49 CEST 2012 x86_64 Intel(R) Core(TM) i7-2640M CPU @ 2.80GHz GenuineIntel GNU/Linux

Its a Gentoo machine with 4.9.0 freshly built. i have even tried a fresh profile.
Comment 2 kordovan_ 2012-08-07 15:02:58 UTC
I can confirm this, too.
Arch Linux with Kernel 3.4.7 on a MacBook Pro 7,1
Comment 3 rail02000 2012-08-07 18:47:51 UTC
I can confirm this,too.

Kubuntu 12.04
Updated to KDE 4.9 via  ppa:kubuntu-ppa/backports

Linux Cheng-Laptop 3.5.0-7-generic #7~precise1-Ubuntu SMP Tue Jul 31 14:34:54 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux 
Running the system on a Acer Aspire 4741G
Comment 4 Eric Donkersloot 2012-08-14 21:10:40 UTC
Confirmed:

System:    Host: vaiopro Kernel: 3.4.8-1-ARCH x86_64 (64 bit) Desktop: KDE 4.9.00 Distro: Arch Linux
Comment 5 valdikss 2012-08-20 04:40:42 UTC
Confirmed.
ArchLinux x86_64 with kernel 3.5.2 on Lenovo ThinkPad X220.
Comment 7 railmaniac 2012-08-27 08:36:13 UTC
(In reply to comment #6)
> This functionality was removed:
> https://projects.kde.org/projects/kde/kde-workspace/repository/revisions/
> a6f8377ccdec5bf3a9a07c4f217583b7d18b7133

Thanks for the info. Any idea why this was done, seems to me a perfectly good feature was removed for no reason...
Comment 8 Nikita Skovoroda 2012-09-04 14:10:15 UTC
*** This bug has been confirmed by popular vote. ***
Comment 10 Igor Strelnikoff 2012-09-04 14:31:51 UTC
In KDE which is in default delivery opensuse 12.2 is also present
Comment 11 Igor Strelnikoff 2012-09-04 14:38:22 UTC
Please do not remove this functionality, it's worth it to finish so that the evidence was more correct.
Comment 12 Viranch Mehta 2012-09-08 15:17:42 UTC
The remaining time option was decided to be removed because it is very very inaccurate. It is shown and will be shown for users who chose to see it in previous versions of KDE (checked the "show remaining time" configuration option), but the option itself is removed from the configuration dialog for new version of the applet. There was a long discussion about this on developer mailing list and this is what they decided.
Comment 13 Viranch Mehta 2012-09-08 15:26:58 UTC
If anyone's interested, the discussion: http://osdir.com/ml/plasma-devel/2012-06/msg00300.html
Comment 14 Kevin Funk 2012-09-08 19:07:29 UTC
Sorry, but I seriously consider that the most stupid feature removal ever. At least you could have kept the configuration option in the dialog. For me it was accurate enough (and very useful) for multiple systems.

I'm not amused.
Comment 15 railmaniac 2012-09-12 08:38:02 UTC
Here seems to be the origin of the trouble: http://osdir.com/ml/plasma-devel/2009-05/msg00269.html

> when they do it varies according to workload that can and does shift wildly from minute to minute on most machines.

And you know what? That does not matter one bit. I don't really care whether the remaining battery time is 1 hour 51 minutes and 23 seconds or 2 hours 5 minutes and 13 seconds. All I want to know whether a) I have plenty of time or b) I need to charge now.
Comment 16 Sebastian Kügler 2012-09-12 13:12:16 UTC
You can have the remaining time shown by enabling "showRemainingTime=true" in the battery's config (usually to be found in plasma-desktop-appletsrc).

We will not add a visible config option for this (rationale is in the linked email threads), but for those that really want it, we've added this hidden config option.

I'm closing this bug as that's really all we want to do about this.
Comment 17 Viranch Mehta 2012-09-13 05:36:56 UTC
For those who really want this feature, I've patched the current plasmoid and put it up on kde-look.org under the name "Enhanced battery monitor". Users can download this via the usual "add widgets" -> "get new widgets" option and searching for the above name. This plasmoid has an option to show/hide the remaining battery time.
Comment 18 Tom Kijas 2012-09-18 10:44:42 UTC
Wow, great decision (joke). I simply NEED to know how long my battery will last, at least aproximately, e.g. to travelling. I think you must be mad, do you know you'll lose many people that use KDE?? Including me, I was with KDE from 4.1, reporting bugs etc. but this bu***t decision is enough. Absote ignorance of people's need, really...
Comment 19 Tom Kijas 2012-09-18 10:52:16 UTC
Tell me something, is it some kind of makeover about fact that KDE is more power-demanding than e.g. Unity? And that's a FACT in Macbook Air 2012. Or is it too old computer for you? I don't think so.
What feature will you remove then? Ability to reboot? Turn off pc? The only way is to run it FOREVEEER without rebooting, isn't it?
Something like Nautilus 3.6.
Comment 20 Tom Kijas 2012-09-18 10:53:01 UTC
And for interested - Ubuntu 12.04 KDE 4.9.1 64bit
Comment 21 Sebastian Kügler 2012-09-18 10:54:44 UTC
If you need to know, you surely don't want unreliable information, I'd think. You can enable the option in your config, or download Viranch's version from kde-look. It's really quite trivial. however.

Please try to keep posts here non-offensive.
Comment 22 Christian (Fuchs) 2012-09-22 00:49:27 UTC
This was already removed once in the past and brought back because users wanted it, now you removed it again which again brings up users wanting it. 

I can see absolutely no reason why to do this as a hidden config option. If the option is there, users should be able to enable it in a way they are used to, and use it at their own risk. In the end it should be the user who decides what information is valuable to him and whether it is useful or not.

If the developers decide that it is not of value, then just change the default to "not show it". But please do not patronize your users in such a bad way that you not only decide what is good for them, but also remove sane ways for them to change things back according to their preferences. 

Summary: I see valid reasons for changing the default value, but I see absolutely no reason to remove any visual configuration for this. Please think about this again. 

Kind regards, a long time KDE user and supporter.
Comment 23 Sebastian Kügler 2012-09-22 09:55:51 UTC
First of all,nothing changed. The option wasn't in the UI before, and it's still not in the UI, for good reasons. The option was never "brought back" to the UI in the first place, as it was never there (not in a released version, anyway). This bug is about a regression in a hidden config option.

I'm really puzzled why people rather complain pointlessly (pointlessly, because it adds no useful information) to this bug, rather than just changing their configs manually or picking the Plasmoid we've put on kde-look and using that. It's the exact reason why we have this architecture, so people can replace even default components according to their wishes.

Let me reiterate the point: We cannot expect from normal users to fully understand what affects power usage, it depends on the use, it depends on the device, and many devices out there are just plain broken.
Comment 24 Christian (Fuchs) 2012-09-22 13:19:39 UTC
(In reply to comment #23)
> First of all,nothing changed. 

Not entirely true. I don't know whether it is due to the fact that the plasmoid is a new one, which might have added a new configuration (or: a new ID to the plasmoid) or due to other reasons, but as the initial reporter I can confirm that upgrading KDE will in fact remove the display of the remaining time. So for the end user something did change. 

>The option wasn't in the UI before, and it's
> still not in the UI, for good reasons. The option was never "brought back"
> to the UI in the first place, as it was never there (not in a released
> version, anyway). This bug is about a regression in a hidden config option.

Which still makes it a valid bug. 

> I'm really puzzled why people rather complain pointlessly (pointlessly,
> because it adds no useful information)

Users complain because they are not happy with a part of your product, and it adds the information that users disagree with you on the decision whether this should be an option to have. You are more or less calling your users preferences pointless here. 

> to this bug, rather than just
> changing their configs manually 

Because having to edit a config file manually, which also can has funny side effects if doing it while plasma is running, is not really user friendly. And having to edit it in an environment with no plasma running is definitely not userfriendly.

> or picking the Plasmoid we've put on kde-look and using that.

Because it has a bug which makes it unusable. I would have reported it on bugs.kde.org, but it is not officially a part of kde, so it is now reported on said kde-apps page. If this think would be in upstream, we could be sure that it is looked after and updated along with KDE updates. Having a separate version on kde-apps.org means that either someone has to update that every time there is a change in the upstream plasmoid, or it will just one day not work properly. And it is harder to report bugs against.

> It's the exact reason why we have this
> architecture, so people can replace even default components according to
> their wishes. Let me reiterate the point: We cannot expect from normal users to fully
> understand what affects power usage, it depends on the use, it depends on
> the device, and many devices out there are just plain broken.

Given that argumentation: I am looking forward to see the remaining time in the file transfer plasmoid and dialog being removed in the next version of KDE, since all of what you say applies to it. 
http://xkcd.com/612/ And if you start to decide what features and options of KDE are fully understandable by your users, I see a lot more of them to be removed, similar to what e.g. GNOME did. Please don't go in this direction. 

This is why you could just change the default. If you think that it is too broken for your users to understand, just disable it by default. But please give users who don't agree with you there a sane possibility to add it back. 

In addition: Other environments that care a lot about what is complicated for users, such as OS X, do show the remaining time as well. And even though it was recently removed to show it right next to the battery applet in OSX, it is still shown on click. 

Also: If you think that the reported time is indeed not accurate enough: have a look at powertop. In addition to the current estimation, it gives out an estimate based on the average of your recent usage. So maybe this would be something good enough for you to show.  

Again, please think again whether this really should be a "WONTFIX". Note the confirmed by popular vote, note the users commenting on it. Please do the users who don' agree with you a favour and give them the possibility to (sanely) configure this option back. 

Kind regards, 

Christian
Comment 25 Sebastian Kügler 2012-09-22 18:50:15 UTC
Wether or not it's a new one should not matter (we're reading the same config we've written before). I just checked how the option was called in 4.8.5 and how it's now called, and the config key is exactly the same ("showRemainingTime"). From a code point of view, I don't know how this regression could have crept in, so in fact why this setting apparently got lost. (I don't say it wasn't lost, just that I don't see how or why.)

Your XKCD reference is one of the reasons why we removed it. We want to build trust with the user, and showing wildly varying information in our primary UI (which varies because we simply cannot predict the remaining time) does not build that kind of trust. That's a conscious design decision.

You're asking for an option to give users this functionality back: We did this in the following:
- we added a hidden config option
- we even added a plasmoid with this feature you want

I've had a look at powertop, not once but actually it's something I'm doing regularly in order to check if we have bugs creeping into our system that hampers battery life. I know exactly what powertop reports, and when it's accurate and when it's not. That's not something we expect from our average users, however. Non-average users might fully understand the meaning of this display, but then, above-average users are quite capable of replacing the plasmoid, or changing its config manually. (Doesn't mean they do not prefer lengthy and philosophical discussions about default settings, unfortunately.)

Let me also reiterate why we don't add a UI option,  disabled by default: Every piece of UI adds maintainance overhead on the one hand. Worse, it creates mental overhead for the user (a UI with less options is easier to understand than a UI with more options). To us, and we made a conscious decision as I've laid out, the added value does not justify the overhead on both accounts.

What's the bug in the plasmoid on kde-apps, btw? If it's not in the shipped version, we can easily fix it.

Please not that *all your arguments* have been discussed at length. There's no difference in input for our decision, so there won't be any difference in the result. If you want to understand the problem even better, the lengthy thread linked earlier in the comments is an excellent pointer.
Comment 26 Christian (Fuchs) 2012-09-22 19:04:32 UTC
(In reply to comment #25)
> [snip]

Yes, I am sure you did discuss this in lenght, however, I (among other users) am not happy with the resolution, since I don't see it as one. 

I mentioned the bug report on IRC today, and then tried to help a person who is rather tech savy in adding the option to the config file. Which again confirmed that it is _very_ far from user friendly.  (You first have to find a givens plugin configuration section in the config, then, if not already existing, add a generic config section for the same ID and then add the line. All of this in a non-plasma environment, since editing it while plasma is running can have nice side effects when the settings are edited by plasma as well). It took us quite a while, which imo is just too long for adding such a simple thing as remaining battery life. 

As all the other environments I am aware of do think that their users are "smart" enough to handle this information and as you are handing the same kind of inaccurate information in other places (note that the file transfer ETA is also dependant on the device, on other operations going on etc.) is something I really fail to understand. I did read through the thread, and if that really is the argumentation, then a lot of other things would have to be removed as well, the file transfer ETA being just the most prominent one. 

If you are feeling that bad about it, maybe add a warning when ticking the checkbox? I really think you should not be patronizing users too much in an environment which is known and loved by many for its reputation to be configurable to fit ones needs. It is what brought me to KDE, it is what held me at KDE and it is what made me support KDE financially and with other means. I really would see it as a great loss if you start removing functionality just because you think users are not able to manage it. 

> What's the bug in the plasmoid on kde-apps, btw? If it's not in the shipped
> version, we can easily fix it.

You probably can, yes. The diff doesn't look all that big. The problem is that the remaining percentage in the tooltip is never updated unless you restart plasma. The data engine below appears to give out the correct data, as it works well in the original plasmoid.

Thanks in advance if you fix it. However, as mentioned before, I am a bit afraid that moving this out to a non-official plasmoid which resides on kde-look/kde-apps  brings a few disadvantages. First of all it is really hard to report bugs to them, since kde-look/apps doesn't offer anything for that other than the comment system (which is imo not well suited for bug reports). Also changes made in upstream need to be added there every time as well, which means you have to take care of two things, which is exactly the overhead you mentioned in your reply. Just keeping a single config option in a qml plasmoid is imo less overhead. 

> Please not that *all your arguments* have been discussed at length. There's
> no difference in input for our decision, so there won't be any difference in
> the result. If you want to understand the problem even better, the lengthy
> thread linked earlier in the comments is an excellent pointer.

Yes. I might have sounded a bit more harsh than intended in here a few times, and I don't want to attack you or your colleagues, as you usually do very good work. But this bug reports, as others I had in the past, just give me the impression of some developers not listening to their user base, which in the end is their "customers". Summarized:  I still see no valid argument to put it in as a hidden config value, which are not really userfriendly especially in plasma. I see valid arguments to not set it on as the default, but then you'd theoretically also have to not display the ETA in file transfer dialogs per default. 

Kind regards

Christian
Comment 27 Sebastian Kügler 2012-09-22 19:23:20 UTC
Sorry, but:

- I've spent more than my personal allowance of time (and energy) on commenting here
- I've not seen new data come up in months
- We've provided multiple ways for those who complain here

I'm now going to unsubscribe from this bug, because I end up spending unreasonable amounts of my free time and energy to explain the same over and over again, without any apparent effect. I give up. :(
Comment 28 Viranch Mehta 2012-09-23 10:54:15 UTC
(In reply to comment #26)
> > What's the bug in the plasmoid on kde-apps, btw? If it's not in the shipped
> > version, we can easily fix it.
> 
> You probably can, yes. The diff doesn't look all that big. The problem is
> that the remaining percentage in the tooltip is never updated unless you
> restart plasma. The data engine below appears to give out the correct data,
> as it works well in the original plasmoid.
> 

FYI, this should now be fixed now.
Comment 29 Igor Strelnikoff 2012-10-18 16:45:12 UTC
OpenSUSE 12.2 KDE 4.8.5

Line items in the battery Aleta formatted on the right side of the applet which takes place and has a lot of extra space.

See attachment
Comment 30 Igor Strelnikoff 2012-10-18 16:47:16 UTC
Created attachment 74635 [details]
Plasmoid battery with alignment lines on the right side
Comment 31 Igor Strelnikoff 2012-10-18 17:11:26 UTC
Created attachment 74637 [details]
I think the battery plasmoid should look like
Comment 32 Viranch Mehta 2012-10-19 06:05:09 UTC
(In reply to comment #30)
> Created attachment 74635 [details]
> Plasmoid battery with alignment lines on the right side

This is the way it is intended to be. Plus, you seem to be using an older version of the plasmoid, though the newer version looks the same as far as the alignment you mention is concerned.
Comment 33 Igor Strelnikoff 2012-10-19 08:52:51 UTC
>This is the way it is intended to be. Plus, you seem to be using an older version of the plasmoid, though the newer version looks the same as far as the alignment you mention is concerned.

But it looks ugly
Comment 34 Igor Strelnikoff 2012-10-19 10:37:36 UTC
+33 You probably kidding me?
Comment 35 Viranch Mehta 2012-10-19 10:51:40 UTC
(In reply to comment #33)
> >This is the way it is intended to be. Plus, you seem to be using an older version of the plasmoid, though the newer version looks the same as far as the alignment you mention is concerned.
> 
> But it looks ugly

Its been like this since the beginning and unlikely to change. There's no scope of further discussion on this.
Comment 36 Igor Strelnikoff 2012-10-19 12:27:38 UTC
I do not even work setting "showRemainingTime = True" what could it be?
Comment 37 Igor Strelnikoff 2012-10-19 14:10:52 UTC
And why get such a difference in the length of the widgets on the two screenshots?

See attached
Comment 38 Igor Strelnikoff 2012-10-19 14:11:44 UTC
Created attachment 74652 [details]
And even so, it greatly extended
Comment 39 Viranch Mehta 2012-10-20 04:42:22 UTC
(In reply to comment #37)
> And why get such a difference in the length of the widgets on the two
> screenshots?

Because your openSUSE plasmoid is an older version.
Comment 40 Igor Strelnikoff 2012-10-20 08:39:27 UTC
>Because your openSUSE plasmoid is an older version.

Thanks for the information.
Comment 41 xor 2012-12-11 23:23:28 UTC
I can confirm this issue on Kubuntu12.10 amd64.
Comment 42 Tom Chiverton 2013-04-25 20:56:53 UTC
You can add time remaining / time to full via  https://bugs.kde.org/show_bug.cgi?id=304510 (though you have to patch it. I am considering a fork with the patch applied).

The most important thing for me (as people mention above) is if I can get stuff done now, or if I need to plug the charger in before I settle in for a few hours of SPAZ or something, So the '15%' low battery alert is no help; seeing I have 5 hours of time left at my current usage *is* though, because you quickly learn that'll be halved once you fire up Quake or whatever.

Tom
Comment 43 Tom Chiverton 2013-05-04 10:58:54 UTC
Oops, I meant to link to http://kde-apps.org/content/show.php?content=153981
Comment 44 zigi 2013-06-30 13:00:33 UTC
(In reply to comment #27)
> Sorry, but:
> 
> - I've spent more than my personal allowance of time (and energy) on
> commenting here
> - I've not seen new data come up in months
> - We've provided multiple ways for those who complain here

Multiple stupid options instead of one simple config checkbox. 

> 
> I'm now going to unsubscribe from this bug, because I end up spending
> unreasonable amounts of my free time and energy to explain the same over and
> over again, without any apparent effect. I give up. :(

Missing respect of users.
Comment 45 H.H. 2013-06-30 15:08:50 UTC
(In reply to comment #25)

> What's the bug in the plasmoid on kde-apps, btw? If it's not in the shipped version, we can easily fix it.

I think it would be a good choice, to ship the alternative plasmoid from kde-apps with kde, as an option for the user to choose from. although the maintainance of a second plasmoid instead of a checkbox is questionable.