Active Topics

 


Reply
Thread Tools
nthn's Avatar
Posts: 764 | Thanked: 2,888 times | Joined on Jun 2014
#81
Originally Posted by Bundyo View Post
Unfortunately this is true - when there is not enoughe memory, the app gets killed in the background, but the cover stays there and showing a slightly blurry screenshot. One time, while I was doing something in MC in terminal, the app got killed while it was focused It automatically transitioned to the task manager (still showing MC in it) and when I clicked it, a brand new terminal was spawn in its place.

I really hope this is going to change in the final release.
How should it change? What difference in a negative way do you perceive compared to previous Sailfish versions?

Edit: of course apps being killed while in the foreground should never happen, but this in itself has nothing to do with the covers.

Last edited by nthn; 2015-09-07 at 16:13.
 
Posts: 752 | Thanked: 2,808 times | Joined on Jan 2011 @ Czech Republic
#82
Originally Posted by nthn View Post
How should it change? What difference in a negative way do you perceive compared to previous Sailfish versions?

Edit: of course apps being killed while in the foreground should never happen, but this in itself has nothing to do with the covers.
I (we) understand what you mean, so you should spare multiple posts.

This is what I think: When OOM kills an app on a phone with 1GB RAM (when no such things would happen years ago on a phone with 256 RAM), then it is a big problem. I would rather have this problem solved with optimization, than hiding the problem by keeping its cover.

So yes, I would like to see this functionality reverted before final Sailfish UI 2.0 comes out - if an app crashes, then I want to see it and deal with the problem, not have it hidden like on Android and other non-multitasking operating systems.

EDIT: I understand that keeping its cover helps keeping other covers in their positions, but I think that this functionality can be easily misused as a mean to have Android-like "single-tasking".
 

The Following 5 Users Say Thank You to nodevel For This Useful Post:
Guest | Posts: n/a | Thanked: 0 times | Joined on
#83
Originally Posted by nodevel View Post
I (we) understand what you mean, so you should spare multiple posts.

This is what I think: When OOM kills an app on a phone with 1GB RAM (when no such things would happen years ago on a phone with 256 RAM), then it is a big problem. I would rather have this problem solved with optimization, than hiding the problem by keeping its cover.

So yes, I would like to see this functionality reverted before final Sailfish UI 2.0 comes out - if an app crashes, then I want to see it and deal with the problem, not have it hidden like on Android and other non-multitasking operating systems.

EDIT: I understand that keeping its cover helps keeping other covers in their positions, but I think that this functionality can be easily misused as a mean to have Android-like "single-tasking".
While great, this doesn't take into consideration code bloat. From my understanding, Jolla/Sailfish is highly optimized, but people just rarely code tightly now.

So from my understanding, this is a dislike on how multi-tasking apps act once either: RAM runs out or upon error, how that app is handled by the OS. The way Android does it, isn't liked, I get that. Until we get to see how Sailfish handles two apps on the screen at once, I'd say that you're still stuck with serial tasking in most apps that are sent to the background anyway since you're not truly interfacing them at the same time.

That's my take, probably inaccurate but as far as it stands, this is not really a big issue unless this affects native audio & video apps that are in the background.
 

The Following 2 Users Say Thank You to For This Useful Post:
Posts: 1,548 | Thanked: 7,510 times | Joined on Apr 2010 @ Czech Republic
#84
Originally Posted by nodevel View Post
if an app crashes, then I want to see it and deal with the problem
Well, if an app crashes a crash handler (such as ABRT used in Fedora/RHEL/CentOS) should show up and ask the user if he wants to report the crash (user writes what caused the crash and confirms what logs to send), so that the application developers know about the crash and can do something about it.

Much better for all parties than having to report all crashes manually.
__________________
modRana: a flexible GPS navigation system
Mieru: a flexible manga and comic book reader
Universal Components - a solution for native looking yet component set independent QML appliactions (QtQuick Controls 2 & Silica supported as backends)
 

The Following 9 Users Say Thank You to MartinK For This Useful Post:
nthn's Avatar
Posts: 764 | Thanked: 2,888 times | Joined on Jun 2014
#85
Originally Posted by nodevel View Post
I (we) understand what you mean, so you should spare multiple posts.

This is what I think: When OOM kills an app on a phone with 1GB RAM (when no such things would happen years ago on a phone with 256 RAM), then it is a big problem. I would rather have this problem solved with optimization, than hiding the problem by keeping its cover.

So yes, I would like to see this functionality reverted before final Sailfish UI 2.0 comes out - if an app crashes, then I want to see it and deal with the problem, not have it hidden like on Android and other non-multitasking operating systems.

EDIT: I understand that keeping its cover helps keeping other covers in their positions, but I think that this functionality can be easily misused as a mean to have Android-like "single-tasking".
Software grows larger as time goes on. You cannot honestly compare things from years ago that ran well on a low amount of resources with things now. Sure, you can program it all in assembly and you'll have an entire OS + all applications running on 4MB RAM, but there's a reason this doesn't happen. If an application takes up a lot of memory, there are two reasons, either it is not optimised enough or it is optimised but simply has lots of functions. The only solution to apps being killed by OOM (again, this has been the case before 2.0 as well, I fail to understand why it's suddenly an issue again, and why it has to be 'solved' [what would that even entail?] before 2.0 is released to the general public) is to optimise them.

Edit: also, this has nothing to do with apps crashing. If that happens, the cover disappears!
 

The Following User Says Thank You to nthn For This Useful Post:
Bundyo's Avatar
Posts: 4,708 | Thanked: 4,649 times | Joined on Oct 2007 @ Bulgaria
#86
On the plus side the browser is accelerated and lightning fast
__________________
Technically, there are three determinate states the cat could be in: Alive, Dead, and Bloody Furious.
 

The Following 8 Users Say Thank You to Bundyo For This Useful Post:
Bundyo's Avatar
Posts: 4,708 | Thanked: 4,649 times | Joined on Oct 2007 @ Bulgaria
#87
Originally Posted by nthn View Post
How should it change? What difference in a negative way do you perceive compared to previous Sailfish versions?

Edit: of course apps being killed while in the foreground should never happen, but this in itself has nothing to do with the covers.
The expectation that what you left is still there.
__________________
Technically, there are three determinate states the cat could be in: Alive, Dead, and Bloody Furious.
 

The Following User Says Thank You to Bundyo For This Useful Post:
pichlo's Avatar
Posts: 6,447 | Thanked: 20,981 times | Joined on Sep 2012 @ UK
#88
Originally Posted by nthn View Post
Software grows larger as time goes on. You cannot honestly compare things from years ago that ran well on a low amount of resources with things now. Sure, you can program it all in assembly and you'll have an entire OS + all applications running on 4MB RAM, but there's a reason this doesn't happen.
Indeed there is a reason. It can be described in just two words:
1) lazy
2) programmers

And yes, I am speaking from experience. I am a programmer myself and the way I see my fellow programmers waste resources pains me to no end. The algorithm runs too slowly or runs out of memory? Why bother investigating the cause, just give it a higher priority or allocate a bigger buffer!

Unfortunately I see this kind of attitude everywhere, its application in programming is that, one application out of many. Other examples: driving in the middle lane of an empty three-lane motorway... a group of three taking a table for 6 when smaller tables are available... buying a bigger pack of yoghurt in the full knowledge that you will never finish it and end up throwing it away - but it's only 10% dearer than the small pack, so what?... keeping the fridge door open while pouring the milk in your tea, because who cares, right?

It is this attitude that means that 1GB is not enough for a task that just a few years ago ran happily on 4MB, nothing else.
 

The Following 12 Users Say Thank You to pichlo For This Useful Post:
pichlo's Avatar
Posts: 6,447 | Thanked: 20,981 times | Joined on Sep 2012 @ UK
#89
Originally Posted by MartinK View Post
Well, if an app crashes a crash handler (such as ABRT used in Fedora/RHEL/CentOS) should show up and ask the user if he wants to report the crash (user writes what caused the crash and confirms what logs to send), so that the application developers know about the crash and can do something about it.
"If an application crashes" and "if the OS decides to kill a well behaving application just because" are two different things. Just thought I'd point that out
 

The Following 7 Users Say Thank You to pichlo For This Useful Post:
Posts: 586 | Thanked: 471 times | Joined on Jan 2010 @ Finland
#90
I like SailfishOS 2.0 MUCH! Looks and feels so smooth and way easier. Now you can open apps very fast, because you can always swipe from bottom and it stays on top. Now it feels multitasking phone... Browser lags littlebit still there 3tab max, but now it is fast again!!!

One thing i can't understand.. Why landscape is not ready? It works, but at home screen it will always turn portrail, with nice turn efect

Way that pulldown menu looks... It's not that sweet, but i actualy like that 2time flicks, when you select something

But thats me. Me like!
 

The Following User Says Thank You to Makeclick For This Useful Post:
Reply


 
Forum Jump


All times are GMT. The time now is 01:42.