- Here's a multi-tasking scenario that I use daily. (I'm not even going to ask if anyone can replicate this on other platforms)
Here I'm browsing Reddit, and I see there is a George Carlin video. It's standup comed, I can play it in a small screen and still enjoy it. At the same time, I can continue browsing Reddit, or read comments other people have made about the video.
Now I have switched over to Chrome, and I am still enjoying the video. I need to do a quick calculation so I pull up AirCalc. Without this app, I would have to switch back and forth between browser and calc, and rely on my short term memory.
I've seen a trend of increasing popularity with this pop-up capability. WIth the larger screens nowadays it is not a novelty but is actually useful. I am envious of the Galaxy Note and the Windows 8 snap-on side by side window. Hopefully the snap-on capability gets brought in into the stock OS.esk369 likes this.10-15-12 08:00 PMLike 1 - But the playbook can I just recorded a six minute video and had the browser open with YouTube then I opened my email and sent one to myself while I also had a song playing the I then opened the video store and browsed a cpl of movies the whole time the video camera kept recording it also picked up the song I had playing in the background.
So it would seem that the pb is so far the only device that can do that.
For me the big picture is the pb imo is a multitasking beast my t300 aint no slacker on the multitasking as well and I'm still learning android os and NAFB Samsung rocks it also impressive for a handset I think they got it right but I'm really looking forward to bbx as I'm really impressed with some of the demos I have seen.
Btw I will not be posting the video of me and the gf in the bedroom as we are in our pj'sesk369 likes this.10-15-12 08:05 PMLike 1 -
Now let's move onto Android example. Devs have no access to hardware for audio. Audio has no hardware acceleration. Resulting in latency in the music therefore resulting in poor user experience.
PB can handle pure 64 bit audio in real time with zero latency. That is an audiophiles dream.
This is just an audio example. BB10 media capabilities will blow people's minds IMHO....that is only if you as an end user appreciate a quality uncompromised experience.notfanboy likes this.10-15-12 08:26 PMLike 1 - Here's another real-world scenario that I use daily on my drive to and from work. It also happens to be a great example of multi-tasking because there are so many concurrent threads of execution.
- Open Glympse - send SMS to friend with link for real-time tracking. Following the link from his phone or desktop browser, it will show my location updated in 30 second intervals for the duration I specified.
- Launch Google Nav. Tap voice action. Say "Navigate to 123 Main Street"
- Google Maps Turn by turn Navigation starts up - this is the will remain in the foreground
- ... I start driving ... everything is hands free after this point ...
- Tap call button on BT stereo headset. This is configured to open up the audiobook player. This is what I listen to during the drive.
- Pause or rewind the audiobook using AVRCP controls on the BT headset
- If a call comes in, answer the call on the headset. Audiobook pauses during the call then resumes afterward.
- Tasker is running in the background. I watches as I enter or exit a geofenced area. An SMS is automatically sent "I've just left the office"
- Tasker can also auto-respond to text messages while I'm driving: "I'm driving at the moment and will get back to you soon"
I am curious to see how much of the above scenario can be replicated on other mobile OSes.Last edited by notafanboy; 10-16-12 at 05:47 AM. Reason: included more detail
10-15-12 08:29 PMLike 0 -
Can you give another example of something that could blow people's minds, because this one isn't that mind-blowing to be perfectly honest.10-15-12 08:35 PMLike 2 -
iOS does not allow custom services. Android does. Their method of hibernation is similar if that's what you meant.
Only "first-party" iOS apps can multitask? Wrong. Examples:
<Apple Services Examples>
Developers can and do choose what they want their apps to do in the background. The difference is that this must be an explicit choice. When it doesn't make sense for the app to run in the background ( eg first person shooters ), they don't. The app is held static in memory, thus CPU cycles etc are not consumed unless necessary.
On Android, the developers can do literally anything in the background. Although, it is less efficient than iOS method as the Android services basically have to duplicate existing frameworks. Also, the developer needs to actually create a service to handle this backgrounding situation which can prove too difficult for many. I still get annoyed that no remote desktop app will continue in the background on my Android tablet.
With your example of a first-person shooter, Playbook will also pause the game (otherwise your character would die). When the game is paused, even though the game is running I can guarantee you it is not using your precious CPU cycles :P That is, unless it is a multitplayer game and you actually need information to be updated while you are paused/in the loading bay. In that case, you would definitely want the game to be paused AND running.
In my opinion, the best multitasking solution should be built on top of a consistent framework or toolkit. Example: Symbian with Qt. You can run more than 50 Qt applications on a Symbian phone and still have your battery last all day. Even if it has only 256MB RAM. The reason is that Qt only loads once (shared libraries). The apps that run on top of this are only a few MB (images mostly). As Qt is already in RAM, the app startup time is incredibly quick too.
With BB10, Qt (and Cascades) is coming pre-packed as well. All the apps will be using it (except for games), so you will see the ability to run many apps with uncompromised multitasking.
With Playbook, it works the same with Flash apps (most apps are based on Adobe Air). You should be able to run a large number of Flash apps without affecting RAM usage or CPU significantly.Last edited by xsacha; 10-15-12 at 08:58 PM.
10-15-12 08:39 PMLike 3 - Thanks for that. I wonder if you can upload a couple of sample audio files using the 64-bit audio, and without so I can compare the two. Admittedly I'm not an audiophile and I find it hard to discern subtle differences in consumer grade audio. Heck I still enjoy FM radio
Can you give another example of something that could blow people's minds, because this one isn't that mind-blowing to be perfectly honest.
Not being rude....I'm just drawing the line........I'm also a funny
If you ever have the opportunity to listen to a high end system.........or even an above entry level set of headphones your ears will notice.
We'll see hardware and software interaction that will push boundaries.Last edited by missing_K-W; 10-15-12 at 09:03 PM.
10-15-12 08:51 PMLike 0 -
So tell me, what exactly is it that you can't do "as a developer" on iOS? It might require a creative solution, but unless you want to write an app to drain the battery while in the background, I don't see too many barriers. I'd be interested in knowing your problem and seeing if I could find a workaround for it.10-15-12 09:31 PMLike 0 - @mike007 apps that sleep and wake have a transition time. Including transitions. App switching, sleep/wake, services are great if you like the in and out paradigm. The whole concept of tablet OS and BB10 is to have the ENTIRE UI feel as though it is ONE PROGRAM. What you consider an app operating in the back ground wasting "resources" is part of ONE program( whatever you like to call it) the invocation framework etc all ties into the one.
People need to wake up to see what's really going on here. On Android, iOS and Windows you have apps segregated and an in and out paradigm. You have to look at BB10(tidbits of this on PB with its multitasking) and see the whole platform as one.
Peek, flow invocation etc is treating the entire platform as one big app seamlessly interacting with it self. Devs even need to start waking up to this.
The multitasking brilliance of BB10 is that it is ONE app. Start seeing the forest through the trees folks.10-15-12 09:55 PMLike 0 - @mike007 apps that sleep and wake have a transition time. Including transitions. App switching, sleep/wake, services are great if you like the in and out paradigm. The whole concept of tablet OS and BB10 is to have the ENTIRE UI feel as though it is ONE PROGRAM. What you consider an app operating in the back ground wasting "resources" is part of ONE program( whatever you like to call it) the invocation framework etc all ties into the one.
People need to wake up to see what's really going on here. On Android, iOS and Windows you have apps segregated and an in and out paradigm. You have to look at BB10(tidbits of this on PB with its multitasking) and see the whole platform as one.
Peek, flow invocation etc is treating the entire platform as one big app seamlessly interacting with it self. Devs even need to start waking up to this.
The multitasking brilliance of BB10 is that it is ONE app. Start seeing the forest through the trees folks.
Honestly though, you're arguing semantics here. We both bake a cake, they both taste great. We both use different methods but come up with a pleasing result. What it really comes down to is: will it matter to the consumer? Swiping between apps on my iPad has zero lag time. The "transition" is virtually the same as the one on the Playbook (new app slides in, old app slides out).
(For anyone who doesn't know, In and out paradigm is a term coined by RIM to describe "traditional" multitasking. It's not actually a "real" paradigm referenced by anyone else in the industry)Last edited by mikeo007; 10-15-12 at 11:19 PM.
10-15-12 10:09 PMLike 0 - One thing everyone seems to have missed here is the hardware component to the equation. "notafanboy" is conducting his tests using an SGSIII, which has 2 either more (quad) cores running at 400 additional MHz or a similar pair (dual) of cores at 500 additional MHz (depending on the phone model/market) when compared to the Playbook. I know from experience that simply overclocking my Tegra 2-based Acer Iconia A200 by 500MHz results in a roughly 40% improvement in integer performance (per AnTutTu) and a 31% improvement in JavaScript performance (per SunSpider). All of which means that, no matter how you slice this comparison, the hardware parameters favor his smartphone.
So, given the above, a proper test would require that you either a) use a Playbook 4G as the QNX test bed or b) use a different Android device with specs comparable to the original PB (e.g. the Acer Iconia Tab A200 running with Stock ROM and CPU frequencies). I don't have the former, but much of my experience with Android is tied to the latter, so the issue then becomes creating a proper test environment, including the identical network infrastructure (router, LAN/WAN speeds, physical location) and also using the identical applications (where possible). And finally, you need to have a very specific script (my original 30 step example is a good start) that includes identical tasks, executed in an identical fashion/sequence (where possible, accounting for UI differences), and over the identical time period.
The above is how we did it at Intel Corporation back when I spearheaded their competitive marketing research vs. AMD, Cyrix, et al. Unfortunately, it requires a considerable amount of planning and setup, and is ideally a two or three person job (i.e. fumbling with the camera while balancing a beer can isn't the most professional way to do this).
Bottom Line: Unless someone wants to step up and do a proper analysis of the multitasking behavior of each OS when running on comparable hardware and in a controlled environment, all of these arguments are moot.
Frankly, if I had access to my old lab setup - and a day or two of my life to waste - I'd grab the necessary devices and run the test. However, I've already done my evaluation of Android vs. QNX for my own purposes, so I know how it ends. And, of course, this time around nobody's paying me to apply my 25 years of experience, including designing and implementing performance metrics for some of the largest companies in the world, to complete their project.
My advice to everyone following this thread: Do your own testing and draw your own conclusions. Get multiple devices (hint: most North American retailers have generous return policies) and spend some quality time trying to implement your own daily workflow on each. Because, barring a comprehensive, independently-funded study along the lines of what I described above, testing it for yourself is the only real way to separate truth from BS.
RCK10-16-12 12:36 AMLike 0 -
'proper routine' with a 'timeout' (so it can't run forever). Multiplayer game? Remote desktop? Anything with a network connection where push messaging isn't a viable solution.
'can't .. update the screen' (well you should be able to regardless of whether it is on the screen). Multiplayer game? Remote peripheral?
And the most important part is that it has to fit in to the categories that Apple allows. This is the most important part that you forgot. It rules out the custom code.
Unfortunately there is no workaround, unless you jailbreak. The limitations are well-known and you seem to have covered them yourself (perhaps unknowingly).
I really got spoiled from using Symbian which is essentially the same as any desktop OS for development. You don't need to handle anything or do any workarounds. Just stop processing when you can and the battery, RAM and CPU will still be fine. If they aren't fine, the OS handles that for you (like any other OS). Playbook is the same as this.adi_vastava likes this.10-16-12 04:50 AMLike 1 - One thing everyone seems to have missed here is the hardware component to the equation. "notafanboy" is conducting his tests using an SGSIII, which has 2 either more (quad) cores running at 400 additional MHz or a similar pair (dual) of cores at 500 additional MHz (depending on the phone model/market) when compared to the Playbook. I know from experience that simply overclocking my Tegra 2-based Acer Iconia A200 by 500MHz results in a roughly 40% improvement in integer performance (per AnTutTu) and a 31% improvement in JavaScript performance (per SunSpider). All of which means that, no matter how you slice this comparison, the hardware parameters favor his smartphone.
So, given the above, a proper test would require that you either a) use a Playbook 4G as the QNX test bed or b) use a different Android device with specs comparable to the original PB (e.g. the Acer Iconia Tab A200 running with Stock ROM and CPU frequencies). I don't have the former, but much of my experience with Android is tied to the latter, so the issue then becomes creating a proper test environment, including the identical network infrastructure (router, LAN/WAN speeds, physical location) and also using the identical applications (where possible). And finally, you need to have a very specific script (my original 30 step example is a good start) that includes identical tasks, executed in an identical fashion/sequence (where possible, accounting for UI differences), and over the identical time period.
The above is how we did it at Intel Corporation back when I spearheaded their competitive marketing research vs. AMD, Cyrix, et al. Unfortunately, it requires a considerable amount of planning and setup, and is ideally a two or three person job (i.e. fumbling with the camera while balancing a beer can isn't the most professional way to do this).
Bottom Line: Unless someone wants to step up and do a proper analysis of the multitasking behavior of each OS when running on comparable hardware and in a controlled environment, all of these arguments are moot.
Frankly, if I had access to my old lab setup - and a day or two of my life to waste - I'd grab the necessary devices and run the test. However, I've already done my evaluation of Android vs. QNX for my own purposes, so I know how it ends. And, of course, this time around nobody's paying me to apply my 25 years of experience, including designing and implementing performance metrics for some of the largest companies in the world, to complete their project.
My advice to everyone following this thread: Do your own testing and draw your own conclusions. Get multiple devices (hint: most North American retailers have generous return policies) and spend some quality time trying to implement your own daily workflow on each. Because, barring a comprehensive, independently-funded study along the lines of what I described above, testing it for yourself is the only real way to separate truth from BS.
RCK has the some of the lengthiest posts in this thread, which took far longer than 13 minutes to write. So much time talking, so little time doing. Don't Blackberry people do anymore?
Here's what I'm reading between the lines: RCK is preemptively making excuses by blaming the hardware. BTW the GSIII model I used is dual core. He says he don't want to do the test because it's not a properly set-up lab experiment. And yet, he was happy enough to make definitive statements based on anecdotal experiences with the Iconia Tab. So sometimes informal tests are ok and other times they are useless, right? Talk about moving the goalposts.
Bottom line: If the PB could perform RCK's own scenario even with less powerful (?) CPUs, then he doesn't need to make all of those excuses. And for all I know, maybe it can. I don't care so much if I "win" the argument or not, for I am not a fan boy of any device. I'm just here to learn about these fascinating gadgets. If I'm proven wrong, so be it. I would be better off because I learned something new. Despite a lack of scientific rigour, I believe we will all learn something from a comparative video. It will be more informative than all of this empty gabbing, theorizing, and speculation.
I fully concur with RCK's advice. Everyone should Perfram your own tests and make your own conclusions, if you are able to. And to that I'll add my own advice to RCK. Post that video - it's another data point so that others and myself can make up our own minds. Unless you think that that video will be unfavorable to your arguments.10-16-12 05:14 AMLike 0 - Sigh. Or instead of writing all of these excuses, RCK could have spent 13 minutes recording a video like I did. (Well maybe a bit longer because I'g guessing RCK can't replicate his own workflow on a PB as efficiently as I did.
RCK has the some of the lengthiest posts in this thread, which took far longer than 13 minutes to write. So much time talking, so little time doing. Don't Blackberry people do anymore?
http://forums.crackberry.com/blackbe...oxnews-749886/
I knew you wouldn't be able to hold back forever, and I was right: You finally showed your true colors.
Now back to the discussion at hand: As I've stated multiple times, a short video showing task switching in quick succession isn't going to cut it - plus, several others have already posted links to similar exercises involving the PB, so why duplicate the effort when it's a major PITA for me to do so?
The fact remains that, in order to get an accurate assessment, you need to use the devices over a period of time and then compare responsiveness, stability, etc. Which is why I'm encouraging everyone to test for themselves with their own workloads and see if they don't run into the same issues I did when trying to migrate full-time to the Android platform. Perhaps Android will prove adequate for their needs. It didn't for mine or for many others in this thread.
And for the record, I type VERY fast (occupational hazard), so...
RCK10-16-12 08:00 AMLike 0 - 'virtually' any code.
'proper routine' with a 'timeout' (so it can't run forever). Multiplayer game? Remote desktop? Anything with a network connection where push messaging isn't a viable solution.
'can't .. update the screen' (well you should be able to regardless of whether it is on the screen). Multiplayer game? Remote peripheral?
And the most important part is that it has to fit in to the categories that Apple allows. This is the most important part that you forgot. It rules out the custom code.
Unfortunately there is no workaround, unless you jailbreak. The limitations are well-known and you seem to have covered them yourself (perhaps unknowingly).
I really got spoiled from using Symbian which is essentially the same as any desktop OS for development. You don't need to handle anything or do any workarounds. Just stop processing when you can and the battery, RAM and CPU will still be fine. If they aren't fine, the OS handles that for you (like any other OS). Playbook is the same as this.
To tackle a few of your examples:
Multiplayer game: depends what type of game you're talking about, but generally if there's a server involved, it's a non-issue. Just re-sync the client when the app gets control of its sockets back.
Remote Desktop: perfectly doable. Just see pocketcloud for an example.
Anything with a network connection: the only thing I see that's not doable would be a client/client connection where both clients could potentially be state saved, resulting in a connection timeout. Even in this case, there are workaround to keep the connection alive (silent audio track, empty VoIP packets)
Can't update the screen: there's no reason to update the screen if you can't see it. Push notifications can pass any necessary messages and if there's something that needs to be rendered, you can buffer it for when the app regains the screen.
Again, I'm not arguing here that iOS provides desktop or Playbook level of multitasking at the core, but I am arguing that with some creative coding, the user experiences can be the same or better.notfanboy likes this.10-16-12 08:09 AMLike 1 - First, you've already proven yourself to be a troll in this unrelated thread:
http://forums.crackberry.com/blackbe...oxnews-749886/
I knew you wouldn't be able to hold back forever, and I was right: You finally showed your true colors.
Question for the mods: are posters allowed to label other posters as trolls without consequence?
Now back to the discussion at hand: As I've stated multiple times, a short video showing task switching in quick succession isn't going to cut it - plus, several others have already posted links to similar exercises involving the PB, so why duplicate the effort when it's a major PITA for me to do so?
1. It is YOUR scenario. Remember you said "my typical workload might look like this" and you even boasted that "All of the above took place before 10:00AM. "
2. It took me only 13 minutes to replicate your scenario
3. If this was MY day to day workload, I could do it much faster.
Just set up a camera and video do what you do every day for 13 minutes. This is a major PITA? Sorry RCK, it's clear to me prefer talk over testable claims. I'm doing my best to incorporate evidence into the discussion, it's a shame that other people just prefer to talk and wave their hands around.Last edited by notafanboy; 10-16-12 at 10:19 AM.
10-16-12 09:42 AMLike 0 -
Does PocketCloud solve this issue?10-16-12 10:09 AMLike 0 - @mikeo007, @xsacha
I have no experience in mobile app development, but am curious to understand one thing.
So lets say when my app goes to background, I want to do some calculation like
n=0;
while( n<10000 )
{
//do some calculation which may take time
}
So what happens in this scenario on IOS and Android?
On Windows (PC), when the App goes in background, the process will still be scheduled, and can do the said calculation, and may be ready with the result(assuming it was in background for sufficient time ), when it comes back to screen.
AFAIK this is not the case with iOS; or does it?10-16-12 10:10 AMLike 0 - Been trying to stay neutral with this pp match but somtimes you have to take sides I read NFB's joke on the other thread in question " media error only on foxnews" How does this make him a troll?????????? I just can't understand the logic behind it he cast no net didn't throw out any bait if anyone imo was trolling it was the op of this thread again just my opinion but you keep throwing that T word around its very unbecoming IMO I just don't get it do you guys know each other personally?
Enough is enough.10-16-12 10:51 AMLike 0 - Not at all. I have never attacked RCK or anyone else for that matter. All I want is an evidence based discussion where we can all learn things. I have posted a lot in this thread, but some of it is just to defend myself from accusations of trolling or lying. In my responses I kept it lively but civil and I have tried not to respond in kind.esk369 likes this.10-16-12 11:04 AMLike 1
- Not at all. I have never attacked RCK or anyone else for that matter. All I want is an evidence based discussion where we can all learn things. I have posted a lot in this thread, but some of it is just to defend myself from accusations of trolling or lying. In my responses I kept it lively but civil and I have tried not to respond in kind.Toodeurep likes this.10-16-12 11:15 AMLike 1
- transitive verb from Webster's dictionary
1 : to cause to move round and round : roll
2 a : to sing the parts of (as a round or catch) in succession
b : to sing loudly
c : to celebrate in song
3 a : to fish for by trolling
b : to fish by trolling in <troll lakes>
c : to pull through the water in trolling <troll a lure>
d : to search in or at <trolls flea markets for bargains>; also : prowl <troll nightclubs>
intransitive verb
1 : to move around : ramble
2 a : to fish by trailing a lure or baited hook from a moving boat
b : search, look <trolling for sponsors>; also : prowl
3 : to sing or play in a jovial manner
4 : to speak rapidly
troll�er noun
Examples
They were trolling the ocean floor.
She loves to troll flea markets looking for bargains.
Origin: Middle English, probably from Anglo-French *troiller, *troller; akin to Anglo-French troil, trolle winch.
First use: 15th century
Synonyms: comb, dig (through), dredge, hunt (through), rake, ransack, rifle, rummage, scour, sort (through), search
2
troll
noun
: a lure or a line with its lure and hook used in trolling
First use: 1869
Synonyms: brownie, dwarf, elf, faerie (also faery), fay, gnome, goblin, gremlin, hobgoblin, kobold, leprechaun, pixie (also pixy), puck, sprite, fairy
3
troll
noun
: a dwarf or giant in Scandinavian folklore inhabiting caves or hills
Origin: Norwegian troll & Dan trold, from Old Norse troll giant, demon; probably akin to Middle High German trolle lout.
First use: 161610-16-12 11:19 AMLike 0 - transitive verb from Webster's dictionary
1 : to act in opposition to : counteract
2 : to incur or provoke the hostility of <his criticism antagonized his friends>
Other forms: an�tag�o�nized; an�tag�o�niz�ing
Examples
He didn't mean to antagonize you.
Her comments antagonized many people.
Origin: Greek antagōnizesthai, from anti- + agōnizesthai to struggle, from agōn contest — more at agony.
First use: circa 1742
Synonyms: embitter, empoison, envenom10-16-12 11:22 AMLike 0
- Forum
- BlackBerry PlayBook Forums
- BlackBerry PlayBook
How does the PlayBook multitask?
Similar Threads
-
How does the Warranty for playbook's bought on Ebay work?
By daytrade in forum BlackBerry PlayBookReplies: 16Last Post: 01-04-13, 11:06 AM -
How often does the playbook receive software updates?
By randomkidsrule in forum BlackBerry PlayBook OSReplies: 14Last Post: 09-03-12, 11:05 PM -
How well does the playbook work with the 9650?
By drudometkin in forum BlackBerry Bold SeriesReplies: 7Last Post: 06-05-11, 02:00 PM -
How does the Playbook screen look compared to iPad 2?
By plucky duck in forum BlackBerry PlayBookReplies: 15Last Post: 04-08-11, 10:37 AM -
How does the Playbook store photos?
By walbern1 in forum BlackBerry PlayBookReplies: 12Last Post: 02-04-11, 10:46 PM
LINK TO POST COPIED TO CLIPBOARD