samsung galaxy s3 broken lock button

>> rhode: android developer relations officehours. i'm sparky from munich, germany and i'm joined by my colleagues, matt and nickand rich in london, uk, matt with google tv. >> butcher: in [indistinct] burrito mike'sthis week. so, we have... >> hyndman: burrito [indistinct] he's okay.>> butcher: we have richard, the editor of geekyouup.com>> hyndman: yup. my account is for... >> butcher: and the android account is matt.>> gaunt: hi. >> hyndman: my clothes today are brought toyou by the dutch android user group. we always do someone [indistinct] and the [indistinct]gtalk [indistinct] on the back. nick brought to us by christmas i think.>> butcher: yes, it's my christmas jumper.

>> hyndman: christmas jumper. and matt by…>> gaunt: nothing goes wrong with a man's style with google s [indistinct]>> rhode: i was going to say matt--matt and i are in--are in uniform today, so.>> gaunt: that's very good. >> rhode: all right.>> butcher: someone had to be, matthew. sorry sparky we high jacked your beautiful intro.carry on. >> rhode: no, no, no. that was--that was farmore interesting than just having me say the usual thing. but anyway, welcome everybody.this is android developer relations office hours. we're here to talk to you about anythingrelated to android application development. usual disclaimers apply which rich will behappy to deliver. basically, we generally

restrict ourselves to sdk related issues.we don't talk about deep internals. we don't talk about unreleased products or our futureplans. but apart from that, we're happy to--we're happy to basically, you know, see what you'reinterested in, talk about what you're interested in. i see we're already joined in the hangoutby several people. we got some folks who have contributed questions in the moderator. hi,andrew. >> kelly: hey.>> butcher: and andrew has a camera because his girlfriend or wife isn't watching x-factortoday. >> kelly: i can't see myself though.>> butcher: oh. >> hyndman: interesting.>> kelly: all right. yeah.

>> hyndman: i've always hope they put it inyoutube live stream at the same time works quite well, if you want to be 10 seconds behindthe rest of us. who's joined us on the far end side.>> butcher: it's yossi. >> hyndman: oh, yossi. there we go. hi.>> rhode: yossi? we have yossi here. >> hyndman: yossi.>> rhode: and so he got the memo about the facial hair.>> elkrief: yup, excellent. i'm about to start a team.>> butcher: join the club. >> hyndman: so we have allan, andy or andrewand yossi all with us in the hangout today. >> rhode: okay.>> hyndman: maybe [indistinct] good quality

camera for elkrief.>> rhode: yeah. so, what's on your mind everybody? who has something that you'd like to sharewith us or want to--have a question you want to ask?>> hyndman: well, i've got something i'd like to share.>> gaunt: go rich, go. >> hyndman: it's just that we did for theuk, specifically may have launched the 2, 3, and 1s that went passed recently. but thenexus 7, tablet of the year. in fact--was it gadget of the year?>> gaunt: gadget. >> butcher: only one vote cast because itbeat out the ipad 3 for tablet of year and it won the overall, the big cache grade, thegadget of the year.

>> hyndman: gadget of the year is the nexus7. so we've been having a few little parties around with google office here to celebratethat. thank you very much to t3. >> butcher: doesn't an old tablet look good?>> gaunt: good indeed. >> rhode: i have to say, i mean it is a marvelousdevice. the nexus 7 is my--basically my go to android device. i mean i'll use my phoneif i'm out and about because i don't generally bring the tablet, you know. i mean i'll evenbring my nexus 7 with me but it's in my bag, the phone is in my pocket. but, you know,anywhere that i'm within range of wifi, the nexus 7 is the device i want to use.>> gaunt: uh-hmm. >> hyndman: okay.>> butchers: agreed. so to developers out

there, if you haven't started looking at yourapps and, you know, if it's running well, if it's optimized for the screen size andso on then, now is a good time. >> rhode: yeah. actually that reminds--thatreminds me. isn't this week tablet week? shouldn't we be talking a whole lot about tablets today?>> butcher: yeah, we got a bit early last week?>> hyndman: yeah, we probably recorded online. >> gaunt: oh, yeah.>> rhode: i missed it. i was... >> elkrief: i didn't call for a week and thenrealized that it was tablet week. >> hyndman: is that leslie? is that you?>> drewery: hey, how are you doing? >> hyndman: he's muting. and we'll say hellothrough sign language.

>> rhode: now, last week while you guys weretalking tablets, i was out there--i was out there celebrating deutschen einheit. it'sa national holiday here in germany last week. >> hyndman: yeah. we're jumping again on tablets.have you joined us today because you’ve got some questions, leslie, or you just comealong to hangout? >> drewery: just to hangout and see what'shappening. >> hyndman: all right, i see.>> rhode: welcome to the hangout. i see that also richard thompson has just jumped on.hello, richard. do you have anything you want to talk about? maybe we'll give you a momentto get yourself establish. >> butcher: yeah, man. give him a chance.>> hyndman: okay. so, yeah. this is a point

that anyone in the hangout can jump in andask a question before we hit the moderator. if anyone has anything they'd like to discuss.>> gaunt: yossi is showing off some rather a fancy [indistinct]>> rhode: oh, is that a serious tree? >> elkrief: i wouldn't like not to interruptyou but i just wanted to show you our office and also these.>> hyndman: nice. you've redecorated your entire office to be android.>> gaunt: excellent, you know. >> butcher: hey, cool. nice sticks.>> hyndman: good work, yossi. >> butcher: explicit. i like it. i like ita lot. >> gaunt: interesting.>> hyndman: i'm sure that it enhances your

ability to do android development. nice talking.>> butcher: and just to catch up when we were talking about tablets as well, i think thereason we declared this tablet week is arbitrary but it's also mostly to celebrate the newquality guidelines we just released. so it's a series of kind of checks you can do to makesure your application is going to look stellar on a--on larger devices. so, hopefully youguys have all seen the blog release on android-developers.blogspot.com and seen the checklist and gone through thatand have a look at your apps and kind of seen do they make use of white space, are theygoing to scale, are they going to be available on all devices which might have differenthardware configurations. so it's kind of what catalyze this week's kind of theme. and richardin the sidebar asks, "will there be google

tv quality guidelines?">> gaunt: i mean that’s--so we've got a lot of…>> butcher: i'm going to pass to matt. >> gaunt: …documentation on google tv designguidelines over at developers.google.com/tv. so, yeah, we've got a whole load of stuffthere about making it very cinematic and accounting for [indistinct] you're not--you're not reallygoing sort of--the interaction is more safer. it's less of a one to one kind of thing wherepeople are usually using a personal device. so, yeah, go check that out. lots of infothere. >> butcher: so that's the checklist more ofkind of document or so? >> gaunt: yeah. it's much more kind of alongthe lines of like the new design stuff that

android went through whereas it's more of--yeah,going through the different sort of design patterns you could follow and other aspects,so. >> butcher: has anyone on the--in the hangouthad a chance to have a look at the tablet quality guidelines yet? any one?>> elkrief: not yet. >> sutton: yeah, i've had a quick look throughthem. it looks cool. it's… >> butcher: do you suppose it's the kind ofthing you can--you can see yourself running through?>> sutton: yeah, definitely. the checklist is definitely something i'm going to run throughin every app because it's one of those things, you make assumptions about stuff. you getcaught up in the design and then you sit there

and i think, "oh, i've forgotten somethingfundamental." so it's really useful to--so i see that checklist is really useful to have.>> butcher: yeah. especially when you're developing like you are into like develop on device perhaps.you might get into a single device mindset where you, you know, going through debug cyclestrying to like get this piece of functionality working and shift it. and you sometimes don'tthink about the brackets or stuff out there. so i think having this simple checklist isreally helpful. so have you guys seen it too? and if you have any feedbacks or suggestionsthen--to any android developers or one of us and we can--we can suggest it. everyoneis getting their [indistinct] heads on. look at this.>> rhode: yeah. that’s just the whole thing.

>> [indistinct]>> butcher: cool. so, are there any live questions in the hangout or any other news that peoplewant to share before we kind of jump into the moderator?>> gaunt: google tv is getting play. it's getting native play.>> butcher: yes. welcome to the club. >> gaunt: which--yeah, i know. it's been along--long awaiting. so, yeah, google tv is finally getting native play. so you can finallyrent and buy movies and play it straight to google tv which is fantastic. we might grabon that. >> butcher: so, any changes for developers?anything, like any different internal docs or whatever they can use now that they couldn'tbefore or anything like that?

>> gaunt: no, i'm unaware. we get subscriptionbilling with the native play. so when that’s come in, that’s one which i think a lotof people been wanting for a little while. >> butcher: cool.>> gaunt: so, yeah. that's basically what's going on.>> hyndman: so you have it up and running right now and that’s going to take…>> gaunt: yeah. for the moment we just got enough billing, no subscriptions. so that’sgoing to come out. >> butcher: that's inevitable. quite likelast night, in fact. free trials, no subscription. >> hyndman: uh-hmm.>> butcher: everyone caught that? if you have a subscription-based product with the [indistinct]without the free period it's kind of getting

used as hooks. oh, that seems--check downthe sidebar. >> rhode: yeah. so leslie asked a questionin the sidebar. he says, "how many companies are using the accessibility framework andhow does it sort of stiffly refute all the good ones?">> hyndman: well, you know, we certainly have recommended many times in the past that peopleshould at least put the content description on all of their html text. so it's just goingto basically work but in terms of using the entire accessibility framework, i don't havea metric on that. >> butcher: no. i have no stats. i do knowthat wolfram basically post a good article on this lately. so if anyone reads grokkingandroid block--blog, it's great, a typical

blog. he recently did a very good post determininghow you would add accessibility towards your apps, so. yeah. [indistinct]>> drewery: well, the actual question is who uses it for anything other than the simplecontent description? i mean we use it quite extensively but the solution to how powerfulit really is to why people are using it? i think richard here is in joint club this month.>> hyndman: excuse me. i've got a bit of a cough going on.>> butcher: which one, joint club uk? >> hyndman: joint club uk?>> drewery: uk, yeah. yeah, let me get there. i'll give you an insight to what we use inthe framework form, then you'll understand the question in context to that.>> hyndman: it's a question of which context

will come three weeks later?>> drewery: yeah. well, remind me when you're there.>> butcher: and i'll see if i can dig out anything relevant for you of that quite aswell, then i'll see you at joint club. >> hyndman: joint club uk? i guess i willbe there as well. >> drewery: yup.>> hyndman: anyone else? >> banes: hello. i'm going to be there.>> hyndman: hi chris. i see you pop in there and matt as well.>> gaunt: yup. >> hyndman: they'll be quite a few of us who'lltrain from uk. okay. so we jump into the moderator then. well, sparky, people hangout for thenext question.

>> rhode: yeah.>> butcher: he's going to do it. >> rhode: first question is from chris inlondon. is there a reason why google does not release updates to new android versionsfor all interested developers of nexus 7 and galaxy nexus devices right away outside alongthe [indistinct] cycle without having to route to the device, etcetera. well, of course,there's a reason. oh, he says, "it would be--" he goes on to say, "it would be convenientto have an option of developer options where you app in to the latest releases immediatelyrather than waiting for your turn on the ota," you know, sort of phase rule out. and he says,"yes, it's only basically a developer thing. not a lot of people would necessarily be usingit." i actually think that's a great idea

and i like it so much that i opened the ticketin our issue tracking systems saying we should do this. so, i don't know that you would savea lot of time because we don't really waste a lot of time between, you know. and then,you've code freeze an ota or perhaps maybe the other way to turn around or just say sometimeswe have a scheduled [indistinct] deployment and we code right up into the last possibleminute. but... >> butcher: plus it seems to me we do a releaseand we normally release the recovery images, right? the--so you can grab one of those andflash yourself. >> hyndman: yup.>> butcher: or as soon as it hits ota, i'm pretty sure you can find the link to downloadit.

>> hyndman: people get the links and you don'tthree devices to apply the official signs, patches anyway.>> rhode: yeah. >> hyndman: that's how we choose it. is chris,was this you chris banes or is there a different chris in london?>> banes: well, it was me. >> hyndman: fair enough.>> butcher: what was the one chris? >> hyndman: i said chris in london. therehappens to be a chris in here. couldn't be in the same room?>> banes: right. >> hyndman: actually, i'm not--and of course,there's something like sometimes in the otas if certain carriers require approval beforewe apply patches to devices on the network.

>> butcher: yup, so a device that's developedvia straight from plato google.com and you don't have a de--a carrier in between theota service and to your device device. >> hyndman: this reference code is there foremea. >> rhode: right. but it's--so, the gist ofmy--the ticket that i opened was i suggested that basically people who app in to this shouldbe able to jump to the head of the cue in the phase roll out rather than, you know,waiting a few days. that way if you have an app that, you know, maybe breaks on the nearrelease you can fix it asap rather than having to wait a couple of days free of term to getsomething wrong. >> banes: the trick of that is wouldn't everyonejust go, "yeah, i'm a developer."

>> gaunt: yeah.>> butcher: i'm just thinking that maybe it has to be like in the pub site or some...>> gaunt: yeah. >> butcher: ...of that actually. you haveto be... >> hyndman: right.>> rhode: yeah, yeah, maybe enroll a device through the developer console or somethinglike that. >> banes: uh-huh. yeah, that's probably...>> hyndman: [indistinct] through the play console and you just grab the [indistinct]>> gaunt: yeah. >> hyndman: if you're a developer, you'regoing to be happy enough with that i'm sure. yeah. so, all good ideas.>> sutton: one of the--one of the ways i read

that is aren't they looking for a versionbetween releases? >> hyndman: uh-hmm.>> sutton: so, instead of waiting for a 4.10 to 4.11 or 4.1--4.0 to 4.1 without lookingfor something between those two. >> butcher: uh-hmm. so that's what we didfor the first time we had jelly bean, right? we kind of had the developer preview. it wasout there for awhile so you could run it in the emulator and you could get to know thenew api's before we went to shifting runs and with the shifting devices and the finalfull image, right? so i think that seems to be the direction we're going. but i'd liketo see us do it earlier if we can but, you know, i don't make these decisions.>> hyndman: excellent. okay. well, seeing

as no one is shouting, let's switch on tothe next question in the moderator. we've got quite a few to get through today.>> rhode: let's jump to the next question. mardon in brayton, uk says, "i know it's stilla fair few amounts away but java 8 is on the horizon." okay. al gave a very good answerto that. al, do you want to--you want to reprise for us want you said?>> sutton: sure. basically and google is, you guys can't answer forward looking statementsabout product development. but one of the things with lambda expressions is they needa change in the virtual machine which came in in java 7 and the dalvik virtual machinewhich android uses isn't equivalent to a java virtual machine. so, you can't exactly saythat version of java would be the same as

the bytecode available in the version of thedalvik virtual machine. so, there would probably be some work to implement the functionalityneeded in the dalvik virtual machine and i see that that doesn't seem to have startedto being happened. so, it might happen but its not going to be for awhile yet would bepretty much is the gist of the answer. >> butcher: uh-hmm, interesting.>> rhode: right. and so the short version of al's answer is we're still in--we're stillusing java 1.6 as our target environment and targeted api level.>> sutton: yeah, pretty much. >> butcher: what else is in java 8? i mean,i miss my lambda so much. i used to use lambdas a lot. what else is in 8? there's not a lotin 7, right? there's the diamond operator

and that's about it.>> sutton: there's a lot of stuff underneath the hood in 7. there's--one of the--its dynamicinvokei think is the bi-code that got introduced in java 7 which allows support for none javalanguages to be compiled. that's... >> butcher: yeah, like gvv and so on.>> sutton: exactly. and it's that which is needed to get lambda expressions sorted aswell. so i think a lot of the 7 was underneath the hoods but...>> banes: it's also a--its project coin so that was the whole rebound that fits in thelanguage. just the thought you can use switch, a lot of strings in switch statements andstuff like that. i think that was pushed back from 7 to 8.>> butcher: that's right.

>> banes: but i can't remember the 8. so,yeah. >> butcher: i was running that because a lotof this stuff at lambda can--is often like syntatical sugars. it's often stuff that mightnot have low down changes but you might... >> hyndman: uh-hmm.>> butcher: ...give a go over [indistinct] or something. but i know nothing about thatstuff. >> sutton: yeah. i mean, you can achieve prettymuch the same stuff with using a list of classes and interphases. and that's the whole thingthat lambda expressions are being sort of touted as replacing.>> butcher: yeah. >> sutton: so there's no absolute need forthem but it's one of those things. the less

case you type, the happy your developers are.>> butcher: yeah. let's close maintain, let's boil the plate. it's easier to read samples.i think--i think lambda is a great [indistinct] mark but yeah, i'd like to see them.>> hyndman: oh, no comment. okay. >> butcher: well, i didn't know if it wasn't--ifand when they're coming but... >> hyndman: by december. anymore for anymore?anything in the chat? anything in the google+ thread? is anyone through in that?>> gaunt: no. i know we had it and now, i've lost--oh, got it.>> hyndman: okay. who's going to get that first?>> gaunt: okay. >> butcher: you haven't joined like...>> rhode: while you're checking google+, i'll

read the next--the next moderator question.igo in lviv, a great city where my previous employer before at google had an engineeringteam, says, "have you heard of android annotations transfuse squares dagger links included? theseare annotation based compile-time code generation controls. what do we think of these?" i mentioneda few more, roboguice, greendroid, ormlite. i think they're all perfectly fine. the problembasically is that every time you introduce any kind of an obstruction layer on top ofan api, it has the potential to simplify matters as long as you stay within it but when you--whenyou start integrating with the native other features of the system sometimes it may notintegrate quite so smoothly. so, i think they're all good ideas and i think it's worth, maybeworth a little bit of time investigating and

see if it works for your used case. i, youknow, nothing against them. it's just a matter of is the, you know, is the benefit that yougain from it worth the potential inconvenience of fitting it in together with the rest ofthe system. >> butcher: understood. in agreement to sparky,one of the thing i'll add to that is the reason i quite like the annotation based one is thatthere tend to be a compile-time kind of stuff rather than looking at runtime things. likelooking at some of that roboguice for example which offers great functionality but you'regoing to take a hit at runtime using the reflection and so on. so, i think that's one of the advanceswhy i'd probably look into something like dagger or, you know, android annotations.>> hyndman: uh-hmm.

>> butcher: instead of the runtime ones justbecause performance is such a consideration on my book. [indistinct]>> hyndman: so... >> butcher: any other comments?>> banes: what you said--sparky mentioned the ormlite and i mean that's a runtime oneso that's runtime annotations as [indistinct] greendroid whereas that's compile-time butit has more work in it. so, i'm just--i'm not agreeing.>> butcher: yeah. my call on these things is they tend to be awesome when you know howto use them. like once you--if you're the developer of the project or something.>> banes: yeah. >> butcher: and you know it inside out, it'slike, "look, how easy it is. i could just

like annotate this one annotation and justa billion things." well, i find them quite like quite hard to learn, quite hard to pickup unless you have that whole surface area of knowledge about the--of what it is. andi find it can actually be hard to debug sometimes if you don't know when stuff--especially wherethere are tons of junctions. if you don't know when stuff is getting crazy and substantiatingwhere it's coming from, you know, if you have this great team and someone is trying to comeand debug an issue then it can make it more complicated. so, those are my only concernsabout--around these frameworks. like if you have just some, you know, not standard but,you know, an android developer who is coming in and want to have experience and it couldput him off but otherwise i think that you

have tons of benefits.>> hyndman: yeah. >> rhode: i'll also add just for the point,just for the sake of adding it that greendroid comes from munich, germany so go munich.>> hyndman: go munich. we've got another question at google+ and another question in the chat.so, leslie asked in the chat. when will be able to add our own channels? maybe referringto--where's leslie? are you referring to google play store? yes, you can nod this way.>> drewery: well, the problem with google tv is that i very would like it to be ableto control my v+ box and stuff like that. and i know the frequencies so i just wantto add them but without routing the device that complicates the channel list.>> gaunt: okay. so...

>> butcher: over to matt.>> gaunt: so at the moment you can't access the channel list in europe. we're workingto bring that over here because over in the us it's a lot easier to get that data forthe channels that you have access to. when you start going internationally it dependson which country you're in. there's a couple weird little quirks. like people in germanythey can change the order of the channels which they're not set out to update here,et cetera. so basically we're trying to find a nice reliable way to actually give you accessto that information. but some of the stuff of you're talking about just then, it remindof an api that we announced at google i/o called the immediate devices api where youcan then essentially pretend it to be a set

of top box or sort of create a driver foran actual set of top box and then you can start communicating between the two. so that'sgoing to come with a later version of google tv which we don't have a date for yet. buthopefully it won't be too long. so, yeah. so, i've answered your question i hope.>> drewery: yeah, sort of. what [indistinct] take to get that upgrade though?>> gaunt: i believe not much will take if you will, i believe.>> drewery: okay. >> hyndman: fingers crossed.>> gaunt: yes. >> hyndman: [indistinct]>> butcher: so in the--in the corner feed, yan is asking about a layout guide for nexus7 in landscape and template one tablets in

landscape. well, like what would you wantto know? >> hyndman: i was thinking use the same layout.you can use the same for tablet [indistinct] and the nexus 7.>> butcher: the answer is it depends. it depends what you're trying to layout. like this iswhy we always say work through your design because you kind of--there is no magic lawthat says, this is the layout that works on the phone that will work in a 7 or work outon 7. what works on 7 will work on 10. there's nothing like that. it really depends on yourcontent. so, what i always end up doing when i'm trying to like say mock up screens forandroid design in action, you might start with the phone layout and then like see whatit looks like on 10 and it might be too empty,

it might be too dense. it really is like akind of a gut feeling thing. you end up kind of trying on different size layouts untilyou get the right amount and the spacing, the right kind of density of information.so there's no magic law. you kind of have to look at it and decide. and in fact, ifyou look at the nexus 7 at the system master sheet where some of them use the ui whichswitch shift on the phone and some of them use the tablet ui. so, people applicationfor example uses a tablet based on ui and it, you know, use this basically nicely withouttoo confusing and so on. but some of the applications like google play music for example who usesthe frame ui. it's a decision made by each kind of an application depending on the informationthey're trying to show and just how well it

works in that given amount of space. so thereis no answer to this. it's try and have a look to see if it works. if not, you mightneed to hint things, you might need to change something, you might need to add more space,bigger text sizes and like that. so there's no hard and fast rule.>> hyndman: and the nexus 7 falls into the sw600dp bucket in the layouts as well. soyou're going to have your--in a 10-inch tablet you can put in the sw720s, your nexus 7 andsw600s. and if you want the same layouts you can just get rid of the shortest width 720folder and you just pick up size that says the 7-inch. but you can very easily distinguishbetween a seven-inch and ten-inch tablets and decide what goes in each one just by testingit and what's best.

>> sutton: i find it quite interesting aswell that the nexus 7 originally the launcher was locked in portrait mode. and the latestrelease of jelly bean which is out and about now allows you to do the wonderful thing ofhaving the launcher in landscape as well. so, i guess it's--it really is just...>> butcher: [makes noise] >> sutton: yup, that's it.>> hyndman: [indistinct] >> sutton: i can do it on mine too.>> hyndman: ...just a matter of... >> rhode: it's just wanting to get it right.>> sutton: yeah. so, it's just goes to show that there's a lot of experimentation thatyou need to do before you get things that you enjoy.>> gaunt: yes.

>> rhode: one thing that i've noticed recentlyis i've been a little bit surprise that even now, like we we're what, a year and a halfafter honeycomb? there's a lot of apps especially games adjust. like even on a 10-inch tabletthey just scale up the phone ui and i think it's--in particular for games it's more ofa problem even than some of the others. i mean yes, they don't have a problem with,you know, huge amounts of white space because they're using opengl and scaling everythingproportionally. but some of these games used gestures where the amount that you travelon screen matters and it can actually be significantly harder to play the same game on a tablet becauseyou have to move your finger that much faster in the same amount or that much further inthe same amount of time. i’m thinking of

games like flight control or flick golf orsomething like that where you train your, you know, you train your muscles to work fasteron a certain screen size and then, you switch to a tablet and all of a sudden everythingis different. and, you know, there's a real fatigue factor that comes in to dragging thoseairplanes all the way across in your screen. >> sutton: it's interesting…>> rhode: so… >> sutton: …that in the tablet app qualitychecklist sort of three or four the points all go across the same thing you're just saying--beensaying, sparky. by using fonts and touch targets and assets designed for that kind of sizeof screen so hopefully it's something that the checklist will hope reinforce with people.>> butcher: yeah. the good news is it can

be a bit scary because you think about thiscontinuous sort of like range of device sizes out there. so, if you think about, "oh, doi have to allocate my layouts and make a layout that works great on a, you know, a 4-inchphone, a 5-inch phone, a 7-inch tablet and 8.9-inch tablet. the good thing about thisis like there's actually kind of a discrete number of switching points at the moment.we kind of see a lot of the stuff kind of setting around four, seven, and ten. it'skind of like come in places and the variations around that are not so huge that you can usea layout that works on like a seven-inch tablet will probably work on someone's--on an eight-inchtablet. and so the way system works, if you use these sw qualifiers for your layouts thenit will get pick up by the kind of the one

slightly larger as well. so, it's not actuallytons and tons of stuff you do but just by providing slightly say--like you said, biggertouch targets or slightly bigger font sizes for example in like a seven-inch tablet canmake all the difference. >> rhode: right. and then another bit of tabletwisdom is that, you know, so you're going back to the game model, you don't have tonecessarily scale up your game to fill the whole screen. you can have a fragment thatvaries it size only a certain amount and then on the larger screens just put more or lessancillary information around--on the sides of it.>> hyndman: yeah. >> butcher: indeed. we asked that one.>> hyndman: who watched last night's android

live in action which covers runtime.>> butcher: yeah. this is--this is the topic we covered yesterday on that.>> hyndman: it's a whole half hour of that. >> butcher: yeah, yeah. giving a few techniquesyou can use to design around this issue. >> sutton: and it's--it was great that youguys were looking at the i/o app which is completely open source. so you can just goin and grab the source and copy whatever it's been done in that wonderful software.>> hyndman: yes, the new templates in the--the android templates and that we use. we createnew project and clips rather than getting into that.>> butcher: and jan has joined us. hello. did that answer your question? you're muted.>> hyndman: is john talking or not?

>> butcher: you're not talking.>> hyndman: yeah. >> souza: yes, thank you.>> hyndman: excellent. you got into hangout. >> butcher: oh, well, thanks for joining us.>> hyndman: okay. let's jump to the next one in the moderator.>> rhode: all right then. yeah, there's--this is yan's question in the moderator. yan, doyou want me to ask your question? i don't see...>> gaunt: [indistinct] >> rhode: okay.>> hyndman: [indistinct] lip sync. >> rhode: right. okay. so, right--yan in thehangout is asking, is it a good approach to in-app purchasing to remove ads versus, youknow, within 1abk versus maintaining two applications,

one with and one without? and i said it'sa great idea to use an application purchasing to remove ads.>> butcher: uh-hmm. >> rhode: then--and provided this developerthat android link for a little more information. you know, since you can make it a managedpurchase, you can query the server anytime and find out whether to--whether that userhas purchased the right to have the whatever unlocked ad free premium pro-version of theapp. there's--i think there's significant benefit in not maintaining two different apks.you--to a certain extent, you're exposing your app potentially to a little more riskin that, you know, the code that queries the license server or the, in-app purchasing serveris another attached surface, but i don't really

don't think it opens you up anymore than justhaving a separate pro-version of the apk where somebody can just, you know, redistributethat. so, i think that the--without a lot of extra headaches involved, you can reallysave yourself some trouble just by not maintaining two different apks for the same application.>> yeah. >> butcher: and two other points to add tothat. is a--sorry chris, is suppose to [indistinct] so it goes out their jumping out to, you knowhit, i want to get [indistinct] and you have to then install a new app. i was like "wait,what? now i have two apps installed, so i have to uninstall an app and also any settingsor configurations i may have done it gets carried over, since it's the same application.i don't have to configure anything again.

so, that's a lot smoother and secondly, i'dsay you that all your reviews and comments will be centralized in one place. and somewhata good or bad thing, you can go find that people who [indistinct] for [indistinct] onyour rating, because they paid money for it. they tend to like, justify that sales i think.maybe that's just my idea. and but--yeah, i think having them all in one place is probablya better thing. >> sutton: that's all so quite interesting.>> kelly: yeah, yeah, the... >> sutton: i'm sorry, chris, good for me.>> banes: no, i'm just saying, you know, i did this for friendcaster. i had made thisbefore in-app purchasing was invaluable so two separate ones the free and the pro. andi use to spend a good 10 to 15% of my time

merging code across because you'd have to--you'dhave to do it in one code device then copy it across. and what you find is you're trying--youwouldn't copy it across, correct? so, you have to spend time fixing lamps and stuff.so, in the end [indistinct] >> butcher: more library projects.>> banes: pardon? >> butcher: was that before library projectsor... >> banes: yeah. so, once library projectscome out which then i can stop moving color code into that and then sharing it. it just--itstill took time and so, yeah, and i've always go now with one apk. and other thing is alsois that--because it will look--because you have the free versions in your listening,people will be more inclined to download it

because it's not paid app.>> hyndman: uh-hmm. >> banes: so, and then they can then do thein-app purchase where if they want to so in other words they're not minor.>> sutton: yeah, that was one of the interesting things from io is that there was some discussionaround monetizing app and all of the data seem to point to having an in-app purchaseas being a better way and a more consistent way of getting users to buy your app and ittends to result in more purchases than trying to get them to download a separate apk andpay for that separately. >> butcher: just devil's advocate as a slight[indistinct]. i'd say some people just hate in-app purchases. you'll see comments in themarket saying "they're trying to nickel and

dime for everything, blah, blah, blah," andsome people are just much happier paying one app price up front. i'm not not saying thisis many users or most users, so some people just don't like that and the second pointis having a paid out, let's you put this way in some kind of like sales and promotions,so for example... >> hyndman: uh-hmm.>> butcher: ...our recent 25 billion promotion was--where were put a bunch of applicationson sale like if you only had a free application then, that would have been available to use.so, i'm not saying these great reasons and perhaps you might have both available. so,you know, a free version with in-app to attach [indistinct] first or like a premium versionor whatever that is just like all in-app purchases

already included in the different costs forexample for those people who just really hate it. it's up you, you know, it depends on whoyou're going for and what the market is and so, but.>> sutton: yeah, i think there's a lot of work that needs to be done in terms of userexperience on the developer side because i think it was yahani point--pointed out anapplication that basically was a free application but the moment you run it, it said buy coinsand you couldn't do anything without making an in-app purchase and that's a straight turnoff for users, it's going to be a bit more subtle.>> rhode: oh, i hate those. >> sutton: it's the easiest. one wasn't theonly that you found was there? i guess there

are few out there, but it has to sparky responseand people just turn off straight away, say trying to force people down in an in-app purchaseis a way around paid app. it's seems to be--yeah, more of a quick way to annoy the user thanto get sales. >> butcher: crappy thing to do.>> gaunt: yeah. >> rhode: so, the applications, let's say,you get ten moves and if you send us a couple of bucks, we'll give you another hundred.i hate those, i hate those. but the apps that say, you know, you can--you can unlock thefirst 10 levels of this puzzle game for free and then, you know, pay us a couple of bucksand we'll give you the next hundred levels and you can, you know, you can play as muchas little as you like. we're not metering

your play, we're just giving you access tothe whole thing, those i love. >> butcher: yeah, i think the mail, are oneof the coolest ones i've seen is the timed office. it's like a wind up night which givesyou--lets you play for your first 10 levels or something like that and then it says, "hey,one time early offer, unlock the rest of this game for this one price won't be offeringagain kind of thing." i mean that kind of taps into the human psyche wanting a bargainright like you can get, okay? you're just one chance to play. i thought that was prettyclever and i think [indistinct] just came on to the show on us and said that one itemwas the bulk of their revenue, i think. >> hyndman: uh-hmm.>> gaunt: uh-hmm.

>> banes: just the app, app store as well?>> kelly: the other good thing... >> banes: sorry, mate.>> kelly: sorry. i was just going to say quickly, the other good thing about the in-app billingstuff is that you make a purchase on one device that alone market for all your other devicesas well. >> butcher: oh, yeah.>> rhode: right. well, i hadn't even thought of that. i mean, talking about saving yourdownload trouble. >> gaunt: yeah, you can have it everywhereand you're going to have the free install everywhere and then.>> butcher: free install, yeah, times end. >> and then you have to go back and reinstallthe new one, yeah.

>> butcher: the clip. chris, so you're sayingas well? >> banes: i'm saying as with the app store--imean, i can't even say it. but the app store on ios you can't actually see what the in-apppurchase are. when you download something was our comment on google play. i could bewrong, but none. >> gaunt: interesting.>> hyndman: you can see it is needs the in-app billing permission, so you can go with outthe in-app payments in there somewhere. >> banes: right, but if you download and seethere's an in-app purchase for playing rest of game or something then you know it's goingto be dodging. >> hyndman: play whole of game.>> banes: yeah.

>> sutton: but there's a lot--there's a lotof things around that because one other things i've seen done in the ios store is some ofthe in-app purchases makes no sense until you start using the application itself, likei think there's an application that converts photos into cartoons and everything is likethe splat in-app purchase and the glam in-app purchase which it--you kind to sit there andi'm thinking, do i really need to buy this to put the word glam on screen and it's--it--it'skind of a double edge sword. >> butcher: yeah, i'm not sure if i see thevalue on that. >> rhode: i think it's the value in it ifyou make it cheap enough. i mean... >> butcher: i know, and [indistinct]>> rhode: ...it's below the threshold where

i'll just say, oh, what the heck, from onephoto, i am willing to spend ten cents, click. >> butcher: no, no, i can see the value. idon't see the value in listing them up front when it's completely out of context.>. rhode: okay, sorry. >> butcher: well one of the ones that i quitelike is a german game [indistinct] which we know who has the ads and this little x nextto it to dismiss them and that launches the in-app purchase switches an interesting location.i think we've done that one to [indistinct]. >> banes: lots of [indistinct] in-app purchase.>> rhode: does anyone here live in the hangout want to bring up something of interest tothem before we go back to the moderator or were there questions in the google+ streamthat we haven't gotten to yet?

>> hyndman: there's one [indistinct] today.>> souza: right, excuse me? >> sutton: there's one for matt in the commentsof the google+ page. >. hyndman: we again got another questionhere. >> butcher: let's take a live question.>> souza: yes. it's not a question but feature request. i like you to include the featuregraphic on my [indistinct]. >> butcher: sorry, say that again, i missedthat. >> souza: feature graphic.>> butcher: uh-huh. >> souza: on my application, my [indistinct]because... >> butcher: you would like to localize thefeature behind, correct?

>> souza: yes. this one...>> butcher: just the same or one of this [indistinct] is on the--it's on the team's list of stuffto do. >> hyndman: it is crazy right now, right?you can localize all the descriptions but if you want to localize the great big graphicthat everyone sees... >> butcher: and screen shots, uh-hmm.>> hyndman: ...and screen shots and all that stuffs, yes. you must list things and targetthem into countries so that's... >> butcher: it's very, very much on the radar,it's just when it gets shift. >> hyndman: should we?>> butcher: we hear you, we hear you. >> hyndman: okay.>> butcher: sorry, did you want to read the

question tonight in the comments, shall iread it? >> gaunt: yeah, go for it.>> butcher: so, i was trying to get that comic app with in-app billing working on googletv. they have to send you box, i always get stuck at the billing process or froze in purchaseup to 30 seconds this application stops responding. >> gaunt: so, i've actually replied to thestock over flow question with a blunt face. >> butcher: it's not.>> gaunt: so, essentially, there's--because we don't support subscription billing, thedungeons and subscriptions sample doesn't work as is. needs a little bit of tweakingto change the trial version number as well as some of the extra variables that you addin, so payment type, you don't do that or

otherwise api starts having some problems.so, you need to that and plus when i start doing this, i have a couple of issues that'sjust the google play service and making it work. so i had to leave it for a little while.which i admit is a bit of a bore thing but i do essentially have one bit of code leftfor it before they came back and then it just started working as it should do.>> hyndman: yeah, that happens to me too. >> gaunt: i think it was just the propagatingthe items that you are adding to google play and make sure that it works and authenticateet cera. so check out the blog post , go through all of that, check that are working with youritems and then hopefully you should be able to figure out what's going wrong.>> hyndman: yup. all the standard things apply.

don't test it with the same account you uploadedthe app to play store you can't buy them for yourselves. there's a list of five thingsyou each need to check when you're testing in acronym in the testing--acronym document.>> hyndman: yup. documentation is ridiculously useful. so, yeah, still do check that. it'sjust the sample which needs some differences which what my blog post is focusing on.>> butcher: okay. and there's a question in there g+ as well. robert is asking why inthe 4.12 update for seven-inch tablets, why search bar angled on the landscape put anoption on portrait? none of us here know the answer. i'm going to speculate here that it'sa ux decision. that in landscape you have so much less vertical space that you don'treally want as many angled items. i'm going

to guess that that being the reason but idon't know for sure. >> sutton: i don't believe it's actually awidget in landscape mode. >> butcher: no, it's part of the home screen,right? >> sutton: yeah, where as in portrait modeit's actually a widget that seems to be used. >> butcher: is it? i don't think it is. ithink it's--you can't pick it up or anything. >> hyndman: you can't regardless...>> gaunt: right. >> hyndman: ...i'm slightly confused by thisquestion. >> [indistinct]>> gaunt: we're all just staring at the seven-inch tablet.>> hyndman: [indistinct] can anyone replicate

this?>> butcher: so, i think it's--i don't think it's a widget. i think it's part of the homescreen. >> hyndman: i think it's part of the launcher,yeah. >> rhode: right. i think so too. i mean, idon't think it's like when you drag a new widget onto the screen. it seems to me likethat search box is out of the accessible space. >> hyndman: can you move it right on yours?>> butcher: yeah. and when you drag an icon around here, it disappears to make room forthe app info and the delete button whatnot. i don't actually think its part of the homescreen like work spaces. where i think it's kind of decoration around it.>> sutton: yeah, i can't actually drag mine

around. it's just a bit confusing. insteadit's anchored in landscape but optional in portrait.>> hyndman: yeah. i have--i can replicate that.>> butcher: i read that as it's kind of always there in portrait. but when you go to landscapeyou just have to google now and the--and the voice action, it's a voice search buttons,right? >> sutton: right. yeah.>> hyndman: it would be neat if you had a vertical text bar.>> butcher: yeah, exactly. >> hyndman: i think that's about. anyway.>> butcher: yeah. so, yeah. so, basically i'm saying.>> hyndman: [indistinct]

>> butcher: yeah, you wouldn't ever have likea vertical search box because that would look crazy. i probably wouldn't. and so i'll likepostulate that they didn't want to have another reserved error space in already limited amountof room when you're in landscape would be my guess.>> hyndman: well, i concur. excellent. >> rhode: i think that's a really good guess.>> butcher: why thank you. >> rhode: since you mentioned google now,this is just kind of a random, what it occurs to me. i'm curious to know, who here bringsup google now by accident more often than on purpose. because i know i…>> banes: [indistinct] >> hyndman: interesting. all right.>> rhode: going back to that flight control

again.>> [indistinct] >> rhode: making those little airplanes aroundthe screen. >> yeah.>> hyndman: i can suggest one solution. if you have that problem with your nexus 7 putit in a case because it's really hard to bring up the tall ones to the case.>> banes: is that the official google answer? >> hyndman: of course it is, chris.>> butcher: [indistinct] answer. >> sutton: that's one of the downside of thosecheap third party cases. you get the official aces while then it leaves you enough screenspace to get the swipe from the problem working. >> hyndman: [indistinct] and that it certainlydoesn't.

>> sutton: yeah, this was 25 [indistinct]from pc world. >> hyndman: [indistinct] magnet in it though.>> suttion: nope. you pay 25 [indistinct] you…>> hyndman: [indistinct] >> sutton: …you don't even get the closemagnet. >> banes: i've got a smart cover [indistinct]or it's like a natural--it flaps over and stuff. it's pretty good. it's like a ten ounce.>> gaunt: i've got one of those as well for my samsung tablet.>> butcher: well, for the--for the nexus 7 i guess.>> banes: yeah. >> sutton: i think my next purchase from amazonis going to be some [indistinct] tape and

a magnet. how are we getting on the moderatorquestions? >> hyndman: yeah, how are we getting on moderatorquestion? >> rhode: i'm taking a look.>> butcher: we've got fifteen minutes left. >> rhode: okay.>> [indistinct] >> rhode: andre in czech republic wants toknow, is there a way how to correctly detect if a tablet has a mobile, it's like [indistinct]3g interface. i need to hide or display some features based on that. feature telephonyand connectivity manager, get all network info aren't reliable enough. i don't know.i think i don't--i think i don't have the best answer to that one.>> sutton: it was--i was kind of confused

because he is saying feature telephony isn'treliable enough. and i was wondering whether he is already using packagemanager.systemfeature to detect whether it's on the device. >> rhode: i'm wondering if maybe he just meansit's not precise enough. >> sutton: right. because there are some features[indistinct] >> hyndman: we've got an extra comments inthe bottom here. thanks for your responses but unfortunately there are some tablets thatdon't report feature telephony as one of the system features. i imagine things like…>> sutton: wifi only devices. >> hyndman: …out of the--no the galaxy tab10.1 may have 3g but no telephony because it doesn't--you can't make phone calls.>> sutton: right. but the feature should be

as said that the device has a telephony radiowith data communication support. >> [indistinct]>> sutton: which to me would indicate it should be present…>> elkrief: [indistinct] >> sutton: …if just got a data modem onit. it doesn't necessarily need to have a voice codec as well to handle calls.>> hyndman: interesting. yeah, i can't remember. if the input requires telephony in the manifestwhether it would appear for a 3g tab or not, i thought it didn't.>> rhode: there you go. i do think maybe there's potential in this telephonymanager get networktype, which should tell you like exactly what kind of radio you're talking to at the moment.>> butcher: but that's what's on at the moment,

not what's available, right?>> gaunt: yeah. >> hyndman: lastly saying, querying the simis a good way to go. >> rhode: querying the sim?>> hyndman: which class would you use to do that [indistinct] we'll give him--i'll readout the url. >> drewery: yeah. you could actually use sometools libraries to do it. we had this problem in htc. where if you got a sim that supports2g but doesn't support 3g it caused a bit of confusion, we had to actually query directly.our htc now could find the--on htc devices the custom api that we used.>> sutton: well, but there is in telephonymanager it gets subscriber id which is supposed toreturn an imsi number. i guess--i would guess,

if there's no 3g or no telephony componentthen you wouldn't get an imsi back. i think you get nulled back in that situation.>> elkrief: it's interesting. i've seen a custom roms that do return imsi or even fakedata just to pass some test on the telephonymanager. so i'll try to do some--try to catch overthere enough on false data. >> butcher: yeah. i think i know we've kindof had a similar question last week. it comes down to why you're trying to do that exactlylike--i guess the main thing you want to do is like work out. do you have a data connection,right? which you can use to get connectivity [indistinct] but it's connected. it's prettymuch on your major check. >> hyndman: i was thinking [indistinct] hadalso a setting which said--let's say it was

oovoo and it wanted to allow calls over 3g.you wouldn't want that setting there at all if you have 3g device. but you do have 3g,one that particular if you have or not. it could make sense to have a [indistinct]>> butcher: i know, i exceptionally use it. >> hyndman: yeah, it's up to you because [indistinct]in the 3g but [indistinct] i think its 3g and wifi.>> gaunt: but you only want to show the option when you have 3g as an actual option to takeaway [indistinct] >> hyndman: [indistinct]>> butcher: but you can [indistinct] if you're on cellular or you're on wifi [indistinct]>> hyndman: no, no, just 3g. >> [indistinct]>> banes: [indistinct]

>> hyndman: if you're on it at the time [indistinct]>> drewery: you forgot one thing guys. >> hyndman: yeah.>> drewery: the same competence of certain [indistinct] in this country will allow youto connect to the 3g network, but not use it. i won't mention any names as a droppingsales, so what that means is that you might have a 3g subscription but it will allow youto connect to 3g matrix saying you got 3g but you can't actually send any data up.>> sutton: that's also another issue which makes the whole thing a bit an unreliableis the use of mifi devices. so i've got a wifi tablet and i'm connecting it to my portablehot spot on the phone in my pocket. the tablet it thinks it's got a wifi connection but theactual end connection is a 3g cellular on

the other one. that's how i got around the[indistinct] as an original restriction of only allowing streaming on wifi which is touse a mifi device. do some tethering between the tablet and the mifi device and then youget around any restrictions. although it may be possible it's not a good thing to relyon looping [indistinct] >> hyndman: yeah. i can see where [indistinct]it would be nice for those [indistinct] i would still use [indistinct] correctly thenso be it. but you want to know whether the [indistinct] in there you want to know ifthe hardware supports that capability then i would go to the feature telephony. it'sgot to be [indistinct] >> rhode: as oppose to its quality of usersexperience issue. say maybe you know if you

can't pull one megabit you don't want to offervideo or something like that just do a little performance test at run time.>> hyndman: yeah. you can do--when you're connected you know what network you're connectedon. it's just you just don't know which networks are necessarily available to you.>> sutton: the performance test would be the best way to go definitely because of wifisituations like io. >> butcher: yeah. and definitely i think aboutnot making a setting as well. think about just doing a test and you get a warning.>> hyndman: uh-hmm. >> butcher: like saying you know signal strengthlooks weak. you sure you want to do a video call or whatever it might be.>> elkrief: yeah, that's what we do.

>> butcher: just making a setting isn't alwaysthe answers. >> hyndman: read reto's blog post on magicboxes as well. what's the problem with network? you probably don't get sub type. maybe youhave to be connected to that network at the time, you know. he--what do we have here?he has said he's using network [indistinct]. have we got anymore in the moderator?>> sparky: we do. and we have a couple more. >> sutton: and questions from andrew kellyin there >> sparky: andrew kelly, yes indeed. aboutviewpager, a list of pages, andrew do you want to ask?>> kelly: yes, since we are--i've got an app that i'm writing for a client at the momentand i got a listview that goes into a viewpager

kind of mode when they want it drilled downinto the details of the list. and so, in the listview i want to show all the items available.>> sutton: uh-hmm. >> kelly: but in this the detail mode runningon the viewpager, i want to skip over items that don't have any further details. and so,i'm just trying to work out if there's a way of skipping with the fourth element of thearray without actually having to create a new array with that fourth element missing.>> butcher: yeah, well that's in the adapter, right? in the get size and method. so, these...>> kelly: yeah, so i got my get count reduced--returns a fewer number of--yeah, a lower value basedon how many rows there are. the get item still gets called for the place holders that i don'twant to--the ones i want to skip over, so

if i want to skip over element number four.>> uh-hmm. >> kelly: element, you know, get out and stillgets called for element four. >> butcher: yeah, you know. you're alwaysgoing to get passed in the time being increasing position numbers, but i think you're goingto have to maintain your own mapping from position to the real item index in your--inyour data. >> kelly: yeah.>> hyndman: correct. >> banes: [indistinct] it sound like you needto--you get them and you just need to--each time iterate from the start and can't untilyou find x numbers of items which don't have a flat board do [indistinct] whatever.>> yeah.

>> kelly: yeah, i 'm just trying to avoidthe constant to be going looping with the array...>> banes: yeah. >> kelly: ...again and again and again towhere that [indistinct] was. >> butcher: [indistinct] trying to. yeah,can you just do that once? can you do so the first fast item is index one, second itemis index four, the third item is index five or...>> elkrief: no, maybe he has a hashmap. >> butcher: yeah.>> elkrief: only in one second, you have all the index's made through the items with thedetails. and then, when you use the get item than you'll just need to verify that you havethe index on your attachment or some sort.

>> sutton: well you could even--if you put--putit into a hashmap, you could then get the entry set back and iterate over the entryset. >> kelly: yeah, some adjustments. that's agood idea. >> hyndman: yeah, you just run through once.are we talking about an enormous list here though? is this something you need to be relaxed?>> kelly: no, at 20 at most so... >> hyndman: you know, i just run through itto be honest. >> sutton: yeah. they said the possibilityfor some of the items to change. so, could item number two be populated then later downthe line then item two could not be populated? >> kelly: not with out going all the way backout and all the way back in through the original

on a summary list.>> butcher: it's now [indistinct] go take a hit and did it once...>> hyndman: yeah. >> butcher: ...and have this simple of hashmapor something. >> keely: yeah, and that was.>> banes: it's currently [indistinct] >> kelly: and i thought it might be havingsome sort of--is enabled kind of true or false, kind of method that i could just call forthe ones that i want to skip over. >> hyndman: no. i mean, you have full controlof your adapter, it wouldn't... >> kelly: yeah.>> hyndman: ...pass the adapter then ignore it. it would--self service, sorry.>> kelly: that's right.

>> butcher: well, you can extend viewpagerand make your own implementation in that... >> hyndman: ignoring the item in viewpager.>> butcher: yeah. >> sends the bytes and sends it.>> rhode: yeah. the reason that i pointed to the sherlock demo that's out there on code.google.comis because the viewpager implementation on that little sample app basically it, you knowalthough viewpager sequential, it basically just pokes a random access value into thetab manager. and you see--because it's mashing up viewpager in tabs together. and so it'skind of a way to jump from your sequential access of the viewpager into the random access,and all you have to do is intervene at that point to tell it which you know, page or tabto show.

>> hyndman: yeah, you could do it so thatwhen you go to item number four, it will also scroll down. you'd scroll to four and thenit would automatically just move onwards because it's an empty item. basically, would knowthey've gone past item four. this page is intentionally left blank, something. no, buti think... >> butcher: no, don't do that.>> hyndman: no. [indistinct] i would definitely go with what you said before.>> butcher: five minutes left. >> hyndman: we're squeezing the last--howmany have we got left in the moderator [indistinct] can get the next [indistinct]>> rhode: four, four more. >> hyndman: four more and four minutes, go.>> butcher: go.

>> rhode: where in the source code can i findthe 4.1 lock screen, google it. >> hyndman: did you read all of it?>> rhode: i read also that the developer.android.com documentation is part of aosp. i assume so,it's--i mean we put the aosp copyright in our sample code and stuff.>> sutton: it is. i'll put a link to the location in the notes.>> butcher: yeah, and i think the lock screens is called the key guard, so if you grab codeswith key guard, i think it's the thing you're looking for.>> rhode: sounds right to me. plus one. >> butcher: next question.>> rhode: can i retrieve data from google maps apis so i can use it for my locationbased service. yes, use the google maps api.

developers.google.com/maps.>> sutton: next muffin. >> rhode: i have a device with only ethernetconnection, how can i find if device is really connected to the network. try to push sometraffic through. >> butcher: hit, [indistinct] it will getactive in a minute. network [indistinct] is connected. next.>> sutton: try to connect to google's webpage. >> butcher: no, if it's connected it willdo that, right? it's connected; it's like the blue, you know, how your wifi turns bluewhen there's really bits flowing. i'm pretty sure it's based on the same logic.>> rhode: and this just--sort of code features prominently i believe in reto blog articleabout writing battery friendly apps. so, yeah.

you can search for it on developer.android.com.we got blog articles and google io talks that all touch on this. is anybody aware on howandroid handles daylight savings time changes? it does, also...>> hyndman: even if the code... >> rhode: even if it, even if it didn't, imean generally speaking you want your device to get it's time of day from either your phonenetwork or ntp, in with and in either of those cases. you know, their smart to know--smartenough to know about daylight savings time and when it happens, then it--your devicewill notice that it's out of sync and it will reset itself.>> hyndman: but they are also asking which part of the source code deals with this, soit looks like they want to get some, pull

something out and reuse maybe.>> sutton: i think that's actually kind... >> butcher: time is hard.>> sutton: yeah, i think that's actually handled by a lot of the lower level linux stuff that'sunderneath android. >> hyndman: i wouldn't be too surprised. javadeals with a pretty... >> butcher: yeah, maybe, no. java does ithorribly. look at like [indistinct] the time or something like that perhaps.>> banes: don't you share the time, it's massive. >> rhode: i seem to remember, in the run offto jelly bean for a little while we actually had a bug where the ntp fallback wasn't workingright. >> hyndman: uh-hmm.>> rhode: so.

>> sutton: android actually has a utilityclass which re-implements and the getting the number of milliseconds from the apk. becauseit's so unreliable incident past the java, so.>> butcher: was that in system.get current millis or...>> sutton: there's a--i can dig out the method for this. there's actually an android versionof system.get current time in millis which is used for key events.>> butcher: uh-hmm. >> sutton: and you can pick that out and basicallyit does the same thing but i'm guessing it's there because there's some discrepancies abouthow it gets reported through. >> butcher: it was interesting because i readabout the three or four different ways you

get the current time. i think the most accurateone was someone from the kernel which is exposed at for a native [indistinct] think you cando. [indistinct] trying to get out. >> hyndman: and with that.>> butcher: are we done? >> hyndman: i think we are done.>> rhode: that's it from the moderator. >> hyndman: tablets.>> gaunt: tablets. >> hyndman: we were going to say tablets tentimes we didn't say enough the last time. >> butcher: not enough time.>> elkrief: even to, speaking to. >> hyndman: damn it.>> butcher: any last minute questions from the hang out?>> elkrief: yeah, just a little question about

the subscription and trial version periods.>> hyndman: yes. >> elkrief: if i want to do and app 12 fora trial period, but it's not--this is basically a subscription app. say you want to do likea snooze button or a silent button widget and instead of doing it into a paid app orsomething like that, can i operate on a trial version with some sort of subscription?>> hyndman: do you have anything--if you're not using subscriptions, the only option wouldbe to not [indistinct] and then pop up a dialogue after 30 days or however long you want tothe free period to be. subscription is the only part of google play that offers a freetrial period for you. and for that free trial, you have to sign up to the subscription, youhave to have the valid kind of details in

place, and at the first billing cycle is completedand starts charging you. >> elkrief: uh-hmm.>> hyndman: so, yeah... >> butcher: it's cool. all right. i thinkthat's all we had time... >> elkrief: do you want to see more androidson my offices? >> sutton: [indistinct] after that.>> elkrief: will close up with a tour of the google office.>> butcher: hey, how about that? >> elkrief: oh, and over there we have...>> hyndman: well, there's an apple. what's that [indistinct]>> elkrief: no, yeah, it's--at the bathing room, you know, everybody is fighting so;we're doing it, so.

>> butcher: have fun.>> hyndman: cool, we just did. thank you very much for joining us today. we will be backagain next week. >> elkrief: will be back.>> butcher: yeah, so hopefully you'll maybe see some people watching in dã©fis, berlinthis weekend. matt you're going to be in paris tomorrow, you'll speaking to the wonderful[indistinct]. >> gaunt: paris on friday, yeah. berlin onsaturday. >> butcher: don't forget there are some greatshows on friday. there's the app clinic and the friday games review and join us next tuesdaycenter line action or see you all next week, next wednesday hopefully for more office hours.>> sutton: and as ray turns one [indistinct]

three and everything that's going on [indistinct]in the way. i swear in last week he was offering badges.>> butcher: bye. >> hyndman: bye.>> bye 23

Share this post