hello, Ivano!

Good afternoon Mark!

Version 7.5 of the book was released about 20 minutes ago, so be sure to grab your copy

beyond that, how can I help you today?

yesssss

well,I had an existential trouble over db. I updated the db mysqlite adding a fifth column, it works pretty coolish. Just I choiced the way easy, that is just adding one column with ALTER SET. Now my concern is, in terms of performance, would have not be better to use a different table instead of a fifth column? So I would have use an Inner Join linking the primary keys.

eeek

on the whole, I would expect an inner join to be worse, not better

eek

i am surprised

now, if the cost of the ALTER TABLE is such that you can't afford the upgrade time, then perhaps going with the inner join is the lesser of two evils

why then there is all this fuss in making different table if the performance become worse?

tables*

I don't know why you would make a separate table versus just adding a column to an existing table, so I do not know what "all this fuss" is

( i can afford upgrade time really well)

lol

OK fine with me

glad to have discovered that easy is better

OK thanks

I am going to follow a movie, but leave the chat to observe if somebody will come

OK

Hello!

hello, Lawrence!

Hi Mark, thanks for the wonderful book!

thanks!

Version 7.5 of the book was released less than an hour ago, so be sure to download your copy

Ivano is lurking, so if you have a question, go right ahead!

I just did. I wanted to ask about Wear and whether you were planning to add material on that, other than notifications.

(BTW, Ivano, if you come up with another question, chime in)

I do not plan on covering Wear apps at this time

(i am not a pokemon to catch although)

(not Mark one per day is more than enough thanks)

Okay. It's a bit tricky and Google's docs on it are sparse, so I was hoping to find more thorough stuff in this book.

sorry

you'll see that I tend toward stuff that is pure Android, versus things with proprietary ties (Play Services, Wear, etc.)

Oh okay. What fraction of Android users are on non-Google platforms, do you know?

well, pretty much all Chinese users have non-Google devices

the one survey I saw a couple of years ago indicated that the Play ecosystem was less than half of the world's Android devices

but I haven't seen anything much recently on the topic

Oh interesting. I only sell on Play, mostly because I figure that the rest of the ecosystem has poor ad revenue.

not surprisingly, Google doesn't comment :-)

the rest of the ecosystem is fairly fragmented

there's probably 100+ markets in China alone

I had a conversation with a large Chinese device manufacturer, and the representative I was talking to was stunned that many Western developers would ship only through the Play Store

it's pretty much par for the course in China that you ship through lots of stores

and I presume that there are services that help developers do that

Yeah, like I said my impression is that ad revenue is poor in developing countries, but I could be totally wrong about that. Many of the deals I make only make me money for U.S. users.

(interesting- thumb up)

I only have a high-level view of the situation, so I can't make any recommendations on ad networks, specific non-Play markets, etc. that would be worth considering

(:) positive)

but, with regards to Wear, while I won't rule out writing about Wear apps, it's not on my to-do list

I get a decent number of requests for Wear support -- more than I'd expect. Judging from the relatively little activity about it on SO, I'm guessing not too many developers support it.

for these sorts of specialized segments, I try to do a bit in the space (e.g., Wear notifications), then wait to see how the segment fares

I think all the specialized segments (Wear, TV, Auto) are only doing so-so for developer adoption

which is why Google went to some lengths to make it so that more apps would ship on Chrome OS "out of the box"

at least on touchscreen-enabled Chromebooks

I admit I'm a bit surprised how much coverage you're giving to Chrome OS -- perhaps I don't realize the importance of the Android-on-ChromeOS thing.

it's only a chapter

Oh I guess I'm referring to your various blog posts recently.

length-wise, I probably spend more time on Wear-friendly notifications

the reason for the blog posts is partly because it's new, and partly because "you only have one chance to make a first impression"

and this is the first impression of Android on a desktop-y environment

at least in terms of official support (vs. Bluestacks, JIDE, etc.)

I am surprised google did not exploit the fact an ad hoc linux distro could have been devised, i am assuming they did not meet Mr. Penguin philosophy not big data oriented. after all I am playing android studio on an I3 lubuntu and is fastest than on my i7 4th gen windows 8.1

Ivano: I cannot really comment on that, sorry

sure sorry for the lack of neutrality, I fully understand:)

Lawrence: since Android apps will ship to touchscreen Chromebooks automatically when support rolls out "for realz", I worry that developers will ignore the platform and users will have a lousy experience as a result

this is different from Wear/Auto/TV, where developers have to explicitly opt into the platform

Makes sense.

but you'll see less on Chrome OS going forward, unless I run into show-stopping problems

Is writing the book your full-time gig?

I do a mix of things on a freelance basis

historically, the book had been a full-time gig

hello, Steve!

Hi Mark!

Steve: Ivano and Lawrence have had time to ask questions already, so I'll turn to you -- do you have a question?

Yes.

I'm trying to do something pretty basic: I want to put activity files into their own package/folder.

I'm having problems getting that to compile.

so, if your package name is com.steve.awesomeapp, you are trying to put your activities into com.steve.awesomeapp.ui?

Exactly.

other than having to add the import statement for com.steve.awesomeapp.R, that should work fine

plenty of apps go that route

I don't think I have any book samples that do, just because my samples are all fairly short

but the APK edition of the book itself has activities in a variety of packages

what is your specific compile problem?

I think it's what your solution addresses: "package R does not exist"

OK

in Java, you automatically import all classes that are in the same package

ok

so, if your activities are in com.steve.awesomeapp, you do not need to have an import statement for R, as R is also in com.steve.awesomeapp

What is the APK edition of the book?

the book is published in PDF, EPUB, MOBI/Kindle, and as an Android app (APK)

you'll find a download for it on your Warescription page

Great. I'll look for that.

I think I should be able to get this working now and will give it a try.

anyway, if you move classes to a different package, that's fine, but you then need to add an import statement for the R class

That makes sense.

I'll also need to modify the manifest to include the subdirectory in the activity name, I take it.

yes

So those two changes?

if you used Android Studio's refactoring tools to move the class, the manifest may already be taken care of for you

I didn't use that but will give it a try.

Android Studio supports drag-and-drop to move classes to other packages

it then tries to fix up all external references to that class to match

That's good to know about.

Ok. I'll try what you said. Thanks so much!

you are very welcome

if anyone has any questions, go right ahead!

Do you have questions for app developers? :-)

besides Wear apps, what things do you want covered that are not being covered in the book?

I can't promise anything, but it's always good to know where pain points lie

javaRX, syncadapter, refactoring for android

Ivano: I am not sure what you mean by "refactoring for android"

RxJava might get in there next year, as with SyncAdapter

well you magistrally covered in your chapters for testing, the concept of Junit vs Instrumentation but...

I switched a bunch of my app to RxJava a few months ago. It was a good experience, but it's a large library that caused some dex problems.

testing and refactoring are a combined skill today in times of TDD and BDD, so android has some particularities, like a lot of super classes, overrides everywhere, ad hoc classes, that could be covered with some tips, more android SDK focused than java.

yes, I do not cover the 64K DEX method reference limit in much detail

the problem is that since I do not do that much in the way of massive refactoring, I cannot really provide advice

with new APIs and such, I can craft examples and explore beyond what documentation offers

clear and understandable. Also I would add regarding RX java

Mark: To answer your question, though, I don't know. I tend to just jump into the book looking for whatever I'm working on at the time. It was helpful for home page widgets, especially since I have a list widget. Wear was the one time I didn't find what I needed.

Lawrence: that's fair, thanks!

now java 8 is supported this means users will be able to use the new functionalities as lamba, default interfaces, and mainly streams that are really similar to rxjva concepts

yes, one of the reasons why I held off on RxJava was waiting for first-class lambda support

that is why I might cover it next year

I don't know if the book covers this, but it's related to Steve's question: How to structure the app? Google's examples (and yours) are simple, so everything is thrown together. I wish Google would open-source their larger apps so that we could see what a well-structured large app looks like. Stuff like how do you decompose classes into packages, are fragments still a good idea, etc. Large-program issues.

by and large, that's the sort of thing that goes beyond what you are likely to find in most books

Exactly!

But where to find it then?

conference presentations/meetups would be one candidate

blog posts/articles from developers of larger apps

personally, I haven't written what I would consider to be a large app

But if you do consulting, you might have seen some.

hence, as with Ivano's refactoring request, I lack the experience

yes and no

I have to be careful, with respect to non-disclosures and stuff

what matter is that you are a great teacher. I do not know how will go with the new documents SO functionalities, but your way to explain concept is number 1, even better than official documentation imo

Ivano: thanks for the kind words!

is true

Lawrence: plus, I generally want to have code that demonstrates what I am talking about

and I can't devote the time to write a large app for the sake of writing a large app

Related questions, like should Application be subclasses (or used at all), how to deal with globals, singletons, how to pass Context to singletons, that sort of thing.

architecture

Yes, architecture.

Lawrence: while fine topics, those are the sorts of things that usually are better for a Stack Overflow question, or a blog post

architecture gets back to the large-app problem

and I'm not really a strong MVC/MVP/MVVM advocate in the first place

I am not saying that any of those are bad topics to want; they just are bad topics for *me* to deliver

Well I wasn't looking to see a large app, only a bullet list of advice, like "- Try to put all activities into a sub-package" or whatever. That wouldn't violate non-disclosures.

I have lots of limitations

Yeah I'm not a fan of MVC either.

yes, but, that's as much a matter of taste as anything else

Is it?

I was just in a chat with others on that very topic, for code organization

it's not a cut-and-dried, This Is The One True Way sort of thing

it starts to get into team roles (e.g., where do you start dividing into libraries), version control strategies, etc.

and that's waaaaaaaaaaaaaaaay outside what I want to get into

Right.

right

again, all valid concerns, just not what I'm really in position to do, in the format of a book

anyway, that's a wrap for today's chat

Thanks Mark!

the next chat is Saturday at 9am US Eastern

have a nice time mark and Lawrence

this transcript will go up on https://commonsware.com/office-hours/ shortly

thanks

have a pleasant day!