* Android API is not properly used the Java Method.
* Only Show the All Package.
* No guide for Developer.
* Many basic application not implement in Android.
* Bugs,Bugs,...

Hi,
It seems that a sharp and hot discussion going on .. It is really a pretty nice challenge introduced by a very huge firm like google which had created a very huge expectation. As stated by some one if it is from a small company there may be some little expectations and if we are not satisfied up to the expectation we will have some dissappointments.
But when we are not satisfied with a product from company like google the disappointments will definitely going to be very high.

From a developer view
Plus ...

1)Almost tried to cover all the aspects in the mobile platform(Included all the necessary classes and methods which we get from reference index)

Minus.
1) No proper documentation and we can say almost no document at all.
2)Lack of comments on classes and methods
3) Improper tested version of code samples and emulator
4)Functionality of emualtor is not stable at all. Lots of performance issues.
5) Missing of basic features like media handling , menus, recording etc
6) very bad forum support ( Google Android community- Developer Forum) .No help at all from the googlers only some other developers came to help which they also be unaware?
7) very hard to use database implementation for storage

As final it seems to be developed only for high level developers who can understand a bit about the new platformand make use of it .. But not easy for the novice developers or beginners who want to learn a new platform.

As stated the version had got released quite a long before the completion which is not expected from google.

[font=Courier New]Hi,

this is really a very inspiring discussion
gladdly contribute my two cents:

compared to times before - developing on android is really joy
I think this platform has huge potential in pushing things forward and help all these mobile visions finally become reallity

evolution:
the age of personal computer
the age of portable computer
the age of pocket computer

if the iPhone did ring in the aera of always-carried always-on/-line computing
then android could really let it take of (bring it to the masses)

regarding the reservations:
* this is just a 'early access developer preview' = not ready
(in contrast the iphone sdk will be released far after the device came out)
* no bugtrackking? - the apis are in ongoing change - not ready yet.
It's too early to collect bugs because most of them are still 'not yet implemented features'.
* documentation lacks? - personally I'd better like the folks at google implement missing functionallity
than writing documentation.
and community driven knowhow is much better anyway. it thrives for example here at anddev!
* locked phones? apache licence is the only way to gather wider industry support right now because companies want a way to protect their intelectual property
if the whole platform is released open source it would be hard for carriers to justify crippled or locked phones and loose this competitive advantage. After all I think common awareness could arise that in order to use all the features you have to somehow pay for the hardware. if you can get a subsidized phone but can't use all your favourite apps (and gain the value they provide) a lot of people would rather choose to pay for the (full) hardware as we all already do for other devices (benefit you win is valued / worth money)

nevertheless privacy will be a big issue! in this always connected world
and only time will tell how 'open' this platform really will be...
everyone is free to create a better 'truely open' alternative but why not watch learn and use what google has done so far?

[/font]

Although I am still learning, there are two challenges that I have run into:

1) Testing on an emulator is problematic. There are many limitations to the emulator, namely, the GPS and the SMS.

2) The lack of documentations is also a problem. If it weren't this community, it would have taken me ages to figure out how to implement some of the features.


Good Luck on your interview.

Hi there folks,

I’m reading some of the replies and I realize that some folks never use the windows mobile to develop...

But there goes my review:

* The emulator is the critical point, is very slow.
* Lack of documentation is not a problem. We got ironRuby or Ruby.net or ironPython and realize that big companies are not interesting having a good documentation. but something to compete.
* Very hard to use database implementation for storage

In general I’m happy to have some possible choice in mobile area to create new stuff.

I think in general the SDK is just ok, it does feel "half baked". But I think that's the idea and I commend Google for that. Unlike the Apple and the iPhone which we've all had for 1/2 a year already with out an SDK. I think this approach shows incredible promise. If this "preview" version of the sdk is already a "C" (see the rating by the dev community in the thread), what will the SDK look like in two years when there are millions of devices out there -- surely it will be an A++.

Pros:
1) It works well and runs fast (on my MacBook that is)
2) UIs can be done in XML!

Cons:
1) Only supports ARM V5TE (won't run on many of HTC's current products including my Touch )
2) No CDMA support...yet
3) I know I can unofficially make programs in C/C++ but why did Google go with their own Java implementation?
(no official J2ME)
4) What's with the awful documentation?

I wanted to produce programs as a way of learning some Java and to enter the developers challenge but the documentation scares me....

Most of my experience with android isn't much since I'm not a java programmer and barely know much about it, but the android SDK was easy enough to install and the tutorials were easy to follow and the help here is great, but the one thing that I don't like is that the emulator won't start for me because I'm using a PPC mac so yeah... android needs this support. Anywho, that's my comment. ^__^

When trying to work on Android to make a commercial product, I think the following requires more attention.

1. API Documentation. It will be great if the documentation can be more meaningful other than just a function head.

2. Information on what Google is currently working on. The current android is still under development and a lot of pieces are still not well-done. Without the information of which part Google is still enhancing, it is very hard for a company to decide which part I should wait for Google and which part I should do my own implementation. For example, there is something looks like a scene-graph toolkit but not announced in the documentation. It is not completed, not used in the Android example. Is Google still working on it or I should implement my own scene-graph API?

3. True Open Source. Google donated Android for everyone to use freely. But this is not Open Source! People out of google cannot contribute to the development directly.

I like to think myself a j2me expert and a competent windows compact programmer. I generally like the SDK, most of the stuff is exactly where I thought it would be, it supports generics and auto boxing, there is lots of javadoc , generally a lot more to like then to hate.

On the plus side:
1) IDE integration, at least as good as eclipseme or using visual studio for windows compact
2) Does(or will do) everything that j2me does and then some
3) Permissions handled at install time. It's a pain dealing with the fact that every other operation you might want to execute might be canceled by the user. The Android way the user can look at the software vendor at install time and just not install the stuff if he doesn't want it to do everything it wants to do.
4) Quake!
5) Hooks into the camera, gps, phone, sms... there is just no solution for j2me that'll work on even half of the phones that have all those features and can run j2me.
6) The strings XML persistence sets everyone up for a very easy time internationalizing.

Now for some negatives:
1) Why is the ui not more like swing(though could be worse, we could have to do our own garbage collection like in SWT)? Why must they provide most of the examples for the ui in XML? I want to use XML for ease of maintenance, I want to code it in java first when I learn it.
2) Using a part of the api in juint is not trivial, so there are very few tests that can be written.
3) No jdbc drivers, or serversocket before the contest is due.

I must say I disagree with a couple of points made in previous posts:
1) Lack of documentation - you can not use logarithms before you learn to add, some familiarity with another mobile framework is definitely a prerequisite. If you want to get started right read
Wireless Java, making android do most of the stuff you want is trivial after reading that book.
2) SqlLite is bad - No, SqlLite is good. Its not a replacement for a fully featured relational database, its a replacement for new File("c:test.txt"). If you need a fully featured relational database host it on your server and use http(which is very well documented) to get data from it.

What really bugs me are the errors in the documentation and example code.

Just started with Android. However, I gave it a lower grade because of how hard it is just to get it up and running.

It took me a long time to figure out how to do it (problems w/ Eclipse and the Android Plugin). I couldn't even figure out how to install it on Linux.

However, the SDK itself isn't too bad. However, I am just way to overwhelmed with the new classes and the XML... so it'll take some time getting used to.

compare to pc's development, this is just................

anyway, i found too many stupid bugs and many hardcode part in the SDK, for example

- _id for your database's id.
- listview in tab
- not good enough documentation.
- not good enough error report (you can do something wrong very easy and you don't understand wtf is it)
- freezing emulator.
- it's not that OPEN!!!

i have a lot of problems with the SDK, I just can't remember the shit story

In Android development I'm grade C because I'm beginning, but in PC I'm grade B and I develop in C++, Java, C, Java ME, Java EE, Pascal, VB, Delphi and Assembly(8086).

I've tried out bestratingfor.me. They can add excellent ratings to your app. Wonderful result.