3 ms·

Inside the gPhone: What to expect from Google Android

5 points by bfioca 19y ago

stillmotion 19y ago
This article makes out that there will be a gPhone. They denounced it, the machine doesn't exist, nor will it in the near future. Google is an application company, they develop for platforms.
dkokelley 19y ago
Has anyone downloaded and played with their SDK? Any projects to report?
- davidw 19y ago
  I'm porting Hecl (hecl.org) to it, and went to meet some of the Google guys in Munich last week: http://journal.dedasys.com/articles/2008/01/29/munich-android-mobile-meeting-report http://journal.dedasys.com/articles/2008/01/29/munich-androi... There are definitely some open questions, but I'm fairly sold on the idea.
  - Tichy 19y ago
    You were there? So was I - too bad, it would have been nice to meet somebody from news.yc...
    - davidw 19y ago
      Yeah... I asked Dan a question in the first group of the 'internals' session after the introduction, if you happened to be in that group. I stuck around pretty much until the end. Would have been fun to meet up indeed.
- iamwil 19y ago
  No project, but I did play around with it. Aside from the general complaints about getting use to Java after Ruby, I liked how the application fit into the overall OS. However, one has to adhere to an application lifecycle, which was a little different, since we're use to pretending that the application we're writing is the only one on a machine. The UI isn't programmatic, like Swing, but rather is reminiscent of web programming by defining the UI through an XML file.
  - ardit33 19y ago
    no, the ui is programmatic also. But if you just follow their examples, they mostly use the xml descriptor, which I find are not useful for small applications.
- ardit33 19y ago
  I am playing around with it, and it is good and bad at the same time. The good: 1.Their SDK and emulator are pretty good, even at this early stage (few bugs). Not everything is documented, but at this early stage it is good enough. 2.There is a lot flexibility, many widgets to choose. It is very powerful when you compare to other platforms. Lot's of layouts, lots of things you can use. You almost feel like a kid on the candy store, not knowing what to eat first. 3.You can do UI both programatically or with predetermined descriptors (xml). In large applications this is pretty useful. 4. There is a resource file called R, which get's autogenerated and includes all the resources. Given my experience, (I have a Constants file in very popupal J2ME app my company is doing, that does the exact thing) it would end up with too much stuff into it. Very useful for small applications though. >>>>>>>>>>>>>>>>>>>>>>> >>>> >>> The Bad: 1.The API is not elegant as it could have been. Some classes extend other classes, which extend other classes, the result? Something like Gallery,class has over 50 methods, and a bunch of constants. 2. They included everything that is under the sun, making it confusing, as there are 10s of ways to do something. This is a Java problem also, but their API made it worse. (compare it with the pythonic way of thinking as there should be one very obvious and easy way to do something). 3. Too much XML.(which I really disslike). You can describe screens, and UI with XML descriptors, (xml is very lame, something more looking like CSS would have been much better). Even the Manifest is fricking XML. There was absolutely no need for this. Sun's J2ME has the manifset in plain text, which makes it simpler. 4. Wierd namings and terminologies, make it confusing to get started. Events, are called Intents, there icicle, bundle, and all other really wierd naming allover this place. Did they do it purposely, so beginners will be so confused and wont be able to do anything with it? >>>>>>>>> So, KISS was defenitetly not the mantra when building the android. It was more "Let's include everything under the sun", so the developers will like it. Well, that's great, but don't give me 10s of way to do something, which are just little variants of each other, and none does the exact thing I want to do. Since I do Java for living, and Python for fun, I can say that I was slightly disapointed by android. Lot's of power, but not elgant enough to be called a perfect platform. So, yes, I probably would be developing for it, but not necessary just for fun.

5 points by bfioca 19y ago

This article makes out that there will be a gPhone. They denounced it, the machine doesn't exist, nor will it in the near future. Google is an application company, they develop for platforms.

dkokelley 19y ago

Has anyone downloaded and played with their SDK? Any projects to report?

davidw 19y ago
I'm porting Hecl (hecl.org) to it, and went to meet some of the Google guys in Munich last week: http://journal.dedasys.com/articles/2008/01/29/munich-android-mobile-meeting-report http://journal.dedasys.com/articles/2008/01/29/munich-androi... There are definitely some open questions, but I'm fairly sold on the idea.
- Tichy 19y ago
  You were there? So was I - too bad, it would have been nice to meet somebody from news.yc...
  - davidw 19y ago
    Yeah... I asked Dan a question in the first group of the 'internals' session after the introduction, if you happened to be in that group. I stuck around pretty much until the end. Would have been fun to meet up indeed.
iamwil 19y ago
No project, but I did play around with it. Aside from the general complaints about getting use to Java after Ruby, I liked how the application fit into the overall OS. However, one has to adhere to an application lifecycle, which was a little different, since we're use to pretending that the application we're writing is the only one on a machine. The UI isn't programmatic, like Swing, but rather is reminiscent of web programming by defining the UI through an XML file.
- ardit33 19y ago
  no, the ui is programmatic also. But if you just follow their examples, they mostly use the xml descriptor, which I find are not useful for small applications.
ardit33 19y ago
I am playing around with it, and it is good and bad at the same time. The good: 1.Their SDK and emulator are pretty good, even at this early stage (few bugs). Not everything is documented, but at this early stage it is good enough. 2.There is a lot flexibility, many widgets to choose. It is very powerful when you compare to other platforms. Lot's of layouts, lots of things you can use. You almost feel like a kid on the candy store, not knowing what to eat first. 3.You can do UI both programatically or with predetermined descriptors (xml). In large applications this is pretty useful. 4. There is a resource file called R, which get's autogenerated and includes all the resources. Given my experience, (I have a Constants file in very popupal J2ME app my company is doing, that does the exact thing) it would end up with too much stuff into it. Very useful for small applications though. >>>>>>>>>>>>>>>>>>>>>>> >>>> >>> The Bad: 1.The API is not elegant as it could have been. Some classes extend other classes, which extend other classes, the result? Something like Gallery,class has over 50 methods, and a bunch of constants. 2. They included everything that is under the sun, making it confusing, as there are 10s of ways to do something. This is a Java problem also, but their API made it worse. (compare it with the pythonic way of thinking as there should be one very obvious and easy way to do something). 3. Too much XML.(which I really disslike). You can describe screens, and UI with XML descriptors, (xml is very lame, something more looking like CSS would have been much better). Even the Manifest is fricking XML. There was absolutely no need for this. Sun's J2ME has the manifset in plain text, which makes it simpler. 4. Wierd namings and terminologies, make it confusing to get started. Events, are called Intents, there icicle, bundle, and all other really wierd naming allover this place. Did they do it purposely, so beginners will be so confused and wont be able to do anything with it? >>>>>>>>> So, KISS was defenitetly not the mantra when building the android. It was more "Let's include everything under the sun", so the developers will like it. Well, that's great, but don't give me 10s of way to do something, which are just little variants of each other, and none does the exact thing I want to do. Since I do Java for living, and Python for fun, I can say that I was slightly disapointed by android. Lot's of power, but not elgant enough to be called a perfect platform. So, yes, I probably would be developing for it, but not necessary just for fun.