Romain Nurik posted an excellent article today on why you need to hand edit layouts for Android rather than using a WYSIWIG tool

First off, there will be some obvious bias in this post. I have been very fortunate in my working career to work with some truly great product managers, designers and artists on mobile applications and so for the benefit of the app consuming public none of my frankly brilliant crap UIs have ever seen the light of day. I do however still look forward to the day where by I can put a cheat mode into one of our own applications that resorts to developer place holder graphics! Some great examples of what games look like before an artist breathes on them can be found on this thread at Toucharcade

If you take the same view point as me, that to make a truly decent user experience for a given platform especially on mobile,  then you need to understand the technicalities and nuances of developing for that platform. A great example of this would be the difference in the back metaphor on various devices,  on Blackberry and Android you have a dedicated back button, iOS the use of UINavigationController and for Java mobile devices the use of soft keys and context menus.

Knowing these interface cues is why there needs to be a close interaction between developer and designer. It makes sense if a fast utility application is required to make  sure standardised controls of the platform are used where possible  (ie ContextMenus on Android vs UIActionSheet on iOS vs Softkey driven menus on Java based devices) otherwise a lot of development time can be expended on making custom UI controls that may or may not be necessary, if all stakeholders are aware of what components are available on a given platform. For this reason alone it is worth at least having a WYSIWIG  editor so that designers who may be new to a platform (Java MIDP2 is younger than publishing to the web, and Android, iOS, WebOS are relatively new, let alone the even newer platforms mobile designers will have to contend with in the future ie Meego, Bada etc etc) have the ability to experiment with the UI widgets that are available for a platform and produce mockup guis that a developer can then bring to life!

The other area where at least having a fast preview of changes to the UI code is where a WYSIWIG editor can really help, or at the very least a very fast way of previewing on a decent simulator, in fact it has to be said MicroEmu perhaps leads the way here when it comes to very rapidly changing screen sizes

Where both Android and iOS have succeeded in my view is that they have made the information on what designers have to work with is far more readabily accesible. A simple Google search for icon design guidelines for three popular platforms will tell you a lot, bearing in mind with the current trend of app stores the icon is the first piece of information that you are likely to convey to your user therefore it pays to have the perfect dimensions. For example try the following searches

“Blackberry icon sizes”

“iOS icon sizes”

“Android icon sizes”

“Nokia icon sizes”

Having to dig for this sort of information is what in the future will infuriate both developers and designers as new devices, screen sizes and user interactions are released into the market

Therefore until Android becomes as set in stone as HTML/CSS I don’t believe many designers have the time to teach themselves another XML based layout manager and need to work closely with developers and hopefully have some sort of ability to experiment with UI metaphors without a developers assistance. This is something which a WYSIWYG tool whilst most likely constrained to a single resolution at least provides. Once this point is reached, much like a sane server admin will never let a developer loose on production infrastructure without supervision, a designer will not let the developers near the UI without their guiding light

However until that day I concur with Roman a hand coded UI will be superior this is still true even with WYSIWIG tools such as Interface builder on iOS

http://www.spotify.com/en/mobile/overview/install-for-android/

Clearly this is aimed at the Finish owners of Android handsets where there is no Android Market.

The downsides of this ad hoc distribution is that an application could potentially be malicious and just alter the users dialler to re route calls through a premium number, (similar things happened back in the days of dialup with peoples machines getting compromised by malware)

However the pros for the consumer and third party developer are much like the internet and that seems to manage quite well despite the threat off malicious software getting installed on an end users device

iPhone Video

Android Video

One thing that strikes me as a little odd though is the amount of coverage this application is getting, if I want a DRM music solution for a mobile phone there is another option, Nokia comes with Music which wont cost £120 a year…..

Maybe the catalogue is much better on Spotify this is obviously enough of a draw, I have been pleased with Spotify and even paid for a few day passes, so would be tempted to purchase a subscription to utilize this on my G1 if it had a standard headphone socket (I seem to consume ear buds in my sleep) or if the iPhone if it could play in the background, as the times I am listening to music on my iPhone is usually on the train when I like to clear email etc.
Maybe this has all come down to clever marketing on the part of Spotify compared to a seemingly more complex solution even though it is easier Comes With Music solution

Sonys Walkman phones would be a seemingly obvious addition as well going back to one of my earlier posts!

http://docs.google.com/Presentation?id=dc2cgzjn_72grzmfqcv