This seems like it's going about solving mobile UI problems backwards.

I think a better approach would be to fill in the gaps between what a graphic designer is used to (Photoshop, Illustrator, Fireworks), and the UI tools that the different platforms offer - Interface Builder and the XML Layout editor for Android.

Interface Builder is definitely visual enough for designers, but most of the "cool" effects for modern iOS apps can't be done purely through Interface Builder.

The Android Layout Editor in Eclipse has a lot more control over features, but I wouldn't expect graphic designers to start creating drawables in XML either.

My thoughts would be to work within the Photoshop ecosystem - a plugin that exported XIBs and Layout XML would probably be the easiest way to go.

Check out Expression Blend.

Yeah, it's for Windows apps, but it's still pretty sweet. The designers at my old company could basically use it instead of photoshop and turned out some pretty impressive stuff.

http://www.microsoft.com/expression/products/Blend_Overview....

You have a good point. I also think that one should not completely replace IB, because its core functionality is very good: Design UI that is callable in code without naming the elements explicitly in the designer. What's missing are two features IMO:

- style templates (CSS or a subset of it) in order to increase reusability.

- access to preprocessed variables in order to define things like fonts and colors in one place that's also usable in code. (Maybe that's already possible but I'm not aware of it).

Some lightweight style plugin could solve both issues.

I agree. CSS Hat currently does this for CSS (http://csshat.com/), but it would truly be great there was a "CSS Hat for native mobile" UI's.

A 'CSS Hat for native mobile UI's" is basically what Pixate supports. That is, we can take CSS, and style native iOS (and Android) components just like you can web components. Pixate is the engine that enables all of this (whereas in a web page, the browser is doing all the rendering for you).

I think I like the CSS Hat model better because it generates the code that I then have complete control over. First, that can be helpful when learning the language. Second, it's really nice because of its tweakability. If the CSS Hat dev's decided to implement a stroke as a border, but I know it would look better as a layered drop shadow, then I can make that adjustment. If there are any other quirks or bugs, I can fix them with the resultant code. With the pixate model, I can't do that, as far as I understand (please correct me if I'm wrong!!).

You can edit the CSS with Pixate in the exact same way. It doesn't get 'compiled away', the runtime optimizations happen behind the scenes. The CSS can remain in your project as a resource, much in the same way you reference your CSS in an HTML page.

That tool is almost exactly the type of thing I was thinking of. Thanks for sharing!

I think good designers these days are used to HTML/CSS/JS - designing in Photoshop alone these days just doesn't cut it. This is why Pixate is so interesting.

$200,000 doesn't seem an awful lot for 3 software engineers and a graphic designer, including premises, utilities and IT.

I'm pretty sure they're currently in YC. I do wonder then what the campaign is for.

It seems like Kickstarter has been used more and more lately as an announcement platform that has the very nice benefit of generating a little extra income on the side.

For now it's going to work because a lot of backers don't actually realize that they have another (probably much larger) source of income to build the product with, and they feel like backing is the only way this product will get off the ground.

Appcelerator (creators of Titanium) supports Pixate by becoming first corporate sponsor.

http://developer.appcelerator.com/blog/2012/08/join-us-in-su...

I've been watching this with keen interest since it was first announced here sometime ago. Unfortunately, at the pace Pixate is raising funds on Kickstarter, it doesn't look like they'll achieve their goal. Does this mean they're screwed, or will they still push this product out the door through their corporate sponsor Appcelerator?

I was thinking about something like this the other day. Nimbus has something similar I think http://blog.jeffverkoeyen.com/nimbus-chameleon

I wrote this for iOSDevCamp last year: http://cssapply.github.com/

Photoshop is what the vast majority of all web and app designers use at some point in their process.

Great design has almost nothing to do with the tools used.

Seems like an irrelevant point. 89a didn't say "UIs designed in Photoshop are awful", after all. But a tool can be bad for a job, even if it is technically possible to do good work with that tool.

That said, I think that Photoshop does have a few core features that designers really like for web design, and then a ton of features that get in the way for that particular use case. So basing a web design tool on those features of PS that are good for web design seems like a reasonable strategy.

But it's powerful and that is what matters.

What would you suggest as an alternative?

I'm all about Fireworks personally. Love it's ease of altering my work on the fly.