What I don't get is why they keep going down the same path. Their Android application has evolved from a simple webview to many many webviews over the last year and they still haven't gone native. Maybe they're fine with the detriments when compared with the ease of creation?
I can totally relate with your comment. I've found myself converting from web to native as well. Mostly because webkit is still slow for UI operations like resizing elements that require some Javascript to be processed and so on.
The downside of this is that you start to have lots of subclassed objects just for a gradient or slide-up/down animation that you could do with 2 or 3 lines in js or css.
I wish someday simple view design could be done in html/css and leave the logic and core functionality to the native language instead.
As amazing as writing an application relying on a UIWebView for its core functionality, it still crashes when you scroll for long periods of time, and instacrashes when you attempt to look at your likes.
I am admittedly on v.3.0.x, but the version previous behaved the same way, so I have little hope for future versions.
I must admit, the Tumblr app is one of the few apps that make me want to throw my phone against a brick wall.
I'm the developer of Tumblr for iPhone and sorry to hear this. I know that web views have their drawbacks and we're working hard to make the app faster and more stable in every way.
While we have your ear: just wondering what the thought process behind the latest redesign is? Yes it's really~pretty...but the functionalities that made the older app attractive are mostly gone.
- cant long press to copy a tumblr link (heaven forbid I'd like to share a link OUTSIDE of the app
- cant post YouTube urls in video post
- reblogs won't let you delete prior comments by others (sometimes you have to clean up cluttered reblogs)
- you took away options to open in safari in many instances as well.
These were not intentional omissions. 3.0 was a complete rewrite and there are still many features that we know we need to add in order to be on par with the web experience.
I strongly encourage any users with feedback or feature requests to hit up my ask box: http://blog.bryanirace.com/ask. We really are listening and plan on providing the best mobile Tumblr experience we possibly can.
Edit: Worth noting that we added post sharing options in 3.1 (long press on the 'Like' button).
Eh, it's not the only thing I find wrong with Tumblr, and I'm yet to try out your latest release (stepped on debit card, broke it, cancelled it, iTunes charge failed, no upgrade for me)
Please try going native, the performance increase would be in the order of magnitudes.
I'm not the only one but it's a very small team at the moment. We have ambitious goals that we're aggressively working towards though.
Tumblr content can be quite diverse and often contains arbitrary user-inputted HTML, numerous inline images, etc. It's non-trivial to render but as I mentioned, we're working hard to provide an even better mobile experience.
Yup. The latest versions of the app has actually driven me away from tumblr. It is un-intuitive, clunky, and poorly thought out. I'm not the only one who feels this way too.
What I don't understand about Tumblr's iOS app is how they managed to make the dashboard slower, while making everything else native. It literally takes a minute to load the initial content (http://nuut.co/post/31000290048/new-tumblr-app-im-keeping-my...)
At least previous versions got the dashboard right.
AirBNB has recently released a library for infinite scroll lists which has great performance and lower memory usage: http://airbnb.github.com/infinity/
It's a great way to launch your MVP, but I've found myself slowly converting pieces to native due to performance issues and other minor quirks.
What the Tumblr folks have now is a great start, but I have a feeling they will make their way down a similar path over time.