One of the ways I am approaching the crowded iPhone Twitter client space
motionobj.com
motionobj.com
While I'm also doing some of that, I'd like to experiment in a different way, that of building new features. The conversation view, while it is not a totally new concept (it is available in a similar form in the Summize page, and the current Twitter search page), doesn't exist yet on the iPhone clients and it's pretty interesting how much you can do in the backend to churn out just a view of a conversation.
Digress a little: I've been working full-time on iPhone applications for the past 3 months but have only started selling a few weeks ago (a delay of 3 months due to the paperwork blackhole!). And it's been a huge challenge to break into the market and break even, but nevertheless an interesting experience.
As I mentioned in the article, I have a few ideas (for features) in my mind which I'll be experimenting with. If anyone of you have any ideas, please feel free to share them and/or discuss/comment on what I have done.
I am working on (and by "working on" I mean, I created some messy code 5 months ago and then abandoned it but cling to the dream that I may one day finish it) something similar, but wanted a web based UI - I have the same issue whether it is on a mobile client or at my desk. I want a twitter client that is smarter than a linear display of what's going on - and instead figures out the context of tweets (like replies and retweets) and intelligently gets me that context
It looks like you are trying to solve that issue with replies, kudos to that.
This particular functionality is actually a web service though. I built this running on a server and the iPhone client talks to the service. This provides some flexibility in terms of release schedules and most importantly works around the default Twitted rate limit. There is just no way a standalone Twitter thick client can do this without hitting the rate limit all the time, making the feature useless.
It's interesting you mentioned retweets though, I haven't thought about that. One thing I'm looking into is the timestamp, @names and some herestics based on commonalities between tweets. I come from a search engine background, so mining text and relationship is always interesting to me.
Why did you stop your work though?
Interesting. A web based application doesn't need delivery, but I wonder what kind of marketing you could get by leveraging the app store to market your product, but actually deploy it on the web. In other words, put a simple app on the app store that is merely a shortcut to spawning the web browser and taking you to your page. Could be free or paid, depending on how people respond to it. You could do more frequent updates too, without having to go through the app store, since it is web based. Do any apps do this currently?
But then if it has no app store presence, how would you charge the user? I might build a few more features around the web service approach to take advantage of the rapid deployment benefit you mention, but probably not as a full blown web app. I don't know of any app doing the latter, but I'm not surprised if there is.
Have you thought about virtual Twitter follower groups (e.g. groups of people you don't actually follow but still like to overhear and/or summize search terms actings as groups)?
Also, have you considered implementing visualizations like these: http://personalpages.manchester.ac.uk/staff/m.dodge/cybergeo...