I like the use of iFrames always a great way to think of things but easy to forget.
A very interesting although I don’t know when I would ever use it. But this may be a great approach applied to some other problem in the future, well see what happens.
This effect looks great, might be a fun thing to play with on an iPad or other touch device.
I find the problem of temporal combined with spacial area a very interesting area. I would think there is a hidden advantage of the spacial nature of geofencing, that being easy sharding. Areas can be sharded easily and then more complex time calculations can be done. The problems stated of a good background location system will eventually in my mind be replaced with issues of privacy and ability to opt in and out of different kinds of geofencing. I look forward to seeing a lot more development in this area.
I like the idea but I wonder if they are going to have trouble with the fact that they are hiding the ads to a service which is based on ads… but then again it might be a great buy for Google to build into a pay service for Gmail.
It looks interesting but it will defiantly mean we need an overhaul to the market if anyone can push out apps so easily.
Looking at the actual product (here) its a graphical programming language. Which at first reminded me of StarLogo TNG and after look more closely they used the programming block out of MIT same as StarLogo.
I’ve used this basic programming block and this is very approachable, so I can see a lot of new apps coming none of which will probably be 100% but it should be interesting to see what happens.
I wonder if you can start off using the blocks then switch to code. That work flow may make quick early iterations very easy.
A few people have asked recently about how rolling out new features is handled at Forrst. While the answer is nothing spectacular, it is something I think is hugely important to consider implementing when you’ve got a fairly active site. Flickr wrote about a few tricks they use — called feature…
Very nice approach the only thing that jumped out at me was was that in_array would be slow unless they were doing the flip trick. For small arrays I don’t think it would be a problem but keeping the keys as the users ids would probably be much more efficient.
The biggest appeal of this for me is the fact it has a modern browser on it. The more I can focus on web applications while getting to ignore IE6/7/8 (I have hope for 9) the better for me. I think it will be interesting to see what happens to the app store if enterprise does take off big for apple. I would think a special flavor would be needed to publish company internal apps…
Layering location with recommendation seems like a very exciting area. I wouldn’t think you would need persistent background operation to start offering recommendations either. Once a user checks into a location the remote server could start working on a recommendation then if one or two are found a push notification could be triggered. Thinking more about this I wouldn’t want to get that recommendation too quickly after checking in.
In my mind one flow could work like the following. I check in to a bar downtown about an hour into being there I get a push notification about a new place a block away that I would like. At that point I could choose to either dismiss the suggestion or head towards the new place.
But one area that is currently missing is I never give feedback on how much I actually enjoyed a venue i checked into. So this may end up giving odd results. Also I wouldn’t want a recommendation for a bar if I’m out for lunch.
To really tackle this type of recommendation I see the need for more time and sequence based measures used in the recommendations. Both time of year and time of day will effect what places I go. Also I could see sequence of events playing a big role, don’t recommend a restaurant if I was just at one 2 hours ago.
Looks like there is more to come I’m excited to see this area start to move even if it is just in the ether at this point.