Feature Requests--What do you Need?

Hi Bill,
Thanks for the question. Funny, I'm also tripping out right now on how 2 side by side devices are getting incredibly divergent location information on the built in Maps application.
Something might be happening over at Skyhook?

When it's working again, give us another try.
Nearby will bring up a handful of 'popular' places, stories, or photos around you (i guess it's a bug that a manually specified location doesn't come thru with the API call asking for places).

Where are you in Portland anyway? I wonder if the edge network isn't active in some pockets? Are you on a 3G or other device?

I'm on a 3G device over at CubeSpace. I've noticed that this device gets low to no bars most of the time here and as of right now no 3G. It's a fact that my BlackBerry got much better reception.

I'll fire up Nearby when I go outside and I'll let you know.

Mkay. So I manually turned off 3G (in general prefs) and turned the iPhone off/on. Now both Google Maps and Nearby can locate me (indoors no less). Wow Nearby is so much more obvious once location services are working.

phew, I thought maybe we had some incredibly terrible bug :D

but seriously, we could use some better handling when certain things are not available, location data, network connection, etc.

Thanks for retesting and reposting the answer!!

how change the distances from meters to miles.

Thanks.

Hey, it'd be good to have a pub category, right now I've been using the food one which is not always appropriate... That and the clunkiness of the map please.
I really do like this app as I do a lot of travelling and ilike to explore places thoroughly; it lets me share places for other people to find. Thanks!

Also I'd like to be able to add photos to posts that have already been made please!

oh and one more thing :P
Could you have filters on the map, so you could select food and it'd only show the food categories. Things start to get awfully cluttered with everything showing up at once.
THANKS!

Just started using this and it's not bad! Is it possible to add "I've been there" to a post on iphone. Also, it's annoying to be taken back to my exact location when I want to add a marker from my iPhone especially as it's a bit slow loading the map

Hey! it's me again! In my area and there about there are posts for gigs and things that are years out of date. Can there be something in the filters to dismiss ye olde things like that so they don't keep popping up likes?

Tracklogs would be nice. I would have expected to see a simple tracklog app pop up in the App Store by now. I'd guess that certain limitations of the device (e.g. no background apps, etc.) make something like this less appealing. One solution might to tack the feature on to another app like Nearby which can serve as the one app to have open when out and about with the device. I'm thinking of a couple of possibly complementary features here:

1. A traditional tracklog, i.e. an ordered sequence of coordinates along with timestamps stored in a useful (preferably standard) format. Such data would be helpful for geotagging photos and as the basis for new places.


2. The ability to mark places with no details by just tapping a 'mark place' button, which records the date and time and the coordinates of the current location (the ability to add a description immediately would be nice but isn't necessary). It should be possible to access/update these places later from a PC, or the mobile device. Something like this allows for the accuracy and convenience of location awareness. But less fumbling with the iPhone while out (presumably doing other things).

Place history: Everytime I visit a place record the date and time associated with the place. Display a simple visited count attached to the place (unique to each user). Clicking the number or some other interface element would allow me to see the place history. The ability to add descriptions next to each visit would be really nice, but is not strictly necessary.

This could be the basis for a number of other features. For example, the counts could be anonymously aggregated (opt-in/opt-out?) to track th most popular places. A list of the most popular places in a city organized by category is the start of a compelling city guide. Combine this with ratings and we can begin to ask interesting questions about all of these places.

Where do I go the most/the least in my home town? What did I do the last time I visited another area? What's the most visited restaurant in the area? Is the most visited restaurant also the highest rated (Why or why not)?

Itineraries (User submitted): This has me thinking about the limitations of typical travel guides. Let's say I'm traveling to a new place. I pick up a travel guide to plan what I want to do while I'm visiting. What's wrong with this. I would say that there is a lot wrong with travel guides. First they list only a small number of places. Secondly, they are quickly out of date. But that's just the beginning. Actually making a plan requires a lot of work on my part, as someone unfamiliar with the area, because I know what I want to do but not how I should go about doing it. For example, In Boston where I live, Fenway Park is a short walk through the Fens (part of the Emerald Necklace parks) from the MFA. These are two of the more popular attractions in the city (three if you count the park) but most people don't think of them near each other. There must be a better way.

Let users stitch connect places with paths to create itineraries which:

1. help with planning little excursions through an area


2. help us think about the relationships between the places in our lives


3. help other users find their way around.

I know I've been posting bits of information that I think others will find useful such as free parking and interesting places that are a bit off the beaten track.

With the place history idea, I'd be interested in how many people visit a place - yes you'd be able to see what's popular. If I could record the number of times I visited a place it could really skew the numbers - somewhere might have a high count from one person but not be that interesting/popular (if I understand Rob correctly)

I'm thinking of a 'place history' working on two levels.

1. Yes keep a count of the total number of times a user visits a place. This history is available to that user, and is unique for all users (I can access my own place history).


2. Pulling and aggregating just the counts as one measure of popularity. DalekKiller I think you're right that there should be some conversion that takes place when aggregating these values. Such a conversion may or may not discount the number of times an individual user visits any one place. It's certainly something to think about!

In regards to #2, there would be at least a couple of issues to think about:

Though it may be reasonable to count my visits to a place literally, (which may number in the thousands). It may be less good to treat potentially thousands of visits by thousands of users literally. I suppose it's just a matter of whether or not we want to potentially see some big numbers. I don't believe that converting that number to a more abstract value would be difficult.

In response to DalekKiller more specifically...

> If I could record the number of times I visited a place it could really skew the numbers - somewhere might have a high count from one person but not be that interesting/popular...

True. You could of course do something simple like count each user who visits a place just once, regardless of the number of times they visit. These are just two options (count every visit for every user, or count every user at most once). There are many other ways to handle this. With some thought it might be possible to find the right recipe (i.e. a happy level of individual influence).

These are great ideas, I'm especially into the ideas that work with the low accuracy issues of triangulation, where that issue isn't seen as a 'problem', but is seen as 'totally fine'

so, since the app is making an api call each time you launch it already,
it would be simple to just translate those queries into some special sort of 'draft' place inside your platial account.
you wouldn't even have to hit a button, it could just happen every time you launch the app. or is that too aggressive? would you prefer to have control over it?

our buddies here in portland are doing an art project along the 'loose tracklogs' thread, but it just runs a cronjob in the background, uploading your location to a server every ten minutes. obviously the official sdk doesn't allow for background stuff, but the results are quite nice: http://hook.org/whereis is a combination of a few participants' tracks

In response to Jason's post:

> so, since the app is making an api call each time you launch it already,...

I'm not sure that I'm following your description.

Right now, when I launch Nearby and I might briefly see a world map before it begins to fix on my location and zoom the map. At that point I can do a number of things (navigate the map, search, adjust my settings, etc) including add a new marker.

If I want to add a marker, I click the plus button in the upper right hand corner, which puts an icon for a new marker at the center of the map (also my current location if I haven't moved the map). At this point I can move the map to adjust the position of the marker (by the way I find that this is a much better way to handle manually positioning than moving the marker itself the way that the native Maps app works), and then I must either pick an existing place or add a new place before I can save. Either way I must at least enter a value for the Title field before I can save.

Let's pretend that instead I could choose to save the marker for the place without picking the Place, Title, Category, or Description.

Maybe these markers, let's call them 'draft-markers' are treated differently than places. For example, they might be assigned a special icon and not be included on the map normally. At some point later on I could work through the draft-markers adding the details and then saving the marker as a full-fledged place.

Why do I think this would be useful?

Essentially we're unbinding two tasks:

1. the job of locating a place, which is easiest when we're out and can use the location services to do most of the work for us.


2. the job of describing a place, which might be much easier/more convenient to do at another time, maybe at a computer for example (with a full-size keyboard, faster network connection, and without worrying about the limited battery life on the iPhone).

If I don't want to enter any information about a marker at all then it might make sense to skip the Add Marker screen altogether. In this case it should be possible to 'save as draft' directly from the map after positioning the icon.

Ideally, I would have the option of saving none, some, or all of the place information associated with a draft-marker and still have it treated as a draft, so that I would have the option of doing whatever makes the most sense at the time.

For example, If I'm in an area I know very well then it's probably enough just to mark the location as efficiently as possible without any details. Seeing the marker on the map later will be enough to remind of the title, description and all of the rest of the information I might want to add to the place. The ability to quickly mark the map without a lot of fuss while I'm out would allow me to better incorporate Nearby into my normal routine, which means more and better map data.

On the other hand, If I'm someplace unfamiliar, I may want to enter at least a title or description to provide me with enough of a clue to allow me to fill out the place completely later on.

I've tended to just do the basics when I'm out e.g. mark the place, give it a name, take a photo and then do the fine tuning when I'm at my laptop. This method seems to be working quite well for me.

I'm using my iphone out and about more so it would be good to be able to add tags via the iphone

In response to Jason's comment:

> Rob, exactly! The main question, and this could be handled as a user pref,...

In my opinion, it probably wouldn't make sense to create a new marker every time you open the app. That would seem to imply that the only reason to start Nearby is to create some sort of marker (either a place or a draft). But as a user, I'm just as likely to be starting Nearby to look at places that already exist, or other content.

After we launch the app and we come to the map, we already have the + button and there's no reason to distinguish between a draft marker and a completed place at this point. Either way, we're adding something. Clicking the button adds an icon to the map and brings up the little panel at the bottom. If there were a 'save as draft' option on that panel then it would still be possible to adjust the position of the icon first, and then you're done - back to the map and ready to add a new marker.

Otherwise, if I chose to go ahead and bring up the 'Add Marker' screen I would be able add as much of place, title, category, description and photo as I wanted to. If we added a 'save as draft' button to that interface as well, then I would have the option of adding any of that information to my draft. On the other hand, if I entered none of that information the effect would be the same as if I had clicked 'save as draft' without going to the 'add marker' screen.

I think that should do it. Obviously there are other ways that you could go but this seems to require a minimum amount of change in the interface, takes advantage of what you already have, and, importantly, doesn't take away any options prematurely. For example, we can still adjust the icon before saving a draft, but it means that we can create a draft, and get out of that task as quickly as possible.

It's two actions on the part of the user. First clicking the + button and then 'save as draft' We also are using the iPhone to its best advantage - directly positioning the icon on the map and avoiding the data entry.

If the app did automatically create a marker whenever it was launched, potentially we could save a step, but I'd be a little concerned that it might be confusing and a little unexpected for new users and maybe cumbersome to have to dismiss the marker somehow if we weren't trying to create a marker.

By the way, I've been referring to these things as draft-markers or drafts but we can probably do better than that. Without giving it too much thought 'placeholder' might work, or simply making clear the distinction between a 'marker', which is the unpublished indication of a place, and a 'place', with is a completed marker published to the map. Of course, there may be better labels too.

Events: As stand alone markers or attached to a place. Either way an event should be clearly intended on the map, e.g. represented by a unique icon. The way I'm thinking about them, events behave like other places except that they are visible when they occur within some specified date range, or at a given time (e.g. now). (This is probably a slight oversimplification but I do think a simple implementation can work fantastically well).

The significance of a place may be different at different times. It makes sense to acknowledge and represent this. For example, an open area maybe a farmers market, a sports venue, an outdoor theater, or an empty space depending on the time.

Am I in a shopping if all of the stores are closed? Maybe.

In my opinion this is an important aspect of 'place', and one that could have a profound impact on Nearby/Platial.

Glad to see the bulk upload tool. Now how about a way to export place data from Platial.

It's something that I always look for with any service like this. Do I have the ability to take the time and effort spent adding content and repurpose it or recover that data so that I can move it to another service (e.g. if the terms of use change such that I no longer feel comfortable using Platial).

The same csv format you require for bulk uploading would be fine. JSON or some other format would be nice too.

Thanks for all of your efforts.

Placetracking/Geotagging:

This is actually related to the itineraries feature/idea I've already suggested. It's actually a different interpretation of what is basically the same idea - but expanded so that it can be useful for geotagging photos. I'll leave it to you to read the original itineraries feature suggestion to save space here.

Essentially, what I'm suggesting is that it should be possible to link places to form what is sort of a route or path. What I suggested as part of the itineraries idea is that this would be useful for trip-planning or as a sort of tourist/visitor aid. Rather than just talking about these places as if they're discrete things, let's acknowledge that the relationships between places are also important; for example, the distances between them. Typically we organize places into categories, which is nice for when we want to, for example, consider which is the best pizza place in a city, but it's limited in a lot of other ways. For example, if I'm planning a night out or if people are coming to town and I want to plan a day's worth of activities, I only want one pizza place, not all of them; but I also want to do other things. What's important is that I can string together what I want to do in some way that makes sense, so distance is an important factor here. Imagine that I can find a map that lists not only things to do in Boston but is composed of a number of suggested itineraries that I can choose from. Maybe breakfast at a nice local cafe followed by a trip to a museum in the morning when it's less crowded, lunch in the park, shopping in the afternoon in some area close by, and then dinner, a show, and drinks. Not only are we considering place, of course I want to go to good places - but also distance, which doesn't have to be a mileage count just closeness and time. (See my idea about events.)

So that's the old idea. What's the new idea? Well, I've also suggested that it should be possible to easily create markers without completely specifying them. I've referred to these as either draft markers, drafts, markers, or place holders. If we are working toward that feature, I think it should be possible to create one of these draft markers with just one hand. Have the iPhone locate you, click the '+' button to add the icon, maybe fine-tune the location with your thumb, and then save. So let's say that we can do this. Placetracking is the idea of saving these markers as you make your way through some journey. Maybe, for example, walking around a city for a day. It should be possible to establish a route by connecting these places, with the time used to establish sequence. So maybe we could specify that we were starting a route and then all of the markers we created subsequently would be considered part of that route until we indicated that we were done. Or, after the fact you could select a group of markers and tell it to connect them to create a route. That is what I'm calling placetracking. It's nice because as you can see it's nothing but a combination of the other features we've been talking about and we can do a relatively sophisticated thing. The best sorts of features are those that can be combined to do new things with them with little additional effort.

How does geotagging fit in? I geotag a lot of photos, and it takes a lot of time, though I've written a script to automate much of the task. You can't automate the work involved in locating the photos. The best general-purpose solution I've found is setting the location in Google Earth, pulling that location and automatically applying it to a group of photos (by selection), and then everything else can be automated; building a stand-alone gallery or uploading to Flickr, populating a Google Map with placemarks for each photo, even building in the logic to group together photos and represent them by a single placemark which is linked to a gallery corresponding to just those grouped photos. The gallery can regenerate dynamically as we zoom in on the map and create separation between the placemarks down to the level where we can distinguish between separate photos. All of this stuff can be done automatically except the geotagging.

How do we automate that (geotagging)? Well we're promised hardware that will do it for us, integrated into some future generation of cameras. And today there are stand-alone devices which work by periodically recording position along with a timestamp so that after the fact we can automatically compare the creation date on the photo (from the metadata) to the log and then apply the appropriate coordinate data from the log to the photo.

What's the problem? Essentially the problem is that GPS is often unavailable. It doesn't work at all indoors. It sometimes doesn't work in cities where there are a lot of obstructions, etc. It's unreliable. The little devices that are available now have this problem, plus they operate on batteries which typically don't last even an entire day. So after going out for a full day, you come back to find half your day is not recorded. Or you forget to change the batteries before you start the next day and don't get any data at all.

For similar reasons, I don't know when we'll see the capabilities in cameras. If I were a manufacturer, it's something I'd love to be able to advertise, but I'd hate to field the support calls from angry customers who insist the device is broken because it geotags photos successfully less than half the time. Not to mention the negative press about the diminished battery life.

So how does any of this relate to Platial? We've already talked about placetracking. I also have my digital camera with me and I'm taking pictures along the route. Let's pretend that I can select some of those photos, put them in a folder, and compare the time the photos were taken to the markers that I've created in Nearby and then apply the coordinate data to the photos whenever the date and time on the photo falls within some user-specified tolerance of the time of the marker (say 15 seconds, 2 minutes, ...).

Why is this better? Well we're still letting the hardware and software do most of the work involved in locating us. That's how we're creating our markers. We don't start with the world map and have to manually key in a location or place name, then manipulate the map to get to where we are. Also, we get the cell and wifi triangulation when GPS fails. But we're also fine-tuning the location of each marker as we go rather than saving up all that work for the end. What takes a very short amount of time for each marker may add up to hours' worth of work if we do them all at once; especially because we're out of the moment and out of the place so we have no assistance. Am I suggesting that we only geotag photos that coincide with the markers? No, I think close enough is extremely valuable here, because if we could geotag the photos with an approximate location, then we could cut down on the time it takes to fix them later.

My Google Earth solution would take a fraction of the time if I could somehow open the map to the approximate location of the photo so start with and make a small adjustment, save the change and be done. Also, for a good number of the pictures, the location is going to be accurate enough. And this would allow you to geotag photos along let's say a highway, which is nearly impossible to do after the fact with no help because you never know quite where you are on a highway from looking at a map and trying to compare the photo (this is a big deal). Also imagine going to visit a national park or something similar. Afterwards people want to know where a particular picture was taken, but the photo doesn't help them. It's a picture of a hippopotamus (I realize hippopotami aren't native to the U.S. :) with no indication of place.

What would be required? The features we've already talked about, and some capability to identify the photos to compare and then apply the metadata back to the photos. That's a simple client app to write or even works well as a web service because we don't have to actually transmit the photos, just some small portion of the metadata.

The additional benefit to Platial is that we get better photos and more of them, because we can apply photos taken with actual cameras not the iPhone's camera, to our places.

So in summary the two ideas I'm talking about are geotagging photos as described and placetracking.

Thanks.

Set location from coordinate metadata in photo: I've already suggested Placetracking/Geotagging under this topic. What I'm proposing now is the complementary idea of having Platial/Nearby examine the a photo added to a place for a geotag and then set the location of the place based on that data.

I can imagine a number of different ways this might work:

1. Immediately after uploading a photo, the app would automatically check for a geotag and, if one is found, notify the user of the presence of the data and prompt her to reset the location of the photo based on the metadata. This could be refined by prompting only in the case that the location of the photo differs (by more than some tolerance) from the current location of the place


2. Alternatively, you might put a 'check for geotag' button on the 'Add Marker' screen. After adding a photo to a place the user would have the option of checking the photo to set the location of the place.


3. The app might also offer the option of adding a place from a location in a photo earlier in the process.
   
   Currently a user clicks the + button and sets the location before coming to the 'Add Marker' screen. You could add a second button in addition to the current + (Maybe the icon could depict a camera along with a '+').
   
   As I'm thinking of it, clicking this button would have the user select a photo first, which would set the location automatically and then bring the user to the same 'Add Marker' screen.
   
   Alternatively the 'Move map to fine tune position' panel could could allow for setting the location by selecting a geotagged photo.
   

In my opinion this would significantly enhance Nearby. Of course the same sort of could be implemented in Platial. With Nearby, better integration with the bundled apps/device functionality promises to make Nearby a more compelling app.

Would like to print certain maps poster size to laminate and place on wall. Example: Pasadena, CA showing locations of houses designed by architects Greene & Greene. Take a look at http://tinyurl.com/5p2s82
or http://platial.com/map/Greene-Greene-....

Desire to zoom in around 75% then save huge image for printing.

Thanks, Tom

In reply to TomSoCal:

> Would like to print certain maps poster size to laminate and place on wall.

If I understand what you want to do, and I'm pretty sure I do, Platial will probably not be able to help you. It's Google's position that this use violates their terms of service. There was a pretty widely distributed python script that simplified the process of merging map tiles into one larger image suitable for printing. Google requested that the author take down the script.

Search for 'gMerge takedown notice' for more info. You might also be interested in taking a look at http://web.media.mit.edu/~nvawter/pro...

In fairness to Google, I'm sure this violates the terms under which they license their map data from their provider.

Anything form of compositing, enlarging, distributing those maps is not going to be allowable. It's too bad really. The resulting images can be quite impressive and useful.

1. I would like to see more flexibility in regards to linking a device and a map. The current process, which (for anyone who doesn't know) involves specifying a single map on the /account page while logged into platial.com, is clumsy.

Ideally I'd like to be able to choose a map for each marker separately.

Being able to set this preference through Nearby (or alternatively a version of the platial account page optimized for use on the iPhone) would be a big step in the right direction.

2, 3. The ability to browse among maps for the associated Platial account would be good. Also, the ability to search through (match) against the complete list of map names (all maps) would be super. Again, if not directly in the Nearby app then some version of platial.com optimized for the iPhone.

4. In fact it would be nice to be able to browse and search maps in Platial. I don't think that's possible currently. Am I wrong about that?

I know searching works across maps, but what if I want to restrict a search to just targeting the maps themselves, and not the place names on the maps. Why? I may want to search for maps with names containing the string 'cafe' because even a simple search like that may return bunches of cafes in the area vetted by someone who cares about that sort of thing. The results of a search for the string 'cafe' among place names may not be as useful. That's a very simple example, but hopefully you understand what I mean. Again, this may already be possible. If so, good job and thanks.

Hi,

I would love to see the option of adding locations from my address book - just locally. As it would make it a lot easier to decide where to go first and such.

And I do not see why you choose to remove my location, after hitting the target to locate me? I mean, at least I would like to keep my location on screen (similar to the map app that has a aqua pearl).

Regards,
Pan

I'd like to be able to view my Platial maps from Nearby. It's not obvious / working / present on the Nearby app version (latest from AppStore) I have installed on my iPhone.

Once I've selected one my maps then any places I add should be added to that map. I haven't tried this but I assume that when I add a place from my iPhone at the moment I can then add it to the map through the Platial website interface later; but this is laborious if I'm going to be moving around a lot marking many places in one go (I'm building a map of inline skating places around the UK).