GeoStrings (Location-based reminders)

[DanPLC]

I've added 0.1, 0.25, 0.5, and 0.75 miles as options for target radius. The sub-1 mile selections will serve 2 purposes:

1) People walking and using other non-motorized transports

2) People wanting to receive reminders very close to their destinations, especially if they're specifically heading towards those destinations. For example, if you were at work and wanted to remember to do something when you get home you would set the target radius to 0.1.

[FischOderAal]

This App looks awesome. I skimmed through the thread but couldn't find an answer to my question: how does it look up the locations of the stores? Would the app work all around the work?

I'm a german Pre-User (GSM of course), fyi. As well as the fellas in Canada we only get the free stuff and only those apps that have worldwide ticked (some forget). I barely look into the official catalogue, preWare owns it hands down.

[DanPLC]

Originally Posted by FischOderAal

This App looks awesome. I skimmed through the thread but couldn't find an answer to my question: how does it look up the locations of the stores? Would the app work all around the work?

I'm a german Pre-User (GSM of course), fyi. As well as the fellas in Canada we only get the free stuff and only those apps that have worldwide ticked (some forget). I barely look into the official catalogue, preWare owns it hands down.

It uses Google to search for the addresses. It seems to work with businesses in Germany. I plugged in "McDonalds" in Berlin (I set the "State" to be "Germany") and it came back with a bunch of results. So I'll either rename the "State" field to be "State/Country" or add a new optional "Country" field. So I'm hoping it'll be compatible with international users. However, it will be a paid app, so it won't show up in your catalog until they offer paid apps there.

[FischOderAal]

Thought so :/ But thanks for the info! Guess I'll have to wait.

[skidmark]

I just jumped into this thread, and was wondering if GPS will be the only way for this program to work or will it run with "location services?" Just from the standpoint of battery conservation. I do realize it may not be as accurate. Or if that can be set as an option?

Thanks for writing such a useful app. Keep up the good/hard work.

[DanPLC]

Originally Posted by skidmark

I just jumped into this thread, and was wondering if GPS will be the only way for this program to work or will it run with "location services?" Just from the standpoint of battery conservation. I do realize it may not be as accurate. Or if that can be set as an option?

Thanks for writing such a useful app. Keep up the good/hard work.

I'm implementing logic to help with battery consumption. That's what I'm currently working on. But as I've been using the app, it doesn't seem that bad even without this logic in place. I'll implement the logic and then run some more extensive tests.

When asking for the current location, I call the webOS "GPS" service which does whatever it has to do to get the current location. I can indicate the desired accuracy of the result; however, I don't think that guarantees that it will use GPS vs. Cell Tower Triangulation or some other method.

[nodforce]

Originally Posted by DanPLC

I've added 0.1, 0.25, 0.5, and 0.75 miles as options for target radius. The sub-1 mile selections will serve 2 purposes:

1) People walking and using other non-motorized transports

2) People wanting to receive reminders very close to their destinations, especially if they're specifically heading towards those destinations. For example, if you were at work and wanted to remember to do something when you get home you would set the target radius to 0.1.

Thank you! I only had 2 issues that I was concerned about with this app, and now 1 of them is gone. Even though I'm not in a city, I live on a small, cramped college campus. My daily life basically exists in about a 2 mile radius. So this was not only wanted, but necessary! The only other issue I have is my concerns over the battery life, but from what I hear, it looks to be very promising. Good job and keep up the good work!

[RSKadish]

That's great! Can't wait to buy the app!

[Aviator]

This is going to be good. I would certainly try and buy.
Just curious, why are you not releasing it? At least in Preware. So you can get a bunch of feedback? And "Donations".

[DanPLC]

Originally Posted by Aviator

This is going to be good. I would certainly try and buy.
Just curious, why are you not releasing it? At least in Preware. So you can get a bunch of feedback? And "Donations".

Music Player (Remix) is the most downloaded app in the homebrew gallery. I would guess it has one of the biggest user-bases as well. While I do get some donations for Music Player (Remix), compared to the number of people actually using the app, it's a very small percentage. So no, I wouldn't use the prospect of donations as a motivator to release it to the homebrew gallery.

But you're correct, the main advantage of releasing it there is you get a bunch of great feedback and a whole slew of testers. The disadvantage is that you're offering your app for free (including your source code) and all those people that love your app may decide they're content with staying with the free beta instead of purchasing your app in the catalog.

So in the end I decided to have a closed beta program, which I hope to start very soon. Before I release it to them, I want the app to be basically "catalog ready". It'll be the job of the beta testers to point out any outstanding bugs which I didn't find, and also to point out any issues with the user interface or functionality which I may not have considered.

[OneDeep]

Dan your right but you could have a limited version of the app but that would take you 3 steps back instead of forward. And plus you want to get paid and I wouldn't blame you bro.

[DanPLC]

Originally Posted by OneDeep

Dan your right but you could have a limited version of the app but that would take you 3 steps back instead of forward. And plus you want to get paid and I wouldn't blame you bro.

Yeah I wouldn't want to do that because the whole point of releasing it to homebrew would be to have people test the full "catalog ready" version. If I create a lite version, that doesn't help me test the full version.

So I think the closed beta is the right way to go. And I don't expect the beta phase to last long because like I mentioned the version they'll be getting will be pretty much the final version (feature-wise).

My handler from Palm has contacted me in regards to the app, so hopefully once the beta testing is complete, the app shouldn't take too long to appear in the catalog.

[bezoris]

Originally Posted by DanPLC

Can you give me an example of a use for this type of feature? I just can't think of one off the top of my head. (RE: reminders when leaving a location)

I could see these types of departure, or outbound, strings being fairly useful, maybe set by default to lowest possible radius.

For example: Don't forget to power down the server before leaving work, Don't forget to check next week's class schedule, or Don't forget to check on the patient in Room 666, that sort of thing.

Of course, while this functionality could be somewhat replicated using time filtering (for example, I normally work 9-5pm, fire the Server Reminder at 4:45) this wouldn't work for flexible schedules. What if I get tied up until 10:00 pm? By that point I've forgotten all about the 5:00 pm reminder.

Food for thought. The app's looking great!

Best, bez.

[DanPLC]

Originally Posted by bezoris

I could see these types of departure, or outbound, strings being fairly useful, maybe set by default to lowest possible radius.

For example: Don't forget to power down the server before leaving work, Don't forget to check next week's class schedule, or Don't forget to check on the patient in Room 666, that sort of thing.

Of course, while this functionality could be somewhat replicated using time filtering (for example, I normally work 9-5pm, fire the Server Reminder at 4:45) this wouldn't work for flexible schedules. What if I get tied up until 10:00 pm? By that point I've forgotten all about the 5:00 pm reminder.

Food for thought. The app's looking great!

Best, bez.

I can see the usefulness of this type of reminder now that I see some examples; however by the time you'd get the reminder (even at a radius of 0.1 miles), you'd already be out of the building and on the road. Of course you wouldn't be that far away, so you could go back. But I just don't like the idea of GeoString reminders coming up when it's too late. The intent of the app is to remind you before it's too late.

And as you mentioned you can use the time filtering option to mimic this type of functionality. But you're correct that if a person has a more flexible work schedule, this would not be effective.

So I'm not sure what the best solution to this one is, or if there is one at all.

[mobilemitch]

Google map data isn't perfect and searching for businesses or addresses, especially when it is new stuff, like a new subdivision. (just no way to search for those very often.) I hope there is a way we can enter lat-long.

On that point, unless you are at the location currently, finding the lat-long on the Pre is time consuming. (tips anyone?)
Earlier you said that being able to use a 'touch point' on the google map made you use the google ajax code and that it made the app take 6-8 seconds longer to start up. I think that rather than eliminating that option, it would be a good thing to keep this option in a preferences dialog. Perhaps with a note that says, "adds 6-7 seconds to startup time"

[bezoris]

I can see the usefulness of this type of reminder now that I see some examples; however by the time you'd get the reminder (even at a radius of 0.1 miles), you'd already be out of the building and on the road. Of course you wouldn't be that far away, so you could go back. But I just don't like the idea of GeoString reminders coming up when it's too late. The intent of the app is to remind you before it's too late.

And as you mentioned you can use the time filtering option to mimic this type of functionality. But you're correct that if a person has a more flexible work schedule, this would not be effective.

So I'm not sure what the best solution to this one is, or if there is one at all.

I certainly see your concern and understand not wanting to be potentially peppered with criticism [W_T_F dood!? I wuz already 1mlie awy frm my office!] but I would view this as an enhancement with certain acknowledged limitations.

Still, I think most people would rather be reminded while still in the area of, say, work than when they arrive home an hour later (Home:reminder: Did you remember to power down the server?)... Ooops!

This obviously is not a top work item, but I do think outbound strings would eventually add to the usefulness of the app.

Best.

[DanPLC]

Originally Posted by bezoris

Still, I think most people would rather be reminded while still in the area of, say, work than when they arrive home an hour later (Home:reminder: Did you remember to power down the server?)... Ooops!

That's a good point. I'll take a look at this and think about adding this option.

[DanPLC]

Originally Posted by mobilemitch

Google map data isn't perfect and searching for businesses or addresses, especially when it is new stuff, like a new subdivision. (just no way to search for those very often.) I hope there is a way we can enter lat-long.

On that point, unless you are at the location currently, finding the lat-long on the Pre is time consuming. (tips anyone?)
Earlier you said that being able to use a 'touch point' on the google map made you use the google ajax code and that it made the app take 6-8 seconds longer to start up. I think that rather than eliminating that option, it would be a good thing to keep this option in a preferences dialog. Perhaps with a note that says, "adds 6-7 seconds to startup time"

In particular I added the Google Ajax API so I could map all the locations in a GeoString on a single map, or provide a map icon on the "Active GeoStrings" list so they could view all their tasks on a single map. Unfortunately you can't add multiple markers using the built-in Google Maps app.

And as you mentioned, the other possibility would be adding a method of selecting a target location via a map input.

In addition to the slow load time, I found that sometimes the Google API could not be loaded and thus all calls to it failed. I don't know if this is an issue with Google's service or webOS, but in either case, it's unacceptable to the user that some of the app's documented functionality would not work. So no, I'm not going to include this as an option. I may re-explore it in the future to see if maybe a more current release of webOS fixed the issues.

In regards to being able to select latitude/longitude coordinates, I'm not sure most people would know how to do that. And it would require them using a separate application to find the coordinates to enter in which seems clunky. One thing I found is that on the "Address Input" screen, it allows you to enter in an intersection, so that functionality could be used if the specific address is not known.

[DanPLC]

For those of you who were asking for a way to launch Sprint Navigation from the alert popup, I've come up with a pretty good alternative.

Each GeoString on the alert popup now has a map icon next to it. When tapped, Google Maps is loaded with the current and target locations. The user can tap the "Get directions" button to display step-by-step directions to the location. The route can also be displayed as a map.

I'll probably post a demonstration video tonight.

[someonewhobikes]

Originally Posted by DanPLC

Yeah I wouldn't want to do that because the whole point of releasing it to homebrew would be to have people test the full "catalog ready" version. If I create a lite version, that doesn't help me test the full version.

So I think the closed beta is the right way to go. And I don't expect the beta phase to last long because like I mentioned the version they'll be getting will be pretty much the final version (feature-wise).

My handler from Palm has contacted me in regards to the app, so hopefully once the beta testing is complete, the app shouldn't take too long to appear in the catalog.

I've seen a number of homebrew apps that are full-function but time-limited; either by # of days or at a particular date. Note that I only point this out because I'm not eligible as a Beta-tester (I don't use my Pre for music) and want to get my hands on this ASAP.