Hacker News new | comments | show | ask | jobs | submit login
Show HN: Gecko SMS, our (hopefully) revolutionary Android messaging app (android.com)
17 points by asmosoinio on June 10, 2011 | hide | past | web | favorite | 28 comments

Best of luck with your app, but I'd like to take a second to play devil's advocate: what does your app offer me over my existing location sending apps (Latitude and Glympse)? Latitude being there to let me friends know where I am without me even having to initiate anything (for when I don't care about privacy at the moment), and Glympse lets me send temporary expiring location messages via SMS, Email, etc (for when privacy of my location is a concern).

No need to apologize, this is exactly why I posted the app and the idea to HN!

I agree Latitude rocks hard (when you get friends to actually use it), and the idea of Glympse is brilliant. I haven't tried it myself but immediately thought it would be a good solution when I stumbled across it. We may want to implement something similar as well.

But they both fail on a few cases:

- Try getting your non techie friend with a dumbphone to use Latitude -- not possible as the app can't run on the device. Same with Glympse, however less so because it may work ok with pretty simple browsers on the receiving end.

- vs Gecko SMS: If someone can receive SMS, they can receive your location

- And then the "rest of the world" -argument: most people don't understand maps, maps are not available to many areas. Maps don't work well in cheap phones with small screens.

Landmarks is a simple concept that everybody understands, and it's available everywhere. That's what we offer.

I know this partly contradicts with using the Android platform, but that's where we wanted to start. We have a fully working prototype for S60 Symbian as well, but it's no fun to use. :)

Thanks for the quick reply! I guess I missed the part where Gecko would work with so-called "dumb" phones as well, without any mapping software. That's definitely a good enough reason for me to give it a try this weekend.

Hey, I just submitted something of my own for the first time so what else can I do than sit here replying?-)

I don't think you missed it, that information probably isn't available anywhere yet. Unless you read all my comments here, and still I don't think I have really mentioned that.

But had you read our business plan, you would definitely be aware of this.

-- Update: Just realized you probably meant the receiving end, whereas I was talking about sending these messages from a dumb phone. The former is mentioned in the Market text:

This means that Gecko SMS works with standard SMS and MMS, and does not require any special software on the receiving phone.

This is an Android application I have made for my company.

It is a Messaging replacement with a twist: You can include your location in the message, while still using standard SMS or MMS without any special requirements on the receiving end.

To make a location human-readable, the app uses a global landmark database that we have developed and are constantly updating. Our database was originally developed for communicating tracking device location over SMS, and we are now looking for other ways to utilize it.

I am somewhat ashamed of the current state of the application, so this should fit the "release early" philosophy.

We have tons of further development ideas ourselves, but would love to get some feedback from real users to figure out what to do next.

Any comments are highly appreciated (on the app, the Market copy, the look&feel etc.).

On your website you say "Our landmark database is the most accurate in terms of quality, coverage and updates ". That's a pretty big claim. What's so special about it? Where does the data come from?

As for the app, I don't know, maybe I'm missing something but doesn't sounds too exciting. In all cases when I would like to put a "landmark" in my message I would do it myself.

We select each landmark "by hand", and can spend time on the quality of the data while still keeping the data gathering/updating process effective. Quite different compared to how traditional landmark databases, with focus on quantity, are built.

Another one of our strong points is data in areas where there are no digital maps available (but still lots of people with cellphones) -- landmark data is so much easier to gather than traditional map data.

Sources of the data depend on the country: 1. For example in the US or Europe, we can work mainly by combining information from a number of online sources with only some checkups from people actually "on the scene". 2. For a lot of developing countries (Nigeria is one of our strong areas for differentiation from other datasources), we will have some local guy physically go around the city and gather the raw data that we then process for accuracy and quality.

I'll dig up some examples where the usual digital maps fail and we succeed.

I am the head tech guy of the company, and am not so accustomed to preaching about the power of landmarks. So thanks for the opportunity to work on my arguments!

Re the app: Try it! It's much faster with the built-in landmarks compared to writing it yourself, and will change your texting behavior. Once you get used to it, texting without a location feels just so... last season. :)

95% of the time if I put a location in my message it's most likely a location that I and the receiver are familiar with, so such a standardized name is not too desirable. But of course it's just me. I can definitely see, a lot of couples using that all the time (beware cheating girlfriends!).

Does it work backwards? So that if somebody uses it too and sends me a message using that can I look up the place on the map? That would be a big plus.

But yea, now that I thought of it. In very big cities where people move around a lot in a fairly constrained area that is very nice. If you can add more features and market that better I can see some people using it. Not quiet revolutionary, but nice. Of course provided that this landmark database is really really, really good.

<warning, long reply which hopefully also answers all your questions>

You can also easily add you own points, which is what you probably end up using a lot. "I'm Home now", "Still at the Office". It's a one time tap/enter name to add these, and these are the ones you probably end up using a lot.

(Not really related, but: Also usually the app is pretty good at picking up the correct landmark even without GPS (from the inexact cellid location), so you can just forget about it, send your text and automatically also share your location.)

Regarding the cheating girlfriends, we let the user select any point she wants. This is by design, this is not a tracking application but a communication tool. For example sometimes you know your mother will not be familiar with the closest automatically selected landmark, and can easily select another one.

It's not about exact location (going all psychological: there is not exact location, ever), it's about human understandable location.

Currently it doesn't work backwards. On a concept level, we want to think maps suck and landmarks rule. On a technical level, this is something we may well implement. (Sidenote: I guess the view with the landmark points can also be considered a "Map", and that we definitely could want to show on the receiving end.)

Thanks for your positive words! This is version 1.0 in terms of marketing texts and everything, so the journey for this app is just beginning.

Do you have to enter your location manually on each text? If not, how does it know where you are? If yes, how is this better than just typing it in yourself?

I think answering this point on the website is crucial.

I guess the answer is two folded:

1. To get your location (as in coordinates), it uses the default Android APIs that utilize GPS, cellid and WLAN.

2. To make these two numbers human-readable and instantly understandable, it uses our landmark database to look up the nearest landmark.

So that's how we go from


  at Googleplex Headquarters, Mountain View, Sta. Clara, US
You can naturally select another landmark manually if you want to, or edit the landmark text.

Do you think the text we have on Android Market fails to explain this "magic"? What would you add there?

(update: formatting, added question)

I guess the part about your phone already being able to determine where you are. It's quite common to have GPS turned off to save battery afaik, and I'm not sure how Android would get my coordinates from the other two, unless they maintain a database of all wlan networks in the world.

I'd add an explicit 'no need to enter your location, that's what our application does' text.

Have you used an iPhone / Android device?

Because Google and Apple have pretty much nailed this. That is exactly what they have: a database with ENOUGH of the GSM (CMDS) basestatio cellids and WLAN identifiers, linked to real coordinates, so that most of the time things "just work".

So all the phone needs is to use some mobile data to get this database, at least once before it caches the local id => coordinates information. After this, a rough location is available even without GPS.

How can they do this? Because everytime someone somewhere uses Google Maps with GPS, they are also submitting data to Google's cellid and WLAN databases. So it works really well in countries and areas with lots of smartphone users, and less so when you go to the outskirts of the Philippines.

Also: Landmarks work pretty well even with the non exact cellid location, you definitely don't need to have GPS enabled to use the app.

This is basic platform functionality, which is why we didn't emphasize it in the text: However maybe we should add some information about this as well.

Yes, I've had and actively used an Android for about a year and half now. I use it quite extensively, have bought a number of applications and I guess I'm basically the audience you're targeting - people who actively go out to find, and potentially buy, applications (this group is imo a relatively small subset of all Android users, I know many more Android users who have never used the Market than those who have).

Anyway just to make it clear that it is, imo, not to be considered 'well known' information amongst the general Android-using population. Heck, I've even written toy Android applications, and I still didn't know.

Thanks for the insight! I guess I've been to involved in location stuff for a while.

You'll get a feel of the location system with Google Maps quite easily: You get a big light blue circle around "own location" when location comes from mobile network cellid, and a smaller one when it uses WLAN-based location. That signifies the accuracy. And it does indeed find this location even when GPS is off. And this works even when "Mobile network"-setting is off, i.e. the phone is not sending or receiving any data, if you are still connected to a previously known mobile cell.

Anecdotally: I moved a few months ago, and didn't change my WLAN network id. At my new place, when GPS was not available, Google Maps was showing my location to my old place. This continued until I renamed the WLAN and used Google Maps with GPS near my new place.

This shows the weakness of crowdsourced location data, but such incidents are rare enough that it basically "just works".

Can I ask what problem your app solves? Why would I want to add my location to an sms? Good job anyway, sounds interesting.

Because a lot of times you want to tell people where you are, or where to meet. So this app does that for you, automatically. In a way that the recipient will understand.

ยค Gecko Landmarks EU HQ, Turku, Finland :)

if I want to meet up with someone, wouldn't I just give them an exact location? if the nearest landmark is 5 miles away I don't see how that would be useful.

How would you define that exact location, with coordinates?

What if we have gathered just the right amount of landmarks from the area, so it's pretty likely that all the well known places to meet at are there already?

Naturally if you are in a rural area far away from any kinds of "real" landmarks, this will not help you. But by definition, such areas will have low number of potential users as well. We think this works well for a lot of people a lot of the time.

Hi brosephius, this blog post might open our idea a bit more, I think the word "landmark" can have a few different meanings:


So we hopefully have the right McDonald's already mapped for you.

How is this "revolutionary" other than it revolutionarily cuts down on the remaining characters to send?

And it's not really revolutionary in the way that BBM or iMessages is (i.e. taking sms away from the carriers).

It's revolutionary because:

- Sharing you location is a very basic need in communication (don't think Google Latitude on smart phones, think 90.7% of mobile phone calls including the question "where are you?", globally, everywhere)

- Landmarks is a concept that has evolved to solve this; Well known places that can be used to communicate location. And they fit the text based communication perfectly.

- Traditional maps, which can be seen as a competitor, lacks in a few major areas:

โ†’ How do you send a map? SMS is the global standard of communication, you can't SMS a map

โ†’ More than half of the people in the world can't read maps. At all. These people still understand the local landmarks.

โ†’ Good maps are not available everywhere, and will not be anytime soon as traditional maps are expensive to make

So revolutionary in quite a different way than BBM or iMessages.

(edit: formatting)

Sure, but it doesn't seem like your company's offering solves any of these problems _well_. Smartphone solve this problem already by letting one send contact point (such as a location from the Maps app on iPhone) over mms.

For a majority of messages it seems like your product mostly eliminates remaining characters.

Did you try the app? If you did: Where are you located, how did you find the local landmarks? What would you improve in the app?

I agree that if (1) the receiving device is a smartphone, (2) there is map data available for the used area, and (3) the users knows how to read maps, sending a location link with MMS works well. Or a http- google-maps link with SMS, that gets opened in Google Maps at least in iPhone and Android so works pretty brilliantly too and doesn't cost the price of MMS.

I think we have pretty much proven that landmarks in SMS is a killer feature for consumer vehicle tracking with Tramigo (best selling GPS/GSM tracking device), and in that use simplicity is the key: plain SMS, no need for data connection or anything else than just a plain GSM phone. So we are trying to expand from there.

I see that it's free. What's the catch? Ads? Does it spam my friends? Anything else that I might not think of?

The most important thing for us is to spread the understanding of our landmark concept. As such the idea is almost trivial: we have coordinates and names for point-of-interest etc. However we think that with highly optimized data those can be a really effective way to communicate location. We have the global landmark database already, and a sustainable business for it, but now we are looking for new ways to utilize it.

There's ads when viewing a menu ("Landmark" button in compose view), not so much for the money but to study if that might work. I think the ads need to be very unobstructive in any app like this that we expect the user to use a lot, so we are not counting on that.

It definitely does not spam your friends.

It's like the Locate in Handcent, but you're basically "rounding-off" to the nearest "landmark"?

Kind of, but the receiver does not need to click on a http-link to understand the location.

So like Locate in Handcent, but Gecko SMS also works

- when texting to your mother/grandpa/non-techie friend with old dumbphone

- in places like Balikpapan, Indonesia, with zero Google map data coverage.

Handcent Locate can also optionally add an address (created from Google street data), which may work well if the data is there. However landmarks are designed to be well known spots, whereas addresses are often cryptic.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact