Hacker News new | ask | show | jobs
by ChrisMarshallNY 1709 days ago
I’m using Apple Maps for the app that I’m developing now.

The alternative would not have been Google Maps. It would have been MapBox.

Google Maps will give app developers like me severe sticker shock. I watched a friend of mine get it in the shorts (forcing him to switch to MapBox). After watching that, I decided that I will never depend on Google services as a linchpin technology. In fact, it simply hardened my resolve to avoid third-party dependence, wherever possible.

When I started the project I’m working on now, Apple Maps had already grown to cover all of the US, albeit not as well as Google Maps. I’ve been seeing Apple Maps cars driving around here, for at least a couple of years.

I’m pretty confident that relying on Apple Maps was not a mistake.

3 comments

> I’m using Apple Maps for the app that I’m developing now. The alternative would not have been Google Maps. It would have been MapBox.

It's easy to make a theoretical stance like this, but Apple Maps and Google Maps are free in an iOS app, but MapBox is not. If you were forced to switch off Apple Maps for some reason, it seems likely the free option would win out.

> it simply hardened my resolve to avoid third-party dependence, wherever possible

all of these libraries are third-party dependencies...

> Google Maps will give app developers like me severe sticker shock

If we're talking just about the javascript API, in my experience if you're using more than just a very local map, people always have very wishful thinking about MapBox pricing and it usually ends up almost the same as Google Maps pricing. This isn't an accident, they need to be a better priced alternative but they also need to pay the bills. The pull to switch at that point would be if you need better address lookup or something.

If you are doing a very local map (like the equivalent of a few tiles for your average user), you're often better off just rolling your own tile server. People are often surprised how easy it is.

> Apple Maps and Google Maps are free in an iOS app

Google API is not free. It would be at least a 4-digit cost per month. I think that my friend was hit with a $1,500/mo nut. Google changed terms (back when they started demanding that all users of the API give credit card numbers). His app is fairly popular, and he'd need to pay them for the SDK.

> all of these libraries are third-party dependencies...

I've been relying on Apple's toolkits since 1986. They've pulled a couple of stunts (OpenDoc, anyone?), but they have been a reliable partner, for the most part.

> If we're talking just about the javascript API

I don't do JS. I do native Swift. It would require embedding their API. My friend did that, and it was far more cost-effective. Lot less than Google (he sells his app for $0.99, and it's not popular enough to do much more than pay for itself). MapBox is quite nice. His app is native ObjC. He's had it out for quite a while. He originally started with Google's API, as that was the only one that offered the particular services he needed.

Most of the stuff I write is free, open-source. I could probably get Google to let us use it for free (we're a 501(c)(3)), but I've already seen them turn the screws once, and I'm in no mind to set myself up for something like them deciding that we could give them a tithe.

Also, I have had some ... issues ... with third-party libraries. I'm quite particular about the Quality of my work, and I find that not too many others share my passion.

> Google API is not free. It would be at least a 4-digit cost per month

On iOS and Android, google maps are free: https://developers.google.com/maps/billing/gmp-billing#mobil...

Maybe your friend was doing something that the native SDKs couldn't do but the web API could?

I know that he used directions and elevations.

He told me that his use case was definitely NOT free.

I also know that Apple Maps could not give him the specializations he was looking for. He is an experienced, rational engineer, and would not have acted in an alarmist manner. He really liked the Google API. Google has done a wonderful job with their map product, but I also know that it's a "gateway drug."

I won't link to his app, because it can get crazy, with all the anti-Apple stuff. I won't drag him into the fray.

As ever, worth noting that self-hosting maps can be a very viable option.

You can generate OSM-derived vector tiles for the whole world for €15, using tilemaker running on Scaleway bare metal. It'd cost more like ~€100 to run it on AWS but still very doable.

That gives you a big sqlite database file (.mbtiles) to host wherever you want. Use Maplibre, the open-source fork of Mapbox GL, in your app or site.

Did you consider here.com, and if not, why?
Yeah, but I have seen MapBox in action in a number of apps, and have yet to see one native app, delivered with here.com. As it is, I even decided against using MapBox.

I'd need to see it in use a bit more, before relying on it for my users. Most of what I do is free, or low-cost work, aimed at folks in need. In some cases, lives depend on my work, and there's little money or glory.

That Google "squeal piggy!" thing I witnessed was quite sobering.

People seem to be quite sanguine about putting all their eggs into one basket. I guess that's kind of necessary, these days, but we need to be careful about choosing baskets.

I'm kind of a crusty, scarred ol' bastard. Takes more than a pretty Web site to get me on board, and I don't bring buckets of money with me, so I'm just not that attractive.

> Yeah, but I have seen MapBox in action in a number of apps, and have yet to see one native app, delivered with here.com.

Makes sense. I encountered here.com once in a corporate setting, and was surprised that it was actually decent (at the time), but no further experience with it either. It doesn't seem to have a strong brand outside big corp.

At one point, Facebook was using here.com for event locations