keskiviikko 26. syyskuuta 2012

iApple's iOS's iMaps

Recently i[sic]'ve been following the "new iOS maps are rubbish"-discussion with moments of humour, fumour and complete astonishment. First of all, let's get facts straight - most of the time I do like what Apple is doing (so there is your fanboy card) and other times I honestly hate what they're doing (and there is your ... umm, what... "anti-fanboy"-card?)

So... what is the problem?


Well, Apple is the newcomer in the field of mapping. That much is certain.

For some reason, everyone expected their maps to be a marvel in this field. Granted, what they put up in their earliest proof-of-concepts (or actually - exactly what they put up there) was nothing short of what I like to call "the Apple fantastique". Eventually we have found out that those PoCs were exactly that; exaggerated fancies of the company executives of "what this technology can be at it's best" ... ok, we only need to go back a short while to see how Nokia announced their image stabilisation and how they were crusifixed after it became apparent that the technology demos were not taken with the actual piece in question.

Oukidou, I'm straying a bit off the subject here... back to Apple. Before publishing iOS6, we got plastered with some magnificent looking "3D-flyover look of real map data through our acquired partners' software". And yes - those looked magnificent. But no-one explicitly stated that "this is what our maps will look like". It was implied, true ... but in the business they're in - they are selling impressions beforehand. And - we did see that the same imagery they used in the technology demos (from around the Hoover Dam) the imagery provided by iOS6 was way different. Even emparrasingly different, I might add.

Ok, so - what's the catch?


Oukelidou - you might be wondering... why am I blogging about this? They didn't deliver what they promised?

Ah... but there is the catch there... They didn't promise the final user experience was going to be anything like the technology demos. Never. They just said "with this new technology we just bought, this is what we can do". What they didn't say is "...on something your handheld smart phone is ready to handle... at least just quite not yet".

So again, why is it that you (I) are (am) blogging about this again?

Well - guess what..?

What they did is what every operator in the field does. And for some reason, people still do expect more from Apple. Well... they do.

Granted, if mr. Jobs was still heading the company, this writer's personal belief is that the decision to drop Google from their data providers would not have been so quick. Steve would've looked into it that when they bring a "new feature of their own" into the game, it would've matched or would've been better than that of the competition's.

So... again... why am I (you) interested in this?

The problem with people at this point is that they are comparing (pardon the pun) Apples to oranges. All the competition has been there for ages; Google maps was launched 7 years ago, Bing had their mapping system debut almost two years ago. Tomtom et al have been at the business for almost a decade.

Ok... so what? So... interestingly, all of these parties are buying data from each other. They do it to make their maps better. The origin of the maps they buy might be very old (the newer the data you acquire, the more it costs - no matter if it has been changed or not).

Basically; all of the players start with the same set of data (if they're willing to pay the price). Getting the "latest set of data" from any one source will not be enough, since all of the data acquirement will basically be a one-way street; you buy the data, you do whatever you want with it... and it never gets back to the source.

That is the way for mapping companies to keep their advantage ahead of other mapping companies. "We have these updates, you don't. We win." ... or "You want these updates, you pay", if you want ;)

They all start with the same dataset. For example, many of the pictures I've seen posted online show how "empty" or "wrong" the Apple datasets are. Guess what - Google looked the same just a little while ago!

After the companies start with a dataset, they update it ... and (you might've guessed it) these updates never get back to where the data originated from. The "fixed data" in the database is the company's own "we have these, you don't"-way to success. Whoever has the most (and most correct) data is the winner.

I wouldn't be surprised if Apple took the lead of the competition fairly quickly. Not to mention their own "report problems with map", they have quite a lot of information from sites that list irregularities in iOS6 maps. What other companies think as a "win" for them, might actually be a "win" for Apple... Honestly - if you see something funny on a map, are you more inclined to report it to Apple or to just post it online for everyone to see ... hey, guess what - even Apple follows the Interwebs ;)

Oukelidoukeli, this is all important... anything more?

Well... not exactly important, but... honestly - people are sending in a million or so pictures of problems with bridges and airports. Really?!?

With bridges, the companies have very little to go by. They get the generic orthographic imagery of the area and they get the height data (usually) from NASA.

As good as NASA is mapping the universe, their systems are fairly limited when targeted to earth. For example, for a bridge (and - assuming You've read what I wrote earlier about the data sources - a bridge that has not been "manually fixed") they generally do get the "height of the water-level" and either "the height of the deck" or "the height of the pylon" ... depending on the resolution. But never both.

The reason above is the "zigjaw-lines on bridges" seen on so many pictures recently. Honestly, every bloody bridge needs to be "manually fixed" for it to look right.

Back to previous... Apple posted some magnificent footage of the Hoover Dam in the technology demo. However, we've seen how it turned out in real life application (bad). Also, we've seen seom other applications' rendition of the same area (which are also bad). So... Making stuff like that work well in 3D... either you need to have a good 3D source for all of the world or you need to manually fix the problems. No shortcuts.

And now back to the fanboy-choir


Just wait. Few more months, few more years. After that I believe Apple will have no problems in the competition.

Please do not compare a newcomer (2 weeks or what) to oldtimers (360+ weeks) just yet at this point ;)

And I think that final sentence did put me back in the fanboy-choir... oh'dammit ;)