Hacker News Reader

I couldn't agree with this anymore. I'm personally getting tired of reading all the armchair critics on HN bashing the game itself. Here are articles popping up saying in under a week it has more installs than Tinder and is about to surpass Twitter in daily active users. Twitter has had 10 years to learn how to handle this much traffic, the Pokemon Go developers haven't even had 10 days.

Look, it's a new app. There's going to be issues with it. There's going to be missing features that everybody wants. Does that mean they did a shitty job? No, not at all. I get that releasing a buggy app is far from ideal, but does anyone here actually know what they're going through? What constraints they had? What their budget was? No, no we don't. So who are we to sit here on a form and put down something that our fellow developers have built? To talk about how much better we would have done it?

I don't have the game yet, but my wife does, and it's been a ton of fun. We walked around our quiet little neighbourhood the other night. We both honestly figured we'd be the only people playing. Then we saw a mother and her son sitting on a bench at a corner defending their gym. Not 50 meters down the road an elderly couple, easily in their 70's asked us what team we were on, and playfully gave us crap since we were yellow and they were blue! In the end, we walked around for an hour and saw at least 20 different people or groups playing. Everyone with smiles on their faces. Everyone being friendly, outgoing, and talking to strangers. How can you consider this game anything but a huge success?

Start with their choice to use Unity. It's a nightmare for its memory consumption and wasted CPU cycles. It is utterly unsuitable for an app that is concerned with battery life.

I say this as an individual who has made something of a career out of fixing/polishing mobile Unity apps from utterly unshippable garbage to decent products. More than a few have called for significant native code accessed via the FFI, because Unity just can't cut it.

I would recommend not using Unity at all, but that would mean losing out on the cheap labour pouring out of the dime-a-dozen game development schools.

Oh, I have to admit I assumed Unity would be more stable to use as it's so popular and self-contained. That said, I don't recall other apps that use it crashing so often! Any clue if Unity is the cause?

I'm surprised it's just an ingress reskin. There seem to be lots of issues with the networking code. For example forcing login when the server's are down. Why bother with the OAUTH when they know it's down? Many states seem to require server interaction to progress to the next state, which makes sense, but there ought to be a timeout to return to the map. Many many times the game isn't non-responsive, but dosn't appear hung... a little X to abandon the current state chain would make the game far more usable.

I haven't played ingress in a couple of years. if it was simply a reskin, they'd have avoided so many "the network is always available" errors.

> Have you written software? That is the most simplistic advice in the world. I think the above poster was using hyperbolic language to make their point. Niantic was founded by engineers from Google. It seems a bit ridiculous that with all the expertise and funding they have, they couldn't even provide a stable backend architecture. It's not that difficult to scale an api using Google's app engine or aws.

>No caching? How would you cache game state when at any time the board state can change without warning, due to the actions of another player? And cached game state would invalidate your game actions, which would cause inconsistencies. The app could at least cache some of the info of your pokemon as well as the locations of nearby stops and gyms. Then only the status of those gyms and nearby pokemon would need to be synchronized when you reopen it.

It's plausible that the gyms are cached, as well as your mileage tracker - both of these take a few minutes to update (but from a user perspective they just seem broken)

> Niantic was founded by engineers from Google. It seems a bit ridiculous that with all the expertise and funding they have, they couldn't even provide a stable backend architecture. It's not that difficult to scale an api using Google's app engine or aws.

That's like saying "they have Google engineers, therefore they must have Google funding and they should be able to smart things out." AFAIK, Ingress was built on app engine, and had huge scaling issues.

> The app could at least cache some of the info of your pokemon as well as the locations of nearby stops and gyms. Then only the status of those gyms and nearby pokemon would need to be synchronized when you reopen it.

It's not clear how much faster that would be. The majority of the RPC time is probably due to network latency issues, not the amount of data being fetched. That said, without slapping a profiler on it, you can't tell either way.

Not that the comment you answered wasn't facile, but:

- buggy unreliable code is inevitable, but it shouldn't bring down the whole app. It isn't often that I see mobile apps crash to OS. There should be a finite number of things that can cause it.

- The are well know strategies to cache data that may change over time, this is true for all caches, if it can't change over time it isn't even really a cache, its just a static asset.

> By your thoughts, all software should have been updated in a few days to be bug free.

Anyone who has used this app knows it is extremely crashy. It's not supposed to be bug-free, but releasing an app this crashy is fairly sloppy.

Not arguing with the bugginess or not bugginess.

But arguing with "just log it and fix it" is like saying to a plumber "just find the leak and fix it".

It's insulting to both the developer and the plumber.

I think most plumbers wouldn't have much problem with that request, either hearing it or complying with it. After all plumbing is what they do. Hell I'm not even a plumber and I've found and fixed several plumbing leaks already this year.

Don't be condescending, I wouldn't have ranted or cared if I didn't have a clue. Others are commenting on game mechanics, lack of features or use of AR, but I'm not a game dev so I don't care.

This app crashes so much, it must be some really obvious exception they're not catching. I don't expect it to be bug free, but constantly crashing? Since you're obviously such a pro, let me ask you: Would you release software that crashes so often?

Apple doesn't have contigencies for top-tier developers and emergency updates? Huh.

Caching is a famously hard problem, but wiping the entire state and starting again for basic information that doesn't change is sloppy. Deltas and common sense helps here. Same thing for use of GPS and data.

Seriously, I have no idea why you're defending such cruddy work, but even if I was completely clueless or outright hypocritical (I've launched buggy software, trust me), it doesn't excuse Niantic's complete screw-up here. Normally I give devs the benefit of the doubt, but this is crazy. So many people affected for no good reason.

> Apple doesn't have contigencies for top-tier developers and emergency updates? Huh.

Thanks to the mysteries of Apple, sometimes your emergency update to fix a crashing bug takes several more days to get approved than the update that introduced the crash in the first place...

>Don't be condescending

Where was I condescending? I gave good answers to all your points.

>This app crashes so much, it must be some really obvious exception they're not catching.

Who's condescending now? "Since it breaks so much, sheesh, it should be really obvious how to fix it." Nothing is really obvious how to fix, if they knew how to fix it they already wood have

>I don't expect it to be bug free, but constantly crashing? Since you're obviously such a pro, let me ask you: Would you release software that crashes so often?

Inferiority complex much? I never said I was a pro. And releasing an obviously buggy product sometimes is tied to contractual agreements or marketing deals. Once you start a marketing blitz, you _need_to release something. Engineering is subordinate to actually, you know, making money.

>Apple doesn't have contigencies for top-tier developers and emergency updates? Huh.

Not for "top tier developers". Also emergency updates are generally only relate to when there are security or privacy concerns. Crashing doesn't bother them much, if it made it through their initial vetting process.

>Caching is a famously hard problem, but wiping the entire state and starting again for basic information that doesn't change is sloppy. Deltas and common sense helps here. Same thing for use of GPS and data.

You said yourself that the app is really buggy, and then you also want to introduce a level of caching on top of it? That will make it even buggier and harder to diagnose.

>So many people affected for no good reason.

Which lays credence to the fact that Niantic made the right balance between buggy software and shipping. Not many people would be affected if the game weren't popular despite its bugs and limitations, so shipping at this stage seems like the proper thing to do.

Hey, look! An App Store update today! Who could have predicted it??

There's a difference between bug-free and tolerable.

They could cache the images of frequent, nearby poke stops and gyms, which likely a huge portion of the data for any sync.