Traceback (most recent call last):
File "/base/python_runtime/python_lib/versions/1/google/appengine/ext/webapp/_webapp25.py", line 701, in __call__
handler.get(*groups)
File "/base/data/home/apps/mapvelopes/main.341084165938999830/handler.py", line 274, in get
start_loc, end_loc, route = get_route(start_address, end_address)
File "/base/data/home/apps/mapvelopes/main.341084165938999830/handler.py", line 104, in get_route
start_geocode = decode_geocode_response(route['legs'][0]['start_geocode'])
KeyError: 'start_geocode'
Traffic from an HN link is a lot over the course of a few hours, but it isn't that hard to handle. A single dedicated server usually won't even break a sweat. That's why this looks really bad.
What do you want for free? "A single dedicated server" costs something... this guy is almost certainly running on the free quota.
I don't think this looks bad for Google -- it looks bad for the author. The fact that he's using the freebie *.appspot.com rather than a dedicated domain gives you some indication of his dedication to the project.
A single server not breaking a sweat means that this can be handled by a middling VPS that is breaking a sweat, and that load only lasts a few hours. That means that it's stupidly cheap for Google to provide for a few hours of load like this, and the ability to absorb a huge spike on a cheapo plan is supposed to be one of the main points of cloud hosting.
Lots of people make cool stuff without dotting all the i's. That doesn't look bad on them - it just means they're doing the hard stuff and skipping the grungy parts of a project until it seems necessary.
What? You're serious? We should all design massive scaling and redundancy features into our hobby projects? That's absurd.
It's not like Google can't handle the extra bandwidth, it'd be much better if they left the page up and reached out to the developer to upgrade their account.
To be fair to the author, it can be surprising how much traffic you get, especially if it's your first foray into "web stuff".
I recently posted a presentation of mine and was completely blindsided by how quickly I burned through the available quota. Granted, it was showcasing WebGL and WebAudio, so there was a bunch of data per download, but I still only barely managed to put in CC info before I reached quota.
(And the best part, traffic fell off as soon as I signed up for billing. :)
It'd be nice if users could indeed bump quota, as suggested above.
Yeah, I didn't mean to imply that it was the author's fault, I was just flabbergasted by the assertion that it was a lot of work to make a GAE application scalable. Really, it's only the author's fault if he posted this link.
I did like the idea of user's bumping quota, as well.
More than a decade into hosting-as-a-business model, I still don't see why 'temporarily over quota' messages aren't replaced with an offer to let the audience sponsor a bump in quota, either for just the current viewer or however many viewers the next quantum of hosting can support.
Odd that I've never heard this before, this is pretty ingenious. Alternately instead of sponsoring, could display ads, etc, but how ever it's done, the user is currently being punished, with no option to make corrections.
I am also getting a server error, which is a real shame, this is pretty awesome.
Suggestion: let me email/upload a photo or two, as well as the addresses and print retro postcard like views of the scaled down photos on top of the map.
It seems to be a "an App Engine app that generates customized printable envelopes with the map to your recipient on them". Here's the developer's blog post about it: http://bit.ly/sVgTGh
Traceback (most recent call last): File "/base/python_runtime/python_lib/versions/1/google/appengine/ext/webapp/_webapp25.py", line 701, in __call__ handler.get(*groups) File "/base/data/home/apps/mapvelopes/main.341084165938999830/handler.py", line 274, in get start_loc, end_loc, route = get_route(start_address, end_address) File "/base/data/home/apps/mapvelopes/main.341084165938999830/handler.py", line 104, in get_route start_geocode = decode_geocode_response(route['legs'][0]['start_geocode']) KeyError: 'start_geocode'