I noticed this as well. At the default level of zoom my browser was at with the link, the bridge was there. When I zoomed out one level or more, it wasn't.
If you zoom in a little further, you'll notice that the road is actually outlined in a dotted line and notated as "planned". So it's there just as a placeholder.
If editing is too complicated/time consuming, then you can use the fancy new "notes" feature to add a marker that someone else can come and fix. It's the "Add a note" link in the bottom right corner. (The bridge has been removed now, by the way.)
I had look at this history and someone marked the roads as closed some time last night. The bridge itself was later removed by another user. Later still, another user (from Turkey) seems to have completely replaced everything back to normal for some reason. We need a local user to make the correct changes and make sure they're kept that way.
Interesting problem. How do Google's self-driving cars react to road problems such as bridge-collapses and sinkholes? Scanning for objects and avoiding objects which extend above the road level is obvious, but what about when the road cuts off? These things rarely happen, but it would be terrifying to be in a self-driving car in that moment when the bridge ahead of you collapses and the car continues to drive as if there is nothing wrong.
I actually had a similar thought yesterday when thinking about how the cars would handle object-avoidance. Say a deer hops out in front of the car. The car is constantly aware of its environment for 100+ feet in all directions, so it could tell if there are no cars in a neighboring lane and swerve to miss the deer. But what if there _are_ cars in neighboring lanes and the only available direction is to go onto the shoulder or into the ditch? If the ditch is fairly level with the road, then this could be the safest option. But if the ditch is steep, then it would be more dangerous than simply hitting the brakes and hoping you don't hit the deer too hard.
I would think that road integrity would be on the long list of things covered in any licensing process (you and I can think of it in a few moments, ergo). I would also guess it is important for the system to be able to estimate an appropriate speed even when external data is not available.
For the deer, I would expect the programmed response to be to brake as much as possible given traffic. Hitting a deer is unlikely to result in anything but damage to the vehicle. Swerving is less predictable than that.
The vector generated tile at the higher zoom has been redrawn while the tile below has not been redrawn yet, I would assume the redraw rate would be less frequent the more you zoom in do to decreased demand for the tiles.
The second interesting thing is that even though the raster tiles have had the bridge removed from the draw instruction the vector data for the click map on the map is still intact. That is interesting because that means that the vector data used to render tiles is operating at a different version than the vector data that is draw on the map. Which could lead to all kinds of weird scenarios.
Guessing, that gmaps switches to drawing vector data at a certain zoom instead of rendering the road line vector data to the TMS tile.
This simply means that the version of the vector data used to draw at the increased zoom is the older version but is actually being used to draw the roads on the browser instead of rendered into the tile before transfer.
I hadn't heard this story previously, so I was not sure what happened. I guess we're on the wrong site to get general news, of course a tech story would be ranked higher :)
Further to this, it's a pretty interesting technical problem to think about why the bridge collapsed. A compression member or node at the top of the truss that forms the main span seems to have failed for some reason. The news story linked in another comment below says the truck hit something. This[0] secondary(tertiary?) truss looks like it might have taken a hit? It forms part of the horizontal truss above the carriageways which resists shear in the horizontal plane and therefore provides torsional stability; the bridge is effectively a braced tube, I think. Maybe the hit on that pulled the main compression member out of alignment enough for it to buckle? I'm not an engineer, but I'm pretty interested in engineering, I'd be interested to know if anyone with better knowledge can spot an obvious reason for the collapse.
To be fair, the fact that the bridge collapsed, while it is news, it's no news for HN.
The technological aspect of google maps being able to so quickly update it is in fact, new for HN.
Another thing that I agree would be interesting is a detailed technical explanation of why the bridge collapse, that I'd assume would be HN material as well.
I don't think that something has to be strictly "technical" to be on HN. I would like to see the the story on how the bridge came to be in the state it is in on HN - not just the technical explanation for what happened, but why did we choose to allocate our resources in such a way that the failure occurred.
Comments on reddit mentioned that the bridge was condemned in 1992, and repeatedly since then, but no one bothered to fix it. I thought we had a big goverment stimulus package for shovel-ready projects over the past few years...
And here I thought its popularity would be due to someone alleging that the only way Google could respond that quickly is if they knew the bridge was collapsing. :-) But web apps are an amazing thing, and this is an advantage over the old Streets & Trips type databases.
I am a mapmaker editor and I have seen my changes instantaneously get published to prod/GoogleMaps for everyone to consume over Internet. I think this depends on how mature the editor is.
Btw, there are plenty of nerdy ppl to correct the web. Ref: http://xkcd.com/386/
Apparently Google is quick to respond to infrastructural failures.
So why didn't it remove calibration scores, and convert the Perf Room to something fun like an arcade (well, it already is a casino, so maybe something else) many years ago?
There's a difference. You know very little about me and where I come from and why I see things the way I do. Do I bring a lot of negativity into these discussions? Sure, but that's needed because I don't want thousands of people to repeat mistakes that I did. If bringing a realistic or even slightly pessimistic tone to this often out-of-touch startup discussion is the cost to prevent a thousand young kids from wrecking their careers, then good.
However, I've articulated deep managerial problems with Google that the company refuses to solve, and many of those can be fixed easily and for free.
So there's no comparison between what I say and what you said.
Every quarter, managers get together to horse-trade "performance points" (or "calibration scores") and generate a company-wide stack ranking of all engineers. If you end up in the bottom 5% (which sometimes happens to people just because their managers don't show up) then you end up on a PIP and your career at Google is over. That monstrosity is called Perf.
The "Perf Room" is a metaphor on most campuses because most of the smaller ones don't actually have a dedicated room "where it happens" for calibration.
"Take you into the Perf Room" is a common managerial threat, but actually everyone is calibrated and the meeting happens in secret (most engineers never learn their scores) so people don't actually literally get taken "into the Perf Room". It's just an expression for a bad review.
Your comment about infrastructure was actually quite hilarious and interesting, and Vikings's repartee comeback was witty as well. This post though made the whole exchange less fun! Next time, I recommend try either responding with "LOL" to acknowledge Viking's wit, or come back with an even more humorous response.
I also know you think there are easy and free solutions to the problems that every company on the planet struggles with every day (measuring and rewarding performance).
I think you're arrogant, disrespectful, and that you over-simplify problems to an absurd degree.
I can conclude that if I worked at Google, I don't think I'd appreciate working on your team.
And you're right, my post didn't have very much to do with yours. Did yours have a lot to do with Google Maps responding quickly to an emerging tragedy? No, it did not.
I know you're the type of person who publicly tells their former employer, "Choke on a fucking taint, Google. Choke. On. A. Taint."
You shaved off the enclosing context: $100M payouts to nontechnical executives when regular engineers can't even transfer without permission of the perf-industrial complex. I'm sorry, but I just can't defend a company that behaves that way. Seriously, you think it's okay that engineers have to deal with phony scarcity ("calibration score" and "headcount" nonsense) when non-technical executives make several hundred times more?
I think you're arrogant, disrespectful, and that you over-simplify problems to an absurd degree.
It's more that I don't have patience for junk complexity that exists only to keep power for those who don't deserve it.
Did yours have a lot to do with Google Maps responding quickly to an emerging tragedy?
Yes, actually. It's good that they updated their map quickly. If they paid the same respect to internal problems, they'd be a much better company.
You're lying: I did not shave off the enclosing context, I explicitly provided the link for everyone to see.
It's good that they got you to leave. If they got everyone like you to leave, it would be a much better company. See, my post has exactly as much relevancy as yours does.
If you have something to say, why not write it up clearly in a blog post, rather than making vague and confusing off-topic comments like this in any HN thread that mentions Google?
If what you say is true, I would think HN would vote it to the top of the front page, and that could lead to reporters calling you and giving you the opportunity to make your case in front of a much wider audience.
WTF is wrong with you? It used to be the case that you'd only post your rants about Google in threads that were at least tangentially related to working there. But as of late you've started spamming random Google stories with these toplevel comments that have nothing to do with the article or the discussion. Do you really not see that this is not acceptable behavior?
55 comments
[ 3.2 ms ] story [ 119 ms ] threadI tried to mark it as closed, but did not find any option for that in potlatch. Can someone else with more experience with OSM help out?
http://www.openstreetmap.org/browse/changesets?bbox=-122.345...
I actually had a similar thought yesterday when thinking about how the cars would handle object-avoidance. Say a deer hops out in front of the car. The car is constantly aware of its environment for 100+ feet in all directions, so it could tell if there are no cars in a neighboring lane and swerve to miss the deer. But what if there _are_ cars in neighboring lanes and the only available direction is to go onto the shoulder or into the ditch? If the ditch is fairly level with the road, then this could be the safest option. But if the ditch is steep, then it would be more dangerous than simply hitting the brakes and hoping you don't hit the deer too hard.
For the deer, I would expect the programmed response to be to brake as much as possible given traffic. Hitting a deer is unlikely to result in anything but damage to the vehicle. Swerving is less predictable than that.
But at the default zoom level, I do see it: http://puu.sh/30828.png
The vector generated tile at the higher zoom has been redrawn while the tile below has not been redrawn yet, I would assume the redraw rate would be less frequent the more you zoom in do to decreased demand for the tiles.
The second interesting thing is that even though the raster tiles have had the bridge removed from the draw instruction the vector data for the click map on the map is still intact. That is interesting because that means that the vector data used to render tiles is operating at a different version than the vector data that is draw on the map. Which could lead to all kinds of weird scenarios.
Guessing, that gmaps switches to drawing vector data at a certain zoom instead of rendering the road line vector data to the TMS tile.
This simply means that the version of the vector data used to draw at the increased zoom is the older version but is actually being used to draw the roads on the browser instead of rendered into the tile before transfer.
Edit: http://www.seattlepi.com/local/article/Interstate-5-bridge-c...
Here's the Reddit thread a FB friend linked: http://www.reddit.com/r/Seattle/comments/1ey32g/i5_bridge_in...
[0] http://www.seattlepi.com/local/article/Interstate-5-bridge-c...
The technological aspect of google maps being able to so quickly update it is in fact, new for HN.
Another thing that I agree would be interesting is a detailed technical explanation of why the bridge collapse, that I'd assume would be HN material as well.
Btw, there are plenty of nerdy ppl to correct the web. Ref: http://xkcd.com/386/
So why didn't it remove calibration scores, and convert the Perf Room to something fun like an arcade (well, it already is a casino, so maybe something else) many years ago?
(For the record, I feel bad saying this... but I'm also just returning snark with snark, so... oh well.)
However, I've articulated deep managerial problems with Google that the company refuses to solve, and many of those can be fixed easily and for free.
So there's no comparison between what I say and what you said.
Every quarter, managers get together to horse-trade "performance points" (or "calibration scores") and generate a company-wide stack ranking of all engineers. If you end up in the bottom 5% (which sometimes happens to people just because their managers don't show up) then you end up on a PIP and your career at Google is over. That monstrosity is called Perf.
The "Perf Room" is a metaphor on most campuses because most of the smaller ones don't actually have a dedicated room "where it happens" for calibration.
"Take you into the Perf Room" is a common managerial threat, but actually everyone is calibrated and the meeting happens in secret (most engineers never learn their scores) so people don't actually literally get taken "into the Perf Room". It's just an expression for a bad review.
https://news.ycombinator.com/item?id=5508058
I also know you think there are easy and free solutions to the problems that every company on the planet struggles with every day (measuring and rewarding performance).
I think you're arrogant, disrespectful, and that you over-simplify problems to an absurd degree.
I can conclude that if I worked at Google, I don't think I'd appreciate working on your team.
And you're right, my post didn't have very much to do with yours. Did yours have a lot to do with Google Maps responding quickly to an emerging tragedy? No, it did not.
You shaved off the enclosing context: $100M payouts to nontechnical executives when regular engineers can't even transfer without permission of the perf-industrial complex. I'm sorry, but I just can't defend a company that behaves that way. Seriously, you think it's okay that engineers have to deal with phony scarcity ("calibration score" and "headcount" nonsense) when non-technical executives make several hundred times more?
I think you're arrogant, disrespectful, and that you over-simplify problems to an absurd degree.
It's more that I don't have patience for junk complexity that exists only to keep power for those who don't deserve it.
Did yours have a lot to do with Google Maps responding quickly to an emerging tragedy?
Yes, actually. It's good that they updated their map quickly. If they paid the same respect to internal problems, they'd be a much better company.
It's good that they got you to leave. If they got everyone like you to leave, it would be a much better company. See, my post has exactly as much relevancy as yours does.
If what you say is true, I would think HN would vote it to the top of the front page, and that could lead to reporters calling you and giving you the opportunity to make your case in front of a much wider audience.