9 ms·
This is a crap response. I went to Shanghai about 3 summers ago right when Denver was finishing up the "A Line" train from Downtown to the Airport. This project
by snow_mac 8y ago
This is a crap response. I went to Shanghai about 3 summers ago right when Denver was finishing up the "A Line" train from Downtown to the Airport. This project cost over a billion dollars and the damn thing can't run smoothly or stay operational. Tons of issues. Meanwhile we took the Maglev train from the Pudong Airport to the Shanghai City Center. The Maglev cost $1.2 billion dollars [1]. The train can travel up to 270 MPH [1]; it takes about 15 minutes to get from the Airport to the city. The Denver train? 45 minutes, goes somewhere around 45 MPH for $1.1 billion dollars [2]. The train in China goes 6x faster then the Denver train.
So, tell me again how the Chinese infrastructure is not good?
[1] http://www.slate.com/articles/arts/go_east_young_man/2005/03/mines_faster_than_yours.html http://www.slate.com/articles/arts/go_east_young_man/2005/03...
[2] https://pagetwo.completecolorado.com/2016/05/31/otoole-taking-rtds-billion-dollar-a-train/ https://pagetwo.completecolorado.com/2016/05/31/otoole-takin...
- floren 8y agoRemember when they crashed a high speed train and buried all the evidence the next day? https://en.wikipedia.org/wiki/Wenzhou_train_collision https://en.wikipedia.org/wiki/Wenzhou_train_collision
- taobility 8y agoyou remind me of this incident as a joke: https://en.wikipedia.org/wiki/2017_Washington_train_derailment https://en.wikipedia.org/wiki/2017_Washington_train_derailme...
- floren 8y agoHmm yes, this is exactly the same sort of situation: > WSDOT announced on December 21 that it would not resume Amtrak service on the Point Defiance Bypass until positive train control is implemented in 2018. "But America has also had train crashes" is a typical response in this sort of situation, but also typically irrelevant.
- taobility 8y agoAs a software engineer, I knew the bug in every software is inevitable. What we should do is learning from each bug and try to find a solution to avoid repeat the mistake, either by some engineering process, or more tests. And blindly blame the engineer who wrote the codes is stupid.