5 ms·

I'd be easy to pile-on with more criticism so I'll do the opposite. > Nick Van Weerdenburg, CEO and founder of Toronto-based Rangle, said his firm could likely

by belval 4y ago

I'd be easy to pile-on with more criticism so I'll do the opposite. > Nick Van Weerdenburg, CEO and founder of Toronto-based Rangle, said his firm could likely have built the app in less than a month for about $250,000. > “This is a very simple app,” he said. Having used ArriveCan I would agree with his quote. It is a very simple app, but what is not simple is writing the specification for it. The contractors weren't given the app and told "how long would it take you to make this?". They were approached to develop "an app to monitor COVID risk at the border", then they had to work with probably several committees, all with different visions and expectations, all trying to get their idea executed. That's often the part that people forget in favour of easy criticism. ArriveCan is simple, but the process that built it is not, there was no equivalent in the wild that could be used for inspiration, it had a bunch of translation requirements, and needed to interface with existing airport arrival systems. I'm not saying we should be proud of our government for making a 54M$ CRUD app that, as far as we can tell, did not have much of an impact on COVID being spread, but interviewing a bunch of "IT leaders" and acting as if they would actually sign a contract to build ArriveCan for 250k$ is disingenuous and the Globe and Mail should be ashamed of framing it this way.

bel_marinaio 4y ago
so you're saying $250k to build the app once you get specs from the $53,750,000 worth of meetings and planning committees? That sounds much more reasonable and not like $50million worth of tax payer money was stolen /s
- belval 4y ago
  Wouldn't it be fair to say that once an organization reaches a certain scale it is expected that communications constitute most of its overhead? Something like ArriveCan requires the collaboration of several branches of government and ties in with a system that monitors over 2M arrivals in our country every month. Even all the doers attitude in the world won't change the fact that they won't just give you access to those databases without some pretty extensive review. Just think of the QA needed on the arrival machines at the airport, you can't afford to have them all be out of order because they handle much of the incoming traffic, so the "ArriveCan update" has to be thoroughly tested so that it does not cause any significant downtime. This is the exact opposite of the "move fast and break things" environment. To summarize, I do think that ArriveCan (the program) was a waste of money, but I think the program itself did not waste money. It's the price that comes with building multi-systems applications.
  - kennend3 4y ago
    Two things for you. First, I fully agree with everything you said. It wasnt the best use of money, but these people fail to grasp what it takes to build a system integrated with the government. Dealing with any large org requires a lot of back and forth with contracts, legal, etc and the government is worse as they have entire organizations which seem to be created specifically to slow things down. Second, your points are most likely falling on deaf ears. All they see is "simple CRUD app costs $54M" and can never see past this, have next to zero experience dealign with large orgs or their complexity, requirements, etc.
  - amscanne 4y ago
    Your first point was excellent, but now you’re trying to prove too much. QA for a software update on those machines is hardly going to justify the cost. But, along with a million other things, I’m sure it was itemized and subitemized, inflated and billed and that’s how we got into this mess. The fact that doing things in government is much more expensive than people think because you have to go through so many hoops and align so many stakeholders is valid. But in this case, they probably still managed to pay 3-5x more than they should have presumably because of basic incompetence and cronyism. Even though the criticism is imprecise (and you’ve legitimately noted why), it’s still about 80% right.

by belval 4y ago

bel_marinaio 4y ago

so you're saying $250k to build the app once you get specs from the $53,750,000 worth of meetings and planning committees? That sounds much more reasonable and not like $50million worth of tax payer money was stolen /s

belval 4y ago
Wouldn't it be fair to say that once an organization reaches a certain scale it is expected that communications constitute most of its overhead? Something like ArriveCan requires the collaboration of several branches of government and ties in with a system that monitors over 2M arrivals in our country every month. Even all the doers attitude in the world won't change the fact that they won't just give you access to those databases without some pretty extensive review. Just think of the QA needed on the arrival machines at the airport, you can't afford to have them all be out of order because they handle much of the incoming traffic, so the "ArriveCan update" has to be thoroughly tested so that it does not cause any significant downtime. This is the exact opposite of the "move fast and break things" environment. To summarize, I do think that ArriveCan (the program) was a waste of money, but I think the program itself did not waste money. It's the price that comes with building multi-systems applications.
- kennend3 4y ago
  Two things for you. First, I fully agree with everything you said. It wasnt the best use of money, but these people fail to grasp what it takes to build a system integrated with the government. Dealing with any large org requires a lot of back and forth with contracts, legal, etc and the government is worse as they have entire organizations which seem to be created specifically to slow things down. Second, your points are most likely falling on deaf ears. All they see is "simple CRUD app costs $54M" and can never see past this, have next to zero experience dealign with large orgs or their complexity, requirements, etc.
- amscanne 4y ago
  Your first point was excellent, but now you’re trying to prove too much. QA for a software update on those machines is hardly going to justify the cost. But, along with a million other things, I’m sure it was itemized and subitemized, inflated and billed and that’s how we got into this mess. The fact that doing things in government is much more expensive than people think because you have to go through so many hoops and align so many stakeholders is valid. But in this case, they probably still managed to pay 3-5x more than they should have presumably because of basic incompetence and cronyism. Even though the criticism is imprecise (and you’ve legitimately noted why), it’s still about 80% right.