The COVIDSafe app was a failure and waste of tax payer money. Everyone knows it. The federal government has spent the last year and a half trying to change the narrative, mostly unsuccessfully.
State governments developed their own QR code based systems which seem to work.. up to a limit. Those systems worked to stop multiple outbreaks - e.g. NSW contract traced their way out of several outbreaks without significant restrictions.
The question I felt was unanswered by the article is have overseas apps actually delivered significantly better results than that?
The actual big difference between Australia and most other developed nations seems to be that we are behind on our vaccine rollout and more willing/able to restrict individual movement, especially across state borders.
Our vaccine rollout should catch up over the next month or so. NSW is approaching 90% first doses for over 16’s. VIC is approaching 80%. The dose spacing for AZ is particularly long, and at least VIC is using 6 week spacing for Pfizer so Australia tends to lag more on 2nd dose figures but that gap will close over the coming month.
On a side note, all the information the author wishes was in the COVIDSafe app is also readily available on state government websites - e.g. coronavirus.vic.gov.au. Personally I prefer having this information on a website than in an app.
> The question I felt was unanswered by the article is have overseas apps actually delivered significantly better results than that?
My country rolled out an app quite early last year. It worked well, didn't drain battery life, didn't need you to "sign up" or provide any personal details. The problem was pretty much nobody installed it, so it was basically useless.
They also tried strict contact tracing, requiring people to register whenever they go into a bar or restaurant, but that also didn't work (most outbreaks were from meeting family or in the workplace).
Now my country has given up on contact tracing and moved to accepting the virus will be here forever, and is trying to encourage more people to be vaccinated.
My question was have contract tracing apps overseas performed significantly better than the QR based systems used in Australia (instead of COVIDSafe). Neither of these links seem to show that.
The first is a criticism of COVIDSafe. I agree it was crap, but have Bluetooth based contract tracing actually worked well to prevent outbreaks anywhere? Is it actually better than QR code based systems?
The second paper you link at least attempts to model this but it’s just that, conclusions drawn from a simulation. In practice there have been many more cases and deaths in Washington State alone than all of Australia, and that’s without adjusting for the much larger population in Australia.
The thing that pissed me off the most was that they took gpl licensed software and slapped more terms and conditions on it. Some suit didn't read the original license and thought it was just free code.
The thing about Australia is we love to walk eyes wide open into failure. We watch policies being implemented in other countries with real world data to learn from, and will then just unabashedly ignore it and do it all ourselves to achieve the exact same outcomes.
This is across technology, politics, climate, business, the works. It's ingrained culturally and will never change.
> Australia is a lucky country run mainly by second rate people who share its luck. It lives on other people's ideas, and, although its ordinary people are adaptable, most of its leaders (in all fields) so lack curiosity about the events that surround them that they are often taken by surprise.
Australia doesn't have to be competent, it can just dig up its vast mineral wealth, share the riches among its very small population, and think that skill and competence had something to do with it.
By comparison, I'm amazed at places like South Korea, Japan and Israel that get by on sheer innovation and technology.
My favourite is the myki fiasco. We could have had Oyster down here, but apparently Melbourne is too special and needed a full custom solution.
Cue billions of blowouts and late deliveries and the thing still doesn't work well even after all these years. Sydney can use their nfc cards to travel, we still need a myki or Google pay, of which GP takes forever to read your card.
One of the biggest failures of COVIDSafe in Australia was that it was so heavily bashed by everyone on it's initial release. Anyone who was anyone binned it as soon as it was released.
With an app like COVIDSafe, where you require to have mass adoption of the app to work (there is no use having 10% of phones operating in the wild with the app installed, you require more like 80-90%, ideally every phone running the app.)
Instead, every techie I spoke to either hated it wasn't open source; or used the fact that the app was based on Singapore's app, but was not subsequently open sourced itself.
I feel that if the powers that be had changed the way that the source was developed, worked with the public a little closer that maybe these issues would not have existed.
But sadly, the damage for COVIDSafe was already done in the first few weeks. There was no resuscitation possible.
I liken it to how the AstraZeneca vaccine was dissed in the media. A small chance of clots was heralded as a major issue, which lead to a extremely slow uptake of a vaccine that is for the most part perfectly fine.
Add to that the fact that privacy was never really enacted for the use of the data gathered by COVIDSafe (and other similar contact tracing tools, for example in Sydney, NSW https://www.smh.com.au/politics/federal/breach-of-trust-poli... ) and you have a perfect storm of an app that no-one really keeps installed.
> Instead, every techie I spoke to either hated it wasn't open source; or used the fact that the app was based on Singapore's app, but was not subsequently open sourced itself.
Singaporean here. Our app has never really been open sourced either, contrary to claims from our previous Smart Nation Minister [1].
The only things that were open sourced were the BlueTrace protocol, which Australia uses, and some really outdated reference app implementations.
I think Singapore has given up on the Bluetooth tracing approach as well, and has pivoted to requiring everyone to check in at various buildings and eating establishments [2].
Then again, Singapore, despite being one of the most vaccinated countries in the world, imposed further restrictions on eating out that take effect today. [3] Hurray. :) /s
As soon as it was released, there was actually pretty good support for it. It was shortly after - when people started realising it didn't actually work as advertised, that it lost all credibility as the government insisted that it did. But you're right, from there it didn't recover.
Most of the bugs affecting how it worked (especially on iPhone) , security and privacy were found by the community, including critical bugs that stopped the app working entirely that weren't fixed until August. All while the government kept saying there was nothing wrong with it.
So it's hard to argue that the bashing wasn't warranted. The app really did not work.
I really want our government to do technology better. Vaccine passports are already off to a poor start - but I really hope that the community has enough influence that we can start to make a difference.
Don’t forget the detail that the app just flat out didn’t work on iOS and the governments advice was to open the app, turn off auto screen locking and leave the app running, screen on in your pocket.
Shortly after, Android and iOS released a proper API for doing this which worked in the background but the government did not use it.
The distrust I had was also in part because it was rushed out the door and based upon Singapores's app, which used deprecated APIs. This, while Google and Apple were promising new fit-for-purpose and interworking APIs within a few weeks.
At the risk of getting political I would also say that this government has a track record of making big announcements, throwing some money to it's friends (donors, boosters) to do the work, not adequately spec'ing or managing the project, not caring if the expected outcomes are achieved and money was wasted, then blaming someone else should this become an issue in the media. This project seems to be just another example of that process.
Yeah, but the amazing thing about software is that updates can arrive.
It shouldn't matter if the APIs in use are deprecated, or a new API that is designed for the use case arrives in three weeks after the application was initially released (the APIs were not ready when COVIDSafe was made available to the public).
Get the app - even if it's imperfect - in the hands of users. Admit fault for the issues that do exist, and commit to a timeline to remedy the issues and update APIs.
> But sadly, the damage for COVIDSafe was already done in the first few weeks. There was no resuscitation possible.
I (the aurhor) feel similar as well. Interfacing with the media was (and continues) to be the only way problems in Australian IT projects are resolved. The Australian Government needs a Vulnerability Disclosure Policy. I would love nothing more than to be able to deal directly with engineers and have sound knowledge that critical problems will be fixed vs having to interface with the media to get problems resolved.
Australia should get something akin to the USDS (https://www.usds.gov/) that acts as a third party contractor/advisor to the federal and state departments.
Have a rolling panel built from Australians that work in the industry to publicly comment on tech projects run by the Government.
I think that in the big picture, COVIDSafe failing over these issues was a good thing: Hopefully, it will encourage future governments to take a more transparent approach and respect people's privacy, because (maybe for the first time in recent history), there's a concrete, demonstrated downside to trying to do the opposite.
> …there’s a concrete, demonstrated downside to trying to do the opposite.
I feel this way about so much of the things that have happened with COVID.
We have such an opportunity to learn so much here but looking at how we’re handling the outbreaks in real time doesn’t give me a lot of hope. At this point we know which actions or lack of actions will lead to an outbreak and those communities still do little to mitigate.
Combine this with how we continually do the wrong things on so many other issues and then cry crocodile tears when we get the same results, I’m just not sure we’ll behave differently next time.
> I liken it to how the AstraZeneca vaccine was dissed in the media. A small chance of clots was heralded as a major issue, which lead to a extremely slow uptake of a vaccine that is for the most part perfectly fine.
It's less safe, there were some deaths, and also it protects less than alternatives like Pfizer and Moderna. So there is no point in that vaccine, in my country they first banned it for people under 60, and now they are no longer using it, choosing only to use Pfizer and Moderna vaccines (that uses the better mRNA technology).
The problem is that we had put quite a few of our eggs in the AZ basket to begin with, so the loss of confidence seriously delayed our vaccine rollout. And in the end we missed out, by a couple of months, on having good vaccine coverage before the delta strain got away from us.
On the other hand, I think some of the anti-anti-AZ stuff is over the top. The increased TTS risk is by now clearly real, the non-fatal outcomes are often pretty bad, and until recently the fatality rate was expected to be very high.
In the pre-delta period, when our covid prevalence was extremely low and it was reasonable to hope that state would last long enough to see us through to the Pfizer rollout, waiting for Pfizer was probably a sensible decision ex ante for many people, even taking into account a reasonable level of pro-sociality (and even without knowing that there would be a significant difference in effectiveness).
An official or media-led decision to play down the risk would have caused some preventable deaths (and serious harm to a larger number) among people who, from a self-interested perspective, probably shouldn't have taken that vaccine in the first place. With hindsight, those deaths are outweighed by the covid deaths we're now seeing; but that wasn't obvious pre-delta, and in any case the authorities would have burned a lot of trust and given a free kick to the 'sceptics' and conspiracy theorists if they had hidden or misleadingly played down the risk.
Digital contact tracing was a failure in practically every country, and there is a techincal problem: all these apps uses Bluetooth to measure the distance, since GPS impacts privacy too much, and also battery life on the device, and doesn't work well in closed spaces. But you cannot have enought resolution with bluetooth to distinguish between 1 meter and 2 meters (the distance that is considered safe), you can have precision in tens of meters. Also you get a ton of false positives, for example caused by people that passes nearby buy you don't came in contact with (e.g. people in the car near you when you are stopped in the traffic). Finally the idea that people wants to collaborate is pure utopia, yes there are people that wants, but the majority of people don't want to install an app that if they possibly came in contact with one people that have the virus (something that if we all install the app was certain, at least in the last winter) will force you to isolate for 14 days, that includes not going to work without getting money for sickness since you are techincailly not ill. Basically it's a failure the whole concept of digital contact tracing, not the implementations.
There is no reliable scientific evidence that 2m is "safe". It's an arbitrary number. The WHO recommends 1m. It's really more a matter of airflow than distance.
I have no comment on the rest of the-fine-article, bu this statement is incorrect:
Our two largest cities (Sydney and Melbourne) right now are in a continued state of lockdown, with the various states pursuing a COVID19 eradication strategy which has caused immense social and financial damage to our society.
Australia's two most populous states (New South Wales and Victoria, the main jurisdictions presently under lock-down) are not pursuing any COVID19 eradication strategy; that boat sailed months ago when the NSW government failed to lock-down sufficiently when the Delta strain hit, thus infecting many of the other states and New Zealand through that lack of action.
The evident strategy being pursued by these two states is one of protecting the health system by ensuring it isn't overrun by COVID cases, while also giving as many people as possible the chance to vaccinate.
Once 80% of the populace is vaccinated (and I concur with the author of the piece here that that should have been achieved months ago), then most of the lock-down measures will cease.
It also means that anyone who opts not to vaccinate (and kids under 12, and the immuno-compromised who cannot vaccinate) will have to fend for themselves.
> Australia right now is in a continued state of lockdown
_Some_ of Australia is in a state of lockdown. This includes Sydney and Melbourne, Australia's two largest cities, but the majority of states are currently not in lockdown.
Don't get me wrong: the response to the pandemic from Australia's federal government has been dreadful, and I'm not a huge fan of the police-oriented approach some states have used when they've been locked down either. But I think the first few paragraphs are alarmist and exaggerated, and that discredits the excellent discussion of the technical failures later in the article.
>but the majority of states are currently not in lockdown
This is needlessly misleading. The majority of the population is in perpetual lockdown, especially Melbourne which has been in lockdown for the longest period of time globally under much stricter rules than Europe and the US.
"The majority of states" are not the majority of people.
Make no mistake, Australia is locked down. With recent changes, foreign citizens are prevented from returning to their home countries without approval. Not only are citizens trapped, but foreigners who had the misfortune of being here are now trapped as well.
You may claim there are exemptions, but I've seen piles of evidence that foreigners are being denied exemptions to leave even in extraordinary circumstances (leaving permanently to see dying family, for example).
For anyone not from Australia, there is more to the story than "police officers assaulting unarmed civilians" just because they were not following covid regulations. That being said, the footage where that was taken was in response to, I agree, a series of non-optimally handled situations.
It's a shame the article has such a rambling style and seems to take forever to get to the main point, which in the end isn't sufficiently fleshed out. I want to see substantiated the claim that any of these apps anywhere in the world were able to make a big impact on contact tracing. I am not aware of that being the case.
The government report into it is a scandal and should be torn apart. The fact it is almost entirely redacted except for one mildly complimentary sentence is outrageous. But I'm more interested in the simple question of whether it could work or not. There is still time even now in which it could make an impact if it is possible.
Readit News
State governments developed their own QR code based systems which seem to work.. up to a limit. Those systems worked to stop multiple outbreaks - e.g. NSW contract traced their way out of several outbreaks without significant restrictions.
The question I felt was unanswered by the article is have overseas apps actually delivered significantly better results than that?
The actual big difference between Australia and most other developed nations seems to be that we are behind on our vaccine rollout and more willing/able to restrict individual movement, especially across state borders.
Our vaccine rollout should catch up over the next month or so. NSW is approaching 90% first doses for over 16’s. VIC is approaching 80%. The dose spacing for AZ is particularly long, and at least VIC is using 6 week spacing for Pfizer so Australia tends to lag more on 2nd dose figures but that gap will close over the coming month.
On a side note, all the information the author wishes was in the COVIDSafe app is also readily available on state government websites - e.g. coronavirus.vic.gov.au. Personally I prefer having this information on a website than in an app.
My country rolled out an app quite early last year. It worked well, didn't drain battery life, didn't need you to "sign up" or provide any personal details. The problem was pretty much nobody installed it, so it was basically useless.
They also tried strict contact tracing, requiring people to register whenever they go into a bar or restaurant, but that also didn't work (most outbreaks were from meeting family or in the workplace).
Now my country has given up on contact tracing and moved to accepting the virus will be here forever, and is trying to encourage more people to be vaccinated.
1. https://raw.githubusercontent.com/AusOpenTech/COVIDSafe-Repo...
2. https://www.medrxiv.org/content/10.1101/2021.06.04.21257951v...
The first is a criticism of COVIDSafe. I agree it was crap, but have Bluetooth based contract tracing actually worked well to prevent outbreaks anywhere? Is it actually better than QR code based systems?
The second paper you link at least attempts to model this but it’s just that, conclusions drawn from a simulation. In practice there have been many more cases and deaths in Washington State alone than all of Australia, and that’s without adjusting for the much larger population in Australia.
Deleted Comment
This is across technology, politics, climate, business, the works. It's ingrained culturally and will never change.
https://en.wikipedia.org/wiki/The_Lucky_Country
By comparison, I'm amazed at places like South Korea, Japan and Israel that get by on sheer innovation and technology.
Cue billions of blowouts and late deliveries and the thing still doesn't work well even after all these years. Sydney can use their nfc cards to travel, we still need a myki or Google pay, of which GP takes forever to read your card.
With an app like COVIDSafe, where you require to have mass adoption of the app to work (there is no use having 10% of phones operating in the wild with the app installed, you require more like 80-90%, ideally every phone running the app.)
Instead, every techie I spoke to either hated it wasn't open source; or used the fact that the app was based on Singapore's app, but was not subsequently open sourced itself.
I feel that if the powers that be had changed the way that the source was developed, worked with the public a little closer that maybe these issues would not have existed.
But sadly, the damage for COVIDSafe was already done in the first few weeks. There was no resuscitation possible.
I liken it to how the AstraZeneca vaccine was dissed in the media. A small chance of clots was heralded as a major issue, which lead to a extremely slow uptake of a vaccine that is for the most part perfectly fine.
Add to that the fact that privacy was never really enacted for the use of the data gathered by COVIDSafe (and other similar contact tracing tools, for example in Sydney, NSW https://www.smh.com.au/politics/federal/breach-of-trust-poli... ) and you have a perfect storm of an app that no-one really keeps installed.
COVIDSafe was dead on arrival.
Singaporean here. Our app has never really been open sourced either, contrary to claims from our previous Smart Nation Minister [1].
The only things that were open sourced were the BlueTrace protocol, which Australia uses, and some really outdated reference app implementations.
I think Singapore has given up on the Bluetooth tracing approach as well, and has pivoted to requiring everyone to check in at various buildings and eating establishments [2].
Then again, Singapore, despite being one of the most vaccinated countries in the world, imposed further restrictions on eating out that take effect today. [3] Hurray. :) /s
[1] https://www.smartnation.gov.sg/whats-new/speeches/clarificat...
[2] https://en.wikipedia.org/wiki/SafeEntry
[3] https://www.straitstimes.com/singapore/health/dine-in-social...
Most of the bugs affecting how it worked (especially on iPhone) , security and privacy were found by the community, including critical bugs that stopped the app working entirely that weren't fixed until August. All while the government kept saying there was nothing wrong with it.
So it's hard to argue that the bashing wasn't warranted. The app really did not work.
I really want our government to do technology better. Vaccine passports are already off to a poor start - but I really hope that the community has enough influence that we can start to make a difference.
Shortly after, Android and iOS released a proper API for doing this which worked in the background but the government did not use it.
At the risk of getting political I would also say that this government has a track record of making big announcements, throwing some money to it's friends (donors, boosters) to do the work, not adequately spec'ing or managing the project, not caring if the expected outcomes are achieved and money was wasted, then blaming someone else should this become an issue in the media. This project seems to be just another example of that process.
It shouldn't matter if the APIs in use are deprecated, or a new API that is designed for the use case arrives in three weeks after the application was initially released (the APIs were not ready when COVIDSafe was made available to the public).
Get the app - even if it's imperfect - in the hands of users. Admit fault for the issues that do exist, and commit to a timeline to remedy the issues and update APIs.
I (the aurhor) feel similar as well. Interfacing with the media was (and continues) to be the only way problems in Australian IT projects are resolved. The Australian Government needs a Vulnerability Disclosure Policy. I would love nothing more than to be able to deal directly with engineers and have sound knowledge that critical problems will be fixed vs having to interface with the media to get problems resolved.
Have a rolling panel built from Australians that work in the industry to publicly comment on tech projects run by the Government.
I feel this way about so much of the things that have happened with COVID.
We have such an opportunity to learn so much here but looking at how we’re handling the outbreaks in real time doesn’t give me a lot of hope. At this point we know which actions or lack of actions will lead to an outbreak and those communities still do little to mitigate.
Combine this with how we continually do the wrong things on so many other issues and then cry crocodile tears when we get the same results, I’m just not sure we’ll behave differently next time.
It's less safe, there were some deaths, and also it protects less than alternatives like Pfizer and Moderna. So there is no point in that vaccine, in my country they first banned it for people under 60, and now they are no longer using it, choosing only to use Pfizer and Moderna vaccines (that uses the better mRNA technology).
On the other hand, I think some of the anti-anti-AZ stuff is over the top. The increased TTS risk is by now clearly real, the non-fatal outcomes are often pretty bad, and until recently the fatality rate was expected to be very high.
In the pre-delta period, when our covid prevalence was extremely low and it was reasonable to hope that state would last long enough to see us through to the Pfizer rollout, waiting for Pfizer was probably a sensible decision ex ante for many people, even taking into account a reasonable level of pro-sociality (and even without knowing that there would be a significant difference in effectiveness).
An official or media-led decision to play down the risk would have caused some preventable deaths (and serious harm to a larger number) among people who, from a self-interested perspective, probably shouldn't have taken that vaccine in the first place. With hindsight, those deaths are outweighed by the covid deaths we're now seeing; but that wasn't obvious pre-delta, and in any case the authorities would have burned a lot of trust and given a free kick to the 'sceptics' and conspiracy theorists if they had hidden or misleadingly played down the risk.
Dead Comment
https://www.who.int/westernpacific/emergencies/covid-19/info...
The evident strategy being pursued by these two states is one of protecting the health system by ensuring it isn't overrun by COVID cases, while also giving as many people as possible the chance to vaccinate.
Once 80% of the populace is vaccinated (and I concur with the author of the piece here that that should have been achieved months ago), then most of the lock-down measures will cease.
It also means that anyone who opts not to vaccinate (and kids under 12, and the immuno-compromised who cannot vaccinate) will have to fend for themselves.
_Some_ of Australia is in a state of lockdown. This includes Sydney and Melbourne, Australia's two largest cities, but the majority of states are currently not in lockdown.
Don't get me wrong: the response to the pandemic from Australia's federal government has been dreadful, and I'm not a huge fan of the police-oriented approach some states have used when they've been locked down either. But I think the first few paragraphs are alarmist and exaggerated, and that discredits the excellent discussion of the technical failures later in the article.
This is needlessly misleading. The majority of the population is in perpetual lockdown, especially Melbourne which has been in lockdown for the longest period of time globally under much stricter rules than Europe and the US.
"The majority of states" are not the majority of people.
Make no mistake, Australia is locked down. With recent changes, foreign citizens are prevented from returning to their home countries without approval. Not only are citizens trapped, but foreigners who had the misfortune of being here are now trapped as well.
You may claim there are exemptions, but I've seen piles of evidence that foreigners are being denied exemptions to leave even in extraordinary circumstances (leaving permanently to see dying family, for example).
The government report into it is a scandal and should be torn apart. The fact it is almost entirely redacted except for one mildly complimentary sentence is outrageous. But I'm more interested in the simple question of whether it could work or not. There is still time even now in which it could make an impact if it is possible.