My favourite aspect of this is that the name EPA is coming from a long-since closed chain of stores selling low-quality products, and the teenagers driving these vehicles have for sure never seen one.
I'm very sorry to hear you are not following the recommendations of Swedish authorities, I would strongly recommend you do, because they are contrary to what you claim here, not actually based on Tegnells personal whims or vanity, but based on the many experts employed at the agency, and more importantly adapted to local circumstances in Sweden, such as the current stage of the epidemic, which areas are most affected and so on. As long as the German recommendations are stricter this is of course not that problematic, but it could mean you pay less attention to things that are more urgent to adress in Sweden.
I think this is decidedly unfair critizism, and is based on cherrypicking statements out of context especially regarding timing. The statement about the Alps/Southern europe has for sure been shown to be false in retrospect, but when it was made nobody knew there was an outbreak brewing there, so it was very much based on what was known at the time. I don't think any epidemiologist could have predicted the big outbreak in the alps/northern Italy specifically, and as far as I know, the advise from Tegnell was not different from that of the German authorities at the time. Basing your judgement of Tegnell only on this statement seems very ill-advised.
"nobody knew there was an outbreak brewing there"
How do you know that? It's 2020, a globalized world, heavy exchanges between China and Italy, supercomputers, complex models. Wouldn't one expect from experts in virology that they actually do have means to foresee these scenarios? I personally do.
I was monitoring the coronavirus situation at that time due to personal travel plans. And by mid February I saw a high probability (70 % - 80 %) that we would have significant outbreaks in Europe. Thus, I do expect more from a state epidemiologist than such a display of false certainty.
Apart from that, Tegnell has been wrong in so many regards (how often hasn't he claimed that the outbreak is near a peak, or that there is reason for optimism??! I don't have enough fingers to count that). One would rather have to cherrypick in order to find a specific prediction/claim/statement which turned out to be correct.
In the end, it comes down to what we expect from our leading experts. The experts who in fact decide about many people's lives, in this case. Particularly in a country in which most people follow the state experts' recommendations without any questions asked.
What I expect from leading experts is to give clear opinions and recommendations based on their best assesment of the situation given available data. I also expect them to quickly and humbly adapt to new evidence when it is found, while keeping in mind that new is not always better.
I don't expect experts to always be right, sometimes the best prediction turns out to be incorrect. This does not change the fact that it was the best prediction at the time it was made, given the available data.
I personally feel that this is exactly what Tegnell is doing in public discourse, he is very quick to point out when there are uncertainties and what the nature of it is, but is still able to present clear advise. He has also demonstrated willingness to change when new evidence is found. This is a far cry from what I see many politicians are doing, and I think this is the reason he is enjoying very high trust from the Swedish general population.
Thank you, that’s insightful. An honest (not cynical) follow up question: Would that in consequence mean that in your perspective, an expert who in mid February somewhat speculative had pointed to a possible outbreak in Italy or elsewhere in Europe - based on predictive scenario modeling and probabilities - had done a worse job than one who ruled out such a possibility, by relying exclusively on that moment’s available data (despite virus data lagging a few weeks)?
Swede here. There is lots of disinformation about the Swedish approach to the epidemic. Sweden is following the same basic strategy as other countries, which is to slow down the spread by mitigating actions so the health care system can cope and to protect people in risk groups. The primary difference in Sweden compared to outher countries are:
* Sweden decided early on that an evidence-based approach should be followed. Many policies that are implemented in other countries do not have a lot of evidence showing their effectiveness
* Policy is still primarily dictated by the experts at Folkhälsomyndigheten (where Anders Tegnell is from), and not by politicians. Most actions taken by the Swedish government has been pretexted by "as requested by FHM we have...". State epidemiologist Anders Tegnell is generally seen as "being in charge" of the actions taken, even though technically he has very little legal authority.
* Sweden has a very strong history of indepenent government bodies. In Sweden, there is specific legislation preventing government officials from intervening in specific cases if there is a government authority in charge of the issue. The legal term for this is "ministerstyre", "ministerial rule", and even if an action is not technically in violation of this legislation, it is often used if politicians try to intervene in cases without creating general legislation.
* Being a high-trust society some measures that have been implemented as strict laws in countries other than Sweden are instead communicated as recommendations with a very high compliance rate.
Can you expand upon the policies that have little evidence of effectivness that are being done by other countries. are Swedes also aware why the other countries are doing this?
* Closing borders when we already have a pandemic in all countries. Swedish borders are still open.
* Closing schools when most evidence shows that it is not effective. Swedish schools are still open.
There is of course lots of debate about this in Sweden, some are worried that Sweden is not taking enough decisive action. There is however a wide-spread general support that the more measured approach in Sweden is a good one. Anders Tegnell has become a very popular public figure for his low-key bureaucratic but yet humoristic way of explaining the uncertainties involved and why Sweden is doing things this way.
I believe that the reasons these measures are taken in other countries is because it is a way to show strong political leadership in a time of crisis. I think this approach works less well in Sweden because people have very high trust in experts and (well-managed) authorities.
There is a strong recommendation that high schools and universities handle their education remotely.
This recommendation is followed. The Swedish government cannot order schools to be closed (a law was quickly put in place in the case that this be necessary. I'm not sure that it pertains to universities, though).
I'm on mobile so can dig up links later. But one of the things I see mostly cited in foreign press is the decision to not shut down all schools.
Universities and what would be equivalent to high schools are doing distance learning, but lower grades are mostly open. The department of health has mostly cited two things for not closing lower grades.
1. Little evidence that younger people are the main vectors spreading the virus.
And 2. That if we shut down schools parents would need to stay home (and are encouraged to stay home via our social programs), and that would include a lot of healthcare workers placing our system under too much stress.
Some think that the Swedish state epidemiologist is confusing scientific approach and emergency management.
Children are typically main disease vectors in influenza epidemics. In Covid-19 they seem to be nonsymptomatic. It is not known yet if they are important vectors.
Two policy choices based on the information above:
(1) We don't close the schools until we have the information.
(2) We close the schools to buy time to find out. It's too late to act when the evidence comes in.
There is evidence sufficient to make predictions, and evidence sufficient to prove something beyond doubt. When you say Sweden is taking a more "evidence based" approach, what you're really saying is that they're denying evidence of the first kind. What ever happened to the precautionary principle? Does Sweden still practice that in other areas, relative to the US (for example)? If so, then suddenly getting all conservative about what evidence to consider and calling the result "evidence based" is not only insulting to scientists and policy makers in other countries but a bit hypocritical as well. The arrogance astounds.
* Sweden doesn’t have anything equivalent to martial law.
* the government have in general very weak power to enforce rules that effect the population and all changes to current law must pass through their parliament. (This has though due to the current situation been relaxed in certain areas, a decision that went through the parliament).
No, apparently they had to construct a emergency procedure for this particular incident by going through and change the laws necessary to handle this pandemic. The whole opposition was in on those changes. Though most of it is to be able to circumvent laws like: they wanted to build a field hospital but to do that they need to have a permit which they didn’t have/get when they needed it. That delayed things. This is a just an example. Also these those changes have a time limit, but could be extended through parliament.
I'm sure most countries would like to think they are taking an evidence-based approach. A lot of the evidence is not very clear though, there are many unknowns. If you think you understand all the mechanisms, just listen to one of the prominent virologists. It's not any different to other areas: the more you learn, the more you realize how much you don't know. If anyone claims otherwise, they probably don't deserve your trust.
In such an environment, it's impossible for politicians to take decisions that are good in all aspects, but at least regarding health, assuming that's a priority, they can either err on the cautious or on the risky side.
Compared to most other countries, Sweden is clearly taking the latter approach.
That's not at all clear, and won't be clear for a long time. When it comes to optimizing long term public health there are more factors to consider than just coronavirus.
Taking the "evidence-based approach" in the case of Sweden has so far meant being late (in comparison to other countries) with every counter-measure.
Yes, that might mean avoiding a counter-measure which would have been ineffective. But it also means having missed out on the positive effects of a counter-measure which - based on later evidence - proved effective.
I'd argue that the evidence-based approach is great when enough evidence is available. When that is not the case however - as with the new coronavirus - it certainly is imaginable that the evidence-based approach actually is a weakness, not a strength.
It’s really not about “evidence-based approach” per se, it’s more because of the notion of “expert authority”. The notion is that the government are heavily relying on the “expert authority” related to the current situation, in this case, FHM. This means that they follow what the experts are saying, which is by design not as mere information or guidance. This happens pretty much in any government body, like Swedish prison and probation services. The prime minister have very little to say about details of what they do (“minister rule”). The decisions are based on science rather than pure politics because they rely on that current government body for that. That does mean that they are late, which to I agree isn’t optimal but a trade off. The other reason is because the government have been slowed down due to the fact that they have to follow the current law. There’s no way for them to circumvent that as it is.
This has nothing to do with the government.
FHM could have made suggestions/recommendations earlier than when they did it, when it comes to physical distancing, maximum group size, behavior in restaurants, public transport, care homes, or even the usage of masks in public (which still isn't even close to be encouraged). I've been following the situation in various countries, and it became strikingly evident how FHM was late with pretty much every single eventually recommended counter-measure - presumably because they didn't see evidence yet (or, alternative explanation: On purpose, as part of the high-risk herd immunity strategy, which does obviously require many people to get infected).
It's a beautiful thing that recommendations (instead of law enforcement on the streets threatening with fines) work good enough for large parts of the Swedish population. But then everything still comes down to the quality and effects of the recommendations.
An interesting point regarding the "ministerial rule" is that the other Scandinavian countries don't have the same restriction on politicians, at least not in law/constitution.
This is an excellent example, the chief epidemiologist in Denmark had a clash with the Danish PM about this, because evidence does not support closing borders as an effective policy when we already have a pandemic. In Denmark, the measure was intruduced despite this, presumably because the government wanted to demonstrate forceful action. Sweden still has open borders (not that it makes much difference since most other countries have closed them :) ).
I have only encountered public SQL APIs in the wild once, and at that time it made me very happy. I was building a scraper for a manufacturer of bicycle parts that shall remain unnamed. Scraping the whole database using a single SQL query was way better than hacking XPaths. I am not sure if their DBA appreciated it as much as I did.
The tagline of Datasette is "A tool for exploring and publishing data", to me it makes a lot of sense that SQL (with appropriate restrictions) is a suitable API for the kinds of things you want to do with datasets. This is not really the typical web API usecase, you normally have much more specialized and restricted use scenarios.
While I have also seen many poorly secured GraphQL API:s, this seems like an unfair criticism. To me GraphQL is mostly an alternative to REST, which also has no authorization or authentication by default, that is a orthogonal concern, but you can hardly blame REST for that.
I responded in a twitter thread[1], unrolled here here:
As all good things in life, and programming, this is a tradeoff. GraphQL is better when what you are requesting is best expressed as a tree (or a "graph", though only the DAG variety). This is not always the case, but it very often is when building API:s for production use.
Of course, you can express tree structures in table form, but it is not very convenient for clients to consume. In particular if your client is rendering nested component views, what you want is very often something hierarchical.
Another aspect of GraphQL that is better for us production people is that the performance is more predictable, exactly because the language is more restricted. You can't just join in all the things, or select billions of rows by accident. The schema dictates what is allowed.
Of course, again, it is possible to restrict this in SQL, just configure your schemas, limits etc appropriately, but SQL is anything-goes by default, whereas GraphQL is nothing is allowed by default. Whitelist vs Blacklist.
This said, as a language, SQL is clearly superior. It is the most (only?) successful 4GL (declarative) language. I wish more languages were this well-designed, and that there would be more language innovation in this direction.
The way I see it, GraphQL is a DSL for flexibly requesting hierarchical data from API:s in JSON format, optimized for complex evolving API:s. SQL is a full-fledged generic language for relational data transformation. They have different niches, but SQL has a much bigger one.
> a DSL for flexibly requesting hierarchical data from API:s in JSON format, optimized for complex evolving API:s
Perhaps the same goal could be achieved with more flexibility by issuing CONSTRUCT queries to a SPARQL endpoint (CONSTRUCT requests a custom RDF graph as opposed to a set of variable bindings ala SELECT) and obtaining results in JSON-LD format.
Exactly. I was always interested in all this semantic web stuff and as I have time now, I'm experimenting with Apache Jena. It's super neat! SPARQL endpoints use HTTP per default. "API for free" so to say. Plus you can even use a binary format for receiving the data. Querying the database is three lines of code in Kotlin. Fuseki (the SPARQL Server in Jena) also supports crazy stuff like "recursively traverse this graph by following the specified type of link until you can't go any further". I'm pretty sure this is not possible in GraphQL.
> You can't just join in all the things, or select billions of rows by accident.
How does GraphQL prevent you from selecting billions of things? I'm not too familiar, but if the answer is something like "you can limit to N results", SQL does have a LIMIT clause, so I'm unsure what is being meaningfully distinguished.
A good implementation will calculate the estimated "cost" of a given query before executing it. This is quite tricky to do well, and I doubt many people do it. Once you know the cost of a given query, it's trivial to reject it if it's too expensive.
A much better alternative is to only allow specific queries in production. These persisted queries can then be assigned an ID, and making a request is a matter of sending the ID with any variables.
It's very different. The main difference is that you define the capabilities on the back end through your GraphQL schema once, then clients generate the query ids through an automated process in development.
In the old way, someone would have to write backend code to build and maintain each individual endpoint. With GraphQL, you only maintain the schema on the backend.
The author makes a benchmark that compares a go program that explicitly buffers many "hello world" strings, and compares this to an assembly program that does not do this, then goes on to argue that this shows that high level languages can be faster. This can certainly be true, but this particular example is not a great way to show it, because the same optimization can be done in an obvious way in assembly, (just allocate enough space with brk or mmap2, write your strings and send the right memory address to ). This would of course be harder to do in assembly than in Go, but it is still quite straightforward to see what is the optimal way to do it. Furthermore, this would probably still outperform the optimized Go version by a wide margin.
A better example could be where a compiler picks an obscure but faster parallelization operation, or unrolled a loop appropriately in a way that is both faster and unlikely to be written by a competent human, or a complex memory management scenario etc etc.
I think this is not the point of the original article though. I think we all understand that abstractions can in theory bring great benefits, but we do need to scrutinize the cost they add. The hello world examples shows that even with the simplest program we can imagine, the result is extremely far from optimal in popular programming environments. If this is the case, why should we assume that these same compilers are doing an excellent job in situations that are actually hard?
He called this out, saying that the added complexity overhead of Go allowed the abstraction to be done much simpler than it would be in assembly. This is just tradeoff that is worthwhile in most cases.
The gripes about boilerplate overhead in Hello World miss the point that the runtimes involved are themselves making tradeoff about what to optimize for. Go explicitly trades off ultra-efficient binary size for ultra-fast compilation and mostly-static artifacts. Go is not designed to make the most efficient possible Hello World binary, nor should we want it to be. The fact that you can optimize Hello World better by hand than the Go compiler does tells us nothing interesting. How well can you optimize Docker, Consul, or Kubernetes by hand?
