DocSavage

It removes all default browser styling for hyperlinks. It wasn't clear from the docs that it would do that so I'm glad the fix came up: https://stackoverflow.com/questions/63883580/tailwind-css-ho...

I was trying Tailwind out a few weeks ago for a very simple static web page. I'm close to beginner with slight knowledge of CSS and HTML. It was surprising to me that all the styling disappeared from <a> tags, and it took some Googling just to see how to work around that. I wound up just going with Spectre to get an acceptable solution. The Tailwind docs seem more for its target audience and not pure beginners. Is there some resource to learn Tailwind in the manner you are suggesting?

They are also using a pure Go version of leveldb (github.com/syndtr/goleveldb) and not the more battle-hardened Google leveldb, RocksDB implementations, or even the dgraph-io Badger WiscKey-style approach.

C++ is still used extensively, even for completely new projects, in computer graphics, databases, etc. Relatively new projects I’ve seen it used: TileDB, tensorstore, core Apache Arrow, etc. CUDA programming is C++ at the bleeding edge. I’m in the process of updating my C++ to C++17 and there’s definitely been a lot of changes over time in the language. I would agree that for many, they’d go with python and let others accelerate that python with underlying C/C++ they wouldn’t have to see.

The police unions and particularly this administration have been actively undermining strong independent oversight.

You might want to look at some of the work we're doing in "git for connectomics": http://dvid.io

There will be a new backend released soon that should really improve our ability to transfer and backup versioned data.

That’s not how the contagiousness works. Most reports have average of 5 days from infection to symptoms, but many people are asymptomatic or have mild symptoms. At least one study says viral shedding is close to peak the day before symptoms and the time course of shedding goes down over time. Saying you are probably not contagious just 5 days after infection is misinformation, especially since incubation can range from 2 to 14 days.

Looks like a cool device for high-virus areas like front-line health care workers.

Screening tests need to be highly accurate OR you need to be fully aware and manage the fallout from false negatives, which is the point I'm making. First, screening tests have costs, that's why many diseases don't have screening tests. It's what's drummed into us in medical school. Sometimes the medical community backs away from a screening test like has been done in many areas with prostate exams. The cost/benefit isn't worth it. One cost in the COVID case is use of PPE, health care worker time, and test availability hence why CA is backing away from broad testing to triage tests to in-hospital use. Another cost is the possible grouping of symptomatic patients together as they wait for tests. Sure, if rollout can be perfect this isn't a problem, but medicine is about what happens in the real world where delivery is not perfect. By the way, NYC and CA medical community is saying the cost/benefit isn't good enough for broad COVID screening at this time. I'm sure you know better because you know the math of disease spread and are completely unable to grok other reasons why'd they do that.

Second and less talked about, there are consequences of a false negative population which is NOT handled appropriately -- like not letting people know if they test negative they STILL should stay away from high-risk population or take precautions like you still could be infected. That directive has been circulated a little but most people I've seen don't know that. The news reporters don't ask Pence about serial tests. They seem satisfied he had one negative test.

Once again, and I don't know how many times I have to say this, I'm not against widespread testing of the current RT-PCR swab, particularly if we had lots of tests, didn't have PPE shortages in the area, and guidance was given to all those receiving results. Your contention that accuracy doesn't matter is incorrect. It matters because false negatives matter and it's great to do testing to decrease the spread, which is pretty obvious, but you absolutely need additional management depending on accuracy. Like not setting the standard of working with high-risk patients as a single negative test. They need to be tested serially and hopefully get a more accurate test for nursing home workers in future. Like not assuming Pence isn't spreading disease because of a single negative test. He could be one of the many asymptomatic infected. This has been my point for this entire thread. I'll leave you to argue something different.

> It's like talking to a brick wall.

Yes, it certainly is. Did I ever say we shouldn't test? Did I ever argue testing couldn't be improved? Suggest you argue points that were actually put forth instead of using straw men arguments. Re-read the last line of my original post. I'm all in favor of trying to find positive cases.

So after you try to belittle me about not understanding 4^x vs 1^x, you are now talking about future efforts, hopes how an entire medical community can be retrained, etc. Let's stay focused on the real world and not some hypothetical past where we were like South Korea. None of your comments apply to a 1^x spread in the current system since we aren't testing many people, and in hot spots, we aren't testing hardly anyone outside those in hospitals.

Here's where we agree. We agree that the only way this can be contained is massive lockdowns, and I believe this particularly true in the face of unreliable and unavailable testing. I also think it has to be coordinated across the country to prevent constant reseeding. We agree that testing everyone would've been good, particularly in the containment phase, but that ship has sailed.

My original point is that under the current system where we are occasionally testing the public at large, and also using swab RT-PCR just after suspected infected contact (way before symptoms like Pence, etc), the variable sensitivity is troubling.

There's a lot to be said about WH-level leaders with the authority to corral and force action down the chain. It's bad enough that the CDC and FDA screwed up, but there was no urgency from the top-down that solutions had to be rolled out in a timely matter or alternatives had to be explored even in parallel. You see the difference now, when there's a sense of urgency, and we are willing to break rules, engage lots of different groups to work on efforts in parallel, etc. None of that happens without urgency from the top.

Since you are being snarky:

Are you aware that the sensitivity quoted was mostly in symptomatic and hospitalized patients and not those in the very earliest stages who could also be shedding less virus?

Are you aware that each false negative can infect N others, some of whom might have also been previously tested, and also be greenlit to deal with high-risk populations like nursing homes?

Are you aware that your 4^x vs 1^x comparison assumes that in that chain everyone is tested, which is not even close to the case now.

I'm also surprised that there isn't more talk about the sensitivity of the standard nasopharyngeal swab + RT-PCR. When the FDA approved it, they showed it's extremely accurate... if you use swabs saturated with the virus at various concentrations. Then I read about the poor (70-80%) sensitivity in the real-world, which isn't suprising considering you (1) have variably expert technique in positioning swabs into uncomfortable body spaces, and (2) apply it to people at various stages of the disease with low viral shedding initially and also at end stages. Apparently you can swab two different areas and may raise your sensitivity 10% to 85%.

So if you are "negative" with say a 80% sensitivity, that leaves a lot of room for actually infected people to think they are in the clear. There's a reason why screening tests need to be highly accurate. I'm all in favor of at least recognizing the positives and quarantining them, but with a high R0 disease, the false negatives could be a problem.

A good synopsis of current understanding (as of last Friday) was a UCSF Medicine Grand Rounds broadcast. I've bookmarked the most relevant slide to our discussion in particular: https://youtu.be/bt-BzEve46Y?t=2349

Significant lung and myocardial injury and papers have honed in on ARDS as a real problem.

Hence my comment, perhaps positioned less than optimally in my statement, about preventing the cascade to next stage. The degree to which these drugs address the various issues is still in flux. In the UCSF grand rounds recently, there was some question as to the role of immunomodulators.

Stage of disease also is important. In the first symptomatic stage, makes sense to try antivirals and HCQ to reduce or eliminate the virus, and prevent a cascade into the next stage. The problematic second stage - ARDS, a subset of patients who really go critical, seems to call for different approach because at that time, it’s the cytokine storm and our own immune reaction killing us.

Regarding randomization, the HCQ + Azithromycin study showed good results despite the control group being 14 years younger. Sex was about equal so no male/female imbalance. It’s clear that older patients have worse outcomes, but this study measures viral load which also can change depending on the stage of the disease.

On the elderly infections: https://www.wired.com/story/why-the-coronavirus-hit-italy-so...

On antibiotics, aside from secondary infections, there’s a paper out talking about azithromycin: “Our preliminary results also suggest a synergistic effect of the combination of hydroxychloroquine and azithromycin. Azithromycin has been shown to be active in vitro against Zika and Ebola viruses [20-22] and to prevent severe respiratory tract infections when administrated to patients suffering viral infection [23].”

Not sure of they used azithromycin but would’ve been good to see that verified.

I think we're seeing a small case study of this in the nursing home in Washington. Out of 120 residents, 13 are dead and have been identified with COVID-19. There were 11 more who died at the nursing home since Feb 19 and haven't been tested, but they said on average 3-6 residents die per month. I think the nursing home scenario is probably a worst case given both age and underlying health concerns (unlike elderly still living in their own home or with their families). Given the Seattle area medical infrastructure, this fatality rate can't be explained by overloading of our facilities.

I really disliked Fauci's comments that "the risk is low" to the American population, pointing to the rate of infection. He hedges this with some hand waving that it could change at any moment. I thought it was irresponsible to say the risk is low when they hadn't done any real testing and we've actually got outbreaks like in Kirkland. Regarding his comment that the WHO fatality rate is too high because of asymptomatic population and the denominator in reality is much higher. Fine. But he doesn't acknowledge that the numerator might be much higher as well. How many 80+ year olds, especially those with preexisting conditions, die and are assumed to have just passed from old age and the preexisting conditions. It is a characteristic of this virus that the people getting killed are the most likely ones not to raise huge red flags early. So in general, I would've preferred him to not say anything about risk until we actually get more numbers and real testing within our population. The average fatality rate hides the fact that this virus is very dangerous to the elderly.

We’ve released our branched versioning system for data at http://dvid.io. It’s entirely open source and uses a science HTTP API with pluggable data types and a versioned key-value backend. I’m currently developing a new backend for it: DAGStore, a lower-level ordered key-value store that will have an explicit immutable store and is tailored for distributed versioning of big data.

From the article: All of this helps explain why IPCC anticipates climate change will have a modest impact on economic growth. By 2100, IPCC projects the global economy will be 300 to 500% larger than it is today. Both IPCC and the Nobel-winning Yale economist, William Nordhaus, predict that warming of 2.5°C and 4°C would reduce gross domestic product (GDP) by 2% and 5% over that same period.

Question: Aside from what I assume are massive error bars on projecting economic growth due to scientific models 80 years in the future, I read that most of the models do not assume climate change impacts actual growth and therefore is exempt from compounding. On the other hand they use compounding to extrapolate growth. It seems odd that escalating climate damages as outlined in the reports would not impose a sizable tax. Anyone with more detailed knowledge of long-term economic projections care to comment on this?

I agree that my wording is too strong due to my memory of the latency results. As you increase the load, Go is closer to Rust/C than C# (see last benchmarks) and at a given load, C# isn’t in picture. It’s fair to say that Go/C# is similar while Java is far behind depending on the benchmarks that are important to you.

One example is n-dimensional arrays where each element could be a byte, uint16, int16, ... , uint64, int64, float, or even some custom vector. In a language with generics, you could describe the nD array more simply and the operators work depending on what was passed. The implementation can also be shared across a number of similar element types. In Go, though, you’d need to use interfaces and at some point you’d have to make tradeoffs about speed, simplicity of the implementation (e.g. hand-coding each element implementation) vs use of interface function calls to get properties, etc.

One example is to look at how Go handles the Image package vs other languages with generics. https://golang.org/pkg/image/

I hate bad analogies that have us arguing more about how well the analogy fits than simply debating the underlying question. When we talk about concurrency or the ability to make simple yet high-performing server apps, it's ridiculous to say Go is a bicycle compared to the motorcycle that is Java / C#.

User-space network driver? The benchmarks group Go closer to Rust with Java / C# way behind: https://github.com/ixy-languages/ixy-languages

This would be a very complex comparison and simple analogies do us a disservice. I think the author's blanket statement that Go is faster than Java / C# also seems too broad.

The whole thread follows from a comment that the number of bankruptcies doesn't matter if he's rich. The point is not that he's rich. It's whether he's increased or decreased the fortune he was given. If he's given $413 million and now has less than that, he's not a good business man. Since we are blocked from all his finances, we don't know where he stands now, but the multiple bankruptcies and NYTimes investigation suggests he's not accumulated wealth more than spent it. https://www.nytimes.com/2018/10/02/us/politics/donald-trump-...

Unable to look inside to see the textbook content, I went to the Bartlett Publishing website where it’s listed along with alternative textbooks debunking evolutionary theory and fossil forensics. Sorry, I’ll pass.

Scalability of state seems to be addressed by minimizing the need for nodes to retain the full state, and the main mechanism is shards. From whitepaper: “A further benefit is that any Atom that performs an inter-shard transfer is present in both the previous owner's and new owner's shards. This, in part, eliminates the need for a global state and mitigates any expensive inter-shard state verification operations needed to prevent "double spends".”

I’ve only skimmed the whitepaper and the consensus check against conflicts seems to be a series of methods with escalating computational cost. I think they assume Atoms (e.g., Transfer Atoms) won’t grow so big that they span massive numbers of shards.

I agree about the proliferation of buzzwords. Why not use Ledger instead of “Universe”?

Not sure how the eventual consistency model works in an economic context, although that’s how ATMs work.

Here’s a whitepaper on the underlying technology:

https://docs.radixdlt.com/alpha/learn/whitepapers/tempo

The conclusion in the paper:

“This paper has proposed a method for determining total order of events within a distributed system, without relying on trust, that is scalable, efficient and agnostic to its operating environment. To reduce overhead and increase performance, our solution defines a structured, shardable architecture that limits state transfer information to only members of the network that need it. A peer-to peer network provides the means for communication and information routing via a reliable and robust gossip protocol. Nodes maintain a logical clock when witnessing events and periodically produce a tamper-proof commitment representing those events. When witnessing events, nodes collaborate to create Temporal Proofs which contain verifiable space-time coordinates which are used to construct vector clocks and determine when an event was first seen by the network. Nodes can join and leave the network at will and rely on detectable consistency anomalies to synchronize and bring them up to date.”

The periodically produced commitment log is done with a Merkle tree.

I was comparing their #1 Cold Spring Harbor Lab to HHMI Janelia Research Campus, which doesn't even seem to be ranked in the top 100. Yet when I look at the research output, it seems pretty similar with the nod to Janelia in terms of FC, top article Altmetric score and article count.

https://go.nature.com/2x8zE68 (Janelia Nature Index FC = 36.41)

vs

https://go.nature.com/2RsQl5n (CSHL Nature Index FC = 31.24)

So the denominator for the # of articles per Dimensions must be much higher or N/A and disqualifies some institutions? Unfortunately the Dimensions article count isn't viewable for any institution not making the top 100 list.