_benedict

Global strict serializability is coming to Cassandra very soon [1]

[1] https://cwiki.apache.org/confluence/download/attachments/188...

The ICO[1] and case law[2] corroborate that audio recording outside your property with CCTV is generally prohibited.

[1] https://ico.org.uk/your-data-matters/domestic-cctv-systems-g...

[2] https://www.icaew.com/library/subject-gateways/law/legal-ale...

Depending where you live this might be illegal. In the U.K. you have an obligation not to record your neighbour’s property, and not to record conversations of members of the public.

On average the equivalent of 0.05% of the UK population claims asylum each year [1]. Last year about 0.07%. That doesn't sound like an endless flood to me. The world unfortunately does have a lot of refugees in it, but we do not experience the consequences of that here, much as we may act as though we do.

I don't personally have a strong belief that refugees should have a blank cheque to settle anywhere. It might be fairer to permit refugees to travel further, as it's unclear why neighbouring countries should shoulder the burden. However, the only point I was making is that my understanding is that today they _do_ have that right, which is why countries make it so hard to arrive.

[1] https://www.gov.uk/government/statistics/immigration-statist...

Being a refugee does not mean you have no resources. Would you flee your home country without taking your valuables? Would you have no friends or relatives to call upon to offer some monetary assistance, if it meant the difference between life and death?

Most have few resources. The ones with more resources have more options about where they flee to.

It isn’t a blank cheque, because making it to safety in a preferred country is exceptionally difficult and expensive.

However, it is a blank cheque once you’ve surmounted those challenges and have arrived at the safe country you have chosen (and were able to reach), though most countries receiving such refugees are doing their best not to honour this obligation.

It’s worth noting that the UK tries hard to prevent refugees reaching our shores, but the vast majority that do, and apply for asylum, are granted it [1].

https://www.gov.uk/government/statistics/immigration-statist...

Also anything that affects gut motility, so anything that messes with the nervous system in the gut.

If the signalling controlling eg the sphincter is messed up, the bile won’t escape to the intestines, and if the bile stops flowing you’ll get liver and pancreatic inflammation.

It might be that some toxin in the food could cause such an effect?

I’m not sure how easy it would be to tell if this was the case in an acute setting though. It isn’t exactly common, and is usually chronic and intermittent. Gall bladder removal would seem to be the wrong treatment in this case, but it might be hard to establish? My amateur medical knowledge is insufficient to know, just hypothesising other potential mechanisms.

This sort of reminds me of Lupin Beans that I discovered and started eating in large quantities before (thankfully quite quickly) finding out are actually very dangerous if not properly prepared, and that even properly prepared (as they probably were) they may be ill advised for regular long term consumption due to their effects in the nervous system. Their acute toxic effects sound a little different, but I wonder if some similar issue occurred here.

The Apache Foundation has next to zero input into how any of the projects are run, so even if this were true it would be irrelevant. How exactly do you imagine this influence working? It would have to be quite overt to have any effect on so many projects that are self organising.

The ASF is an umbrella organisation. For most projects the ASF’s involvement stretches only to procedures and structures for managing the project. The maintainers are mostly not even ASF Members (even those that steer projects). So, not much inference can be drawn from the failure of one project that applies to another.

If you read the article:

> 12-fold greater counts than when lidded

How did you source your “factor of 2” assertion?

If the average electricity bill includes houses that use gas for heating, then the electricity bill is grossly inaccurate for houses without gas heating (unless we are discussing an area that requires minimal heating demand)

I believe that when you get to 1RT, your LWTs will still effectively be 2RT if you include the coordinator's latency to the leader (except in the leader's region).

Please, no need to apologise! Your criticisms were all entirely well founded and the pain points you mentioned very real. I thought you expressed them considerately (I have seen plenty of vents that did not). I may reflexively defend Cassandra, but healthy and honest discussions around these things is great IMO.

> It did feel from many of those issues like it was a victim of its own complexity

I think there’s some truth in this, but I think the bigger problem was failing to give this complexity its proper respect (which would have been very costly and slowed feature development - perhaps consigning Cassandra to an also-ran position like others in the space, who knows?)

The level of service you recall from Germany was common in the U.K. until a few years ago, or certainly ten years ago this was my experience in every practice area I lived.

GP practices are all private in the U.K., the visits are only paid for by the NHS. Precisely why GP wait times have gotten so bad I can’t say, but the real-terms funding cuts the NHS has experienced over the past decade likely contribute.

Of course, you can top up with private care in the U.K. also, and this seems to be becoming increasingly common.

I probably know who you are referring to, but I don't feel it would be appropriate to discuss a specific identifiable individual in a public forum, no matter how innocuous that discussion might be! My fault for bring it up.

Cassandra was super easy to shoot yourself in the foot with, and it remains quite easy. You mention a few foot guns that have gotten better, a few that remain but that will get better, and a few that are sort of inherent to distributed databases.

Anti-compaction for instance should now not be a huge issue if you're running regular incremental repairs, and I hope even the few remaining caveats will be alleviated soon. Bootstrap is something you mention that is also going to get much easier for users this coming year, so that unsophisticated users can manage their cluster membership safely.

Application-misconfigured consistency levels is a really obvious one that isn't strictly Cassandra's fault but for which much better help could be given to the user, and I expect some major improvements here in the next year or two, so that users can configure tables with consistency properties that the database guarantees (to some extent, the user will always be able to screw it up by providing the wrong consistency identifier, but at least the scope is reduced to accidental misconfiguration rather than misunderstanding). This is something we're considering as part of the introduction of general purpose transactions later this year.

Poor data models and partition keys are things the database can offer less help with, though I anticipate much better support for ordered partitioning in future, that would help poorly-selected partition keys, as clustering keys can be used for partitioning there too.

Regarding the choice of language, Java has upsides and downsides. GC spirals are something we have control over at the end of the day, and we continue to do better at (as does the JVM), but guaranteeing no segfaults (and not worrying about the ABA problem) is a big benefit we get in return. I wish we had more control over things like memory placement and execution, but these things may be coming to the JVM to some extent (Loom I expect to benefit Cassandra hugely, and value types later), but equally distributed systems problems often give you enough things to worry about.

The visibility you have into a Java process is fantastic, however, and we are starting to make use of the ease with which you can modify the code Java runs for system validation, using byte weaving to permit us to simulate clusters as they are run, with adversarial event orderings, to ensure those notoriously hard distributed systems problems are correctly solved.

If you do want to speak privately, about anything Cassandra related, the lowercase part of my username (i.e. without the _ prefix) at apache.org reaches me.

Re: your comparison, LWTs in Cassandra in 4.1 (about to release) offer global 1RT reads, and 2RT for writes. Your comparison suggests you take 3RTs? I suspect you may be moving to offering 1RT for those in the local region, and 2RT for all operations in another region?

You make a point of mentioning your MVs (and presumably your global secondary indexes) may get out of sync with the base table, which is the reason the Cassandra community decided to mark them experimental. The Cassandra community has become very conservative since it began being driven by users of the software. I should not draw any conclusions around the relative merits of features from this conservatism.

The Cassandra community paused feature development for several years, due to this very conservatism, in order to focus on delivering safety and reliability at scale. Now that's done, you can expect a lot more visible activity.

I'm intrigued: The Apache Cassandra developer community is quite small and well known to each other, and the only major contributor I know to have left the community to run a cluster did so almost a decade ago?

Cassandra had a lot of rough edges until very recently, and was only really suitable for sophisticated users, as most contributors would have attested. The project has matured a lot over the past couple of years in particular, as the broader community stepped up in the face of DataStax's (temporary) withdrawal. The investment by some large scale users has transformed it, and the next couple of years will do so even more.

None of the issues the project faced were really related to the language chosen, in my opinion, rather than the maturity of the project and how it grew at the time.

I think you maybe overstate it, but I agree Paxos is nicer than Raft. The papers in which we outlines it are awful though.

Raft is more understandable, because they wrote a paper that was easy to understand. You are much better off reading Heidi Howard to understand Paxos, than Lamport.

It's very slow, particularly since introducing byte weaving to introduce chaos in concurrency-sensitive places. Presently it takes a single thread in the region of twenty minutes to simulate a cluster from startup to shutdown, and to run a thousand linearisable operations (down from a couple of minutes prior to those changes). I haven't got any numbers on the rate of context switches per second, but it is certainly a great deal less than 2.5M/s.

This kind of approach is in my opinion a _requirement_ for work on distributed consensus protocols, which the Cassandra community is improving at the moment: the upcoming 4.1 release has some major improvements to its Paxos implementation that were validated by this simulation, and the following release is expected to contain a novel distributed consensus protocol we have named Accord, for supporting multi-key transactions.

So, I'm not sure it has affected the speed or ease of shipping improvements, rather than enabled a class of work that would have previously been impossible to do safely (IMO).

Cassandra already does this[1] (and a whole lot more besides to introduce chaos to the simulation), and just accepts that there's a huge penalty of pausing and scheduling threads.

I'm really looking forward to being able to use Loom to run much more efficient cluster simulations.

[1] https://cwiki.apache.org/confluence/display/CASSANDRA/CEP-10...

Spam volumes are also down dramatically due to conventional means like making it illegal, and taking down bot farms that were sending it. Also the introduction of DKIM and DMARC making it harder across the board to be seen as a legitimate sender, and the fact that many-to-many emails are a huge red flag for spam filters, and not a concept for social media.

Spam is a dramatically easier problem, and has many more mechanism to suppress it both legally and technologically.

Nowhere did the GP say that it was antithetical to all interests of the ASF.

I think you have still missed the important distinction between the ASF and the ASL.

Your opinion is based around your interpretation of the ASL, whereas the GP was referencing the ASF - an organisation of people, with their own shared values, that created the ASL. You have not referenced the ASF at all, only the ASL and some general concepts around FOSS, which are not particularly relevant to a discussion of the ASF's goals and values.

Perhaps limit your policing of others' phraseology to cases where it's clearly wrong or harmful?

The Apache Software License and the Apache Software Foundation are not the same thing. The latter is a human organisation and its values are determined by its members - some of these values are not even codified, so it may be hard for an outsider to determine them - but in this case I am inclined to agree with the GP.

The ASF promotes permissive software distribution - this includes the right to remove permissions when relicensing. However, distributing such a modified license under the Apache umbrella potentially harms that goal by muddying the meaning of "Apache licensed"

I think in the case of masks, it is highly plausible there was no deliberate attempt to mislead.

While at the time it seemed pretty clear to me that masks had a high probability of being helpful, there was a dearth of research that would support this conclusion - indeed the very limited research conducted on public mask usage did not seem to support it.

On top of this, I believe a major contributing factor was that many medical experts were used to very stringent mask protocols, and did not seem to consider the obvious differences between a hospital setting (where Covid may be rife) and every day life where it is a medium-to-low risk event, and carried over their belief that anything less than their stringent (and unrealistic protocols for the public) would render them useless.

I saw many regular doctors expressing this belief on Twitter at the time, as well as explaining how mask fitting was a laborious process, and that without this the mask would be useless, or that if you touched the mask while taking it on or off it would be worse than useless.

So, I think it is entirely possible that experts were giving leaders and journalists their honest assessment that masks were useless for the public - no doubt more exuberantly than they would have done if masks were plentiful, but not necessarily dishonestly.

I don't disagree that public messaging is highly manipulated in general, but I think it is more plausible that this particular case was fabricated by flaws in the decision process rather than dishonesty.

I believe this refers to replacing plastic packaging with paper packaging, not the removal of packaging.

This also tracks with apoe4 carriers being at greater risk, as they are also at greater risk of raised cholesterol from dietary sources of saturated fat

So, firstly 88% of homes were built pre-2000[1] and the regulations that resulted in these improvements came into force in 2010. Assuming the regulations instantly had the desired impact (unlikely) we're discussing around 6% of UK houses. Those built earlier (on average) achieve significantly worse than 10m3/hr/m2 air permeability of the building envelope[2]. So the vast majority of UK homes have more ventilation than a Passivhaus building, and for the purposes of the initial discussion I think this is enough to demonstrate the inadequacy of this level of ventilation for the stated purpose, else Covid spread wouldn't be a problem in the UK within the home.

It appears that for new builds latest research[3] shows an average of around 4m3/hr/m2, which is better than I expected, however the research suggests this is in part due to temporary measures by builders to achieve higher ratings than they would otherwise, and this is born out by the distribution of ratings achieved with an unlikely precipitous drop at precisely 5m3/hr/m2.

However, for houses built to this standard some form of additional ventilation is a requirement of Part F[4]. It is possible to build a house with as low as 3m3/hr/m2 and provide no additional ventilation, but it must have been designed to > 5m3/hr/m2 so this should not be common. For social housing we can assume a single bedroom home will be around 40m2 with 2.6m between floors, and the required ventilation rate is 13l/s, or 47m3/hr/m2. This is around 0.45ACH.

In fact the legal minimum rate per m2 is 1.08m3/hr, so for a typical home with 2.6m between floors this works out to 0.42ACH. So any house built to modern standards and used correctly should achieve the Passivhaus level of ventilation, unless it has tall ceilings. Certainly any UK home built with continuous mechanical ventilation is pretty comparable to Passivhaus standards for airflow.

> People don’t open trickle vents, they turn down the extract fan timer because of fan noise

People can also turn off the MVHR in a Passivhaus. I'm not sure what purpose it serves to compare correct use of a Passivhaus to incorrect use of this small fraction of UK homes.

[1]https://www.statista.com/statistics/292252/age-of-housing-dw...

[2]https://www.researchgate.net/profile/Dominic-Miles-Shenton/p...

[3]https://discovery.ucl.ac.uk/id/eprint/1573697/1/1-s2.0-S0378...

[4]https://assets.publishing.service.gov.uk/government/uploads/...

Do you happen to know what make and model? The only cooling functions I have found besides summer bypass involve a separate cooling loop that ultimately relies on a heat pump. I'd be interested to read about an MVHR with this facility.