Nextdoor is the only social network I have used that confirms your real-life location, and it's not any better than the planetary scale sites. BBSes were better due to how users self-selected into them. Small geographic clusters don't inherently promote quality.
There are a lot of "better than Bloom" filters that work similarly in some aspects. I have used Cuckoo [1] and Ribbon [2] filters for Bloom-type applications. If you have an application where you do a lot of one kind of searching, it may also be worth implementing a specialized variant of a data structure. I needed a Cuckoo-type filter on the JVM but only for 64 bit integers and I was able to make a smaller, faster code base that was specialized to this data type instead of handling generic objects.
You need to know up front whether you need to be able to dynamically add entries to the filter or if your application can tolerate rebuilding the filter entirely whenever the underlying data changes. In the latter case you have more freedom to choose data structures; many of the modern "better than Bloom" filters are more compact but don't support dynamic updates.
I wonder how often in the wild people are tuning for a 1% false positive rate versus a much lower one, like .1%. You do quickly reach data set sizes where even 1% introduces some strain on resources or responsiveness.
Cuckoo claims 70% of the size of bloom for the same error rate, and the space is logarithmic to the error rate. Looks like about 6.6 bits per record versus 9.56 bits for bloom at 1%. But at .5% error rate a cuckoo is 7.6 bpr. In fact you can get to about a .13% error rate for a cuckoo only a hair larger than the equivalent bloom filter (n^9.567 = 758.5)
Cuckoo filters can do even better with the small adjustment of using windows instead of buckets. See "3.5-Way Cuckoo Hashing for the Price of 2-and-a-Bit": https://scispace.com/pdf/3-5-way-cuckoo-hashing-for-the-pric.... (This significantly improves load factors rather than changing anything else about the filter, and ends up smaller than the semi-sorted variant for typical configurations, without the rigmarole.)
My fairly niche use case for these kinds of data structures was hardware firewalls running mostly on SRAM, which needed a sub one-in-a-billion false positive rate.
Residential electricity service from PG&E has never properly separated the fixed costs of service delivery to a location (maintaining poles, transformers, and wires) from the cost of marginal energy consumption. It has folded much of the fixed infrastructure costs into the per-kilowatt-hour unit price. This functions as an implicit subsidy for households that need grid tied electrical service but do not consume much electricity from the grid.
These implicit subsidies used to mostly benefit lower-income households (though not always: properties like seasonal vacation houses also benefited). Now, higher-income households are more likely to benefit from this structure because they are more likely to install rooftop solar (reducing kWh consumption) but still need the grid to work at night. Crediting solar households for grid exports makes this problem especially acute but it would also exist even if solar households were merely reducing the kWh drawn from the grid during daytime.
One remedy could be to fully separate the costs of fixed infrastructure from per-kWh unit charges and set prices directly proportional to costs. But that is probably politically unfeasible because there will be outcry that prices proportional to costs would hurt low-income, low-consumption households.
Another way to remedy it would use the previous approach but give offsetting vouchers to households that would face financial hardship as a result of the change in pricing structure. I don't know why the underlying issue has remained unaddressed in favor of patchy solar-specific changes to the law.
This is the exact issue in most grids TBH. Not sure I've seen one that has properly priced fixed costs with properly priced usage costs. My grid has the split, but the costs seem out of proportion. I'd say if the concern is low-income pricing, factor out fixed costs and rates so the the mean low-income prices stay the same. Mix that with only offering wholesale rates for customer backfeed and some reasonable controls about how much and when customers can backfeed.
The things that PG&E has notably been neglecting, resulting in highly destructive wildfires?
The idea that it costs every single ratepayer $100/month to maintain the infrastructure is ludicrous. It's just attempting to deflect blame from PG&E's horrible mismanagement to environmentalists.
In places with honestly run utilities that cost is closer to $5-$10/month per household.
Wow that makes sense yeah, and it’s difficult because you don’t want to disincentivize solar but socializing the grid costs, which would normally be a good move, does have that effect when done like this.
I wouldn't care if the money were actually going to improving the system but it's not, it's going to shareholders and paying for their crimes. The fact we have to pay for their crimes and their CEO and execs and shareholders can continue to increase their salaries and bonuses along with our rates just makes me very angry. The entire company should be held responsible and things like THEIR bonuses should be withheld until the company has fixed everything.
The more depressing problem is that the US is raising trade barriers for solar equipment made outside of China nearly as fast. I think that the current administration just dislikes solar power, and only a part of that animosity is due to its tough-on-China stance.
"South Korea files WTO complaint over US solar tariffs"
The US Department of Commerce (DOC) has issued anti-dumping and anti-subsidy tariffs against solar cell imports from Thailand, Vietnam, Malaysia and Cambodia.
"A Casualty of Trump’s Tariffs: India’s Nascent Solar Industry"
I’d be curious how much of the goal of the tariffs is for solar actually produced in those countries vs just slightly modified goods from China. That’s one reason tariffs don’t work when they aren’t applied universally
I don’t doubt that another benefit is the current admin just simply doesn’t like solar of course
I know that at least Malaysia, Mexico, and South Korea had genuinely independent solar manufacturing that was not just repackaging Chinese components. In reading the solar industry trade press, I haven't seen indications that the new tariffs are targeted to fight "leakage" of Chinese products via intermediary countries. They seem aimed at keeping import prices elevated across the board.
In India's case it appears to be an attempt to strongarm a trade deal - Indian energy firms that have invested heavily in the solar and renewables manufacturing chain like Reliance Group and Adani Group are diversified energy companies that are also competing against American ONG majors like Exxon and Chevron.
Heck, the additonal 25% tariff on India for Russian oil imports only came up after Exxon started lobbying to re-enter the Russian market [0], but the Russians sold Exxon's Russian assets to India's ONGC [1].
This alone, brings me to believe that when we get there, there will be some built-in safety mechanism to preserve power for those that are powerful.
I think that people will try yet fail to build perfect security mechanisms for controlling the spread of smart robots. If robots build robots, and smart robots are smart because of software, then trying to stop copies of smart robots is like trying to prevent copies of movies. Most people will come to have their material needs met by smart robots, and in the typical case this will be an improvement in quality of life for people on Earth.
Unfortunately, it also means that when people have deadly impulses disconnected from material deprivation, the power to kill will be greatly amplified. Tiny states and even sub-national groups could easily acquire nuclear weapons. The key technical insights are already published. It just takes engineering work and willingness to violate international norms to develop an arsenal. (International norms are not going to be enforceable by "soft" measures like sanctions if every nation can be simultaneously autarkic and prosperous, thanks to smart robots.)
WMD proliferation in turn may drive more comprehensive, paranoid global surveillance and an increasing number of preemptive attacks on facilities that could become weapons factories. Fear of military attack then drives more small states to actually seek a nuclear deterrent. An increased number of actors with nuclear weapons increases the chance that they will kill people on a large scale either deliberately or accidentally (like a "retaliatory" launch against a falsely detected incoming strike from another nation.)
It's impractical even for tinkerers. YouTube claims to get over 20 million videos uploaded daily and it has well over 10 billion stored videos in its corpus. The metadata alone is tens of terabytes. The usual introduction-to-recommendations approaches out there are going to completely fall over on an item set of this size, even if you have disk space to spare.
France doesn't have any new reactors under construction. Its annual nuclear generation peaked in 2005:
~/git/iaea-pris % sqlite3 pris_data.db
sqlite> select year, sum(electricity_supplied_gwh) from reactor_statistics, reactor_operating_history where reactor_statistics.reactor_id=reactor_operating_history.reactor_id and country_code='FR' and year > '1999' group by year;
year sum(electricity_supplied_gwh)
---- -----------------------------
2000 395392.3
2001 401256.49
2002 415110.33
2003 421028.62
2004 428040.69
2005 431179.56
2006 429819.63
2007 420129.49
2008 419800.32
2009 391752.97
2010 410086.42
2011 423509.48
2012 407437.88
2013 405898.51
2014 418001.4
2015 419035.02
2016 386452.88
2017 381846.02
2018 395908.34
2019 382402.75
2020 338735.78
2021 363394.15
2022 282093.23
2023 323773.23
2024 364390.78
France is planning new EPR2 reactors, but no construction is expected to start before 2027 and none would run before the 2030s. I put little trust in announcements of future plans without actual construction work, whether the plans are for nuclear reactors, wind farms, data centers, or any other major investment.
Look at the number of planned reactors and the number of reactors likely reaching EOL. France plans to build fewer than fifteen new plants (delivery date tbd). If they started building all of them today half of their fleet would be fifty years old by the time construction was done.
Once again: source? Mycle Schneider by any chance?
Anyway: those numbers are not "reactors likely reaching EOL". Those are reactors reaching the end of their original operating license.
These two things are not the same. At all.
Initial operating licenses were intentionally relatively short, because at the time there was no experience with the longevity of reactors. So you conservatively license towards the short end.
Now that we have that experience, reactor operating licenses are getting extended. A lot. The first reactors in the US have had their licenses extended to 80 years, and the current consensus appears to be that 100 won't be a problem.
So France won't be running low on nuclear power anytime soon. Unless you're Mycle Schneider and/or confuse "current operating license expiry" with "EOL".
France does not have any new reactors currently under construction because until March 2023, expansion of nuclear generating capacity was forbidden by law.
So even to build the one Flamanville 3 reactor, they had to shut down two older reactors in Fessenheim in order to not have an illegal increase in capacity.
Now that the law has been rescinded, they are planning 6 simplified EPR2 reactors, taking lessons from the fairly catastrophic EPR project FV3.
(Of course, even that catastrophic reactor will be more profitable than any intermittent renewable projects in, for example, Germany, but hey, the standards for what counts as "success" and what as "failure" are different for nuclear and for renewables).
France also currently does not need to urgently expand their nuclear fleet, so the schedule for the EPR2s matches those needs and the need to fully account for the problems with FV3. Instead, they are increasing the production of their existing fleet, both by operational upgrades and also by increasing use of intermittent renewables to cover variations in demand, allowing the nuclear fleet to run closer to fully rated capacity instead of having to load-follow.
I have wrestled with it too. I believe it's because wikitext is an ad-hoc format that evolved so that the only 100% correct parser/renderer is the MediaWiki implementation. It's like asking for an SDK that correctly parses Perl. Only Perl can do that.
There are a bunch of mainly-compatible third party parsers in various languages. The best one I've found so far is Sweble but even it mishandles a small percentage of rare cases.
That's why global nuclear disarmament seems only slightly more plausible to me than (e.g.) global artillery disarmament. For the foreseeable future there are going to be some nations that see nuclear weapons as the more affordable (or the only affordable) deterrent against rival nations that can field much larger armed forces.
reply