• 2 Posts
  • 393 Comments
Joined 1 year ago
cake
Cake day: July 5th, 2023

help-circle



  • Franchises and ensemble casts really skew those results. So in the end it basically just comes down to “who was part of one or more large franchises” (primarily marvel), which to me is not that interesting.

    For me it would be much more interesting which actors brought the most “value” to a variety of unconnected movies, which would probably boost someone like Leonardo DiCaprio much higher, and in return throw out a bunch of actors from the MCU. For example Don Cheadle, who is on the list because he replaced Terrence Howard as War Machine. Which might have been the right call and an improvement, but imo don’t do the recast and you could swap those names on the list, because i don’t think he majorly shifted the franchise (unlike somone such as RDJ). Recast Leo and who knows how his movies would have performed with someone else in the lead.

    Not sure how an alternative ranking should work, but maybe take either the first/average/highest grossing movie of a movie series instead of adding all up.



  • The issue is that as someone already mentioned i doubt something like that was ever truly on the table.

    I think you can’t give assurances like that in a vacuum. If a nation e.g. the US would grant them, they’d only do so while simultaniously building up a physical presence in the territory and possibly also do deeper integrations military wise. You wouldn’t give such strong assurances while weakening your own ability to act on them.

    For Russia that would have never been acceptable.


  • Since I see this claim constantly: where in the Budapest memorandum did they promise protection?

    Looking at the Wikipedia summary nowhere does anyone give security assurances similar to NATO article 5 or the even stronger worded mutual defense clause article 42 TEU of the EU. The closest it comes to is in the fourth point, but that is only in the case of nuclear weapons being used. Which obviously hasn’t happened yet. Beyond that it is just a promise not to attack, which Russia has broken, but every other singator has kept. And as far as I can see it does not contain anything that compells others to act on someone else’s breach.


  • (disclaimer that this is purely my impression from what i’ve seen mentioned online, not firsthand knowledge)

    Which isn’t necessarily mutually exclusive. I was under the impression that the problems have more to do with high workloads and work environments that are chronically understaffed, not necessarily because of low salaries. Not claiming that all nurses are payed well, but it seems like that at least in the US there is a somewhat reasonable path to making good money (assuming you are willing to switch jobs and maybe continue to get sought after qualifications along the way).








  • I actually think all the posts talking about the size of communities, amount of memes on the frontpage and so on are wrong, since those will naturally change over time and are not fixed.

    Every platform will see changes in their user base to some degree. Reddit now is very different to Reddit 10 years ago. The same thing will happen to Lemmy: If growth continues we will see more engagement in niche communities, but also more low effort posts and reposts.

    Considering it doesn’t do anything fundamentally different to reddit in the way of being a content aggregator with comment section it will be a similar experience. It would be different if it e.g. had a function to make older posts resurface and stay relevant longer to foster longer conversations, or structure comments differently since right now the further down a chain you go, the less people will engage with it.


    Even if the average user doesn’t care about open source or federation, they’ll still benefit (and suffer) from the consequences.

    On a centralised platform like Reddit you are beholden to their will for better or worse, and incentives might change over time such in their case with taking investor money and going public. This can have consequences such as forcing out third party software (one of the events that brought a lot of people here), but also censoring specific content or taking away powers from moderators.

    There are downsides to it, since smaller, less professionally run instances might disappear at some point or have less reliability. But The upside is the option to choose and the resilience that should things change at one instance/community, you can switch without having to leave the whole ecosystem. And for that you do not have to be a moderator or volunteer

    The existence of different instances also to some degree helps identify users to some degree, the obvious choice being political instances like hexbear.


    The average user is not looking for NSFW

    That’s an assumption i’ll challenge. Looking at the amount of porn on the internet, the average person most definitely is looking for it. But that is probably a bit offtopic.


  • Is YouTube doing it with small creators actually in mind? Who knows, other than them?

    I am pretty confident in guessing that they are not doing it for selfless reasons. Imo the reason is that the less information they give the user, the more you are beholden to the algorithm choosing for you.

    But depending how they hide it it actually might not just be users, but also companies that e.g. buy ads from them. The less information they get, the more they need to trust whatever metric google offers them


  • I recently read a plausible reason that I hadn’t thought of yet:

    Apple would need to include a specific flexible cable rated for continuous movement with the mouse. If the port was in the regular spot, then people would ofc also use it wired at times. However if buyers would use regular charging cables, then the experience would both be worse and the cables might get damaged over time from bending.

    I still think the main reason is simply that they value form over function, otherwise the shape would be more ergonomic, but it’s another interesting factor to consider.


  • It was more the relation between them (40x) that struck me as bigger than I expected given the relative performance between photovoltaic and photosynthetic efficiency.

    Honestly i was suprised aswell by the difference. I did some further digging and while i think i found the german source they used, it was a bit harder to comprehend.

    But i also looked at this paper which forexample seems to support the rough numbers for energy/hectare biomass (it’s also on scihub if you dont have institutional access). It’s using fast growing tropical tree varieties as an example, but i imagine that if anything this would influence results favorably for biomass. If you look at figure 5 the yield is between 15-25 MWh/hectare.

    As a swede, energy usage in the winter is warm at heart which is something that is hard to compare and muddles the numbers. In Dec-Jan energy (kWh) output from solar is at best 9-10% of their peak output during summer at my latitudes, (further north, this goes towards zero as there is no sunlight in winter), so with that in mind, the stored 20MWh/hectare, available round the clock, looks apetizing until we find a better solution to store energy.

    Yeah, in the end there probably isn’t one solution. In Sweden for example area efficiency probably doesn’t matter as much due to your low population and large areas of woodland (that wouldn’t be suitable for much else). And you are right that PV probably wouldn’t work, so wind/hydro or maybe even tidal power generation would be the more appropriate competitors to compare biomass to, although those have more specific needs in terms of location.


  • I’ll try to track it down, but am kind of having a hard time finding their methology for those stats. Which ones do you think are unsure about? the 20MWh/year/hectare biomass, 800 MWh/year/hectare solar or the energy loss through stroage with batteries/hydrogen? Or something else?

    I’ll try a bit further to find their specific methology when i find the time. But for the solar part i also did a quick google search and found for example this paper. To quote from their conclusion:

    Based on empirical observations drawn from a large, nearly complete sample of utility-scale PV plants built in the United States through 2019, we find that both power and energy density have increased significantly over the past decade. Modelers and analysts, policymakers and regulators, and others who continue to rely on outdated benchmarks from the last comprehensive U.S.-based assessment of power and energy density conducted nearly a decade ago [6] will, therefore, significantly overstate the land requirements, and by extension perhaps also the land-use impacts, of utility-scale PV.

    Updated benchmarks as of 2019 established by this study are as follows.

    • Power density: 0.35 MWDC/acre (0.87 MWDC/hectare) for fixed-tilt and 0.24 MWDC/acre (0.59 MWDC/hectare) for tracking plants.
    • Energy density: 447 MWh/year/acre (1.10 GWh/year/ hectare) for fixed-tilt and 394 MWh/year/acre (0.97 GWh/year/hectare) for tracking plants.

    It is about the US and not Germany, but i wouldn’t expect there to be massive differences. If we assume that Germany has slightly worse conditions for solar, then 800MWh/year/hectare seem in the right ballpark.


  • I’m learning a lot about energy in Germany this way. Thanks!

    Glad to hear it, by doing some more digging i am learning new things aswell. I actually came across this site (sadly only available in German) by one of our environmental agencies, which i found quite good and mirrors pretty much my opinion (but maybe that is why).

    One particularly interesting piece of information is that they give specific numbers for the energy yield per hectare of biomass vs solar. They estimate using corn as an example crop that for biomass it is on average 20MWh/hectare vs 800MWh/hectare for solar, so a difference by a factor of 40x. Further for load balancing renewables they estimate 10% loss for short term storage through batteries and 40% loss when converted to chemical energy sources (presumably hydrogen).

    And while biomass currently plays a substantial role particularly in electricity production, they link to studies that long term (2050 being the target date) energy demands can be met completely without biomass useage.

    So basically farmers in Germany grow food to make fuel for cars? Like ethanol? Thats an abundance of agricultural lands!

    I mentioned E10 fuel earlier, which mixes 10% bioethanol into petrol. Seems like according to wikipedia it is also a thing in Sweden/Finland/Denmark. So that would be an example where plants ultimately get fed into cars. I think this is also a case where demand is induced by some EU directive that requires reduced emissions in the transport sector, and mixing in biofuels was one of the solutions to achieve this goal.

    I am not 100% certain on the specifics and i assume that it isn’t a pure play where literally nothing from those plants is used for other purposes. But it’s still fair to say that these crops are primarily planted for extracting energy from them, not for feeding livestock or food. So they are directly competing with those and without this demand farmers would plant different things.

    But if I’ve understood correctly from the TEN-T directive, Germany and Switzerland has invested pretty much in H2?

    Yeah, seems like we are investing a good amount in clean hydrogen. That includes storage and pipelines. I wasn’t particularly aware of the ten-t directive, but if i undestand it correctly this is about transportation? I think as far as hydrogen goes here in Germany the main focus for that is on use in industrial settings (particularly stuff like chemical processes and steel production). But ofc it’ll also be used in other areas.

    but most common is district heating. (I got the name wrong in translation earlier)

    Instead of one boiler in every house, there is one boiler per 50-100000 inhabitants or so. Efficiency is great and heat is pipes to where it is used. When it’s cold (-20 or so) those boilers go through tens of semitrucks of wood every day. And as I said, it’s a fairly common set up in parts of Europe, although i understand its not common i Germany.

    Not as common, but Munich for example has a large network, which currently is still mostly based around fossil fuels, but they are investing a lot in renewables (particularly geothermal) and plan to get it climate neutral in the longterm while expanding its reach.

    Also because of a recent law every municipality has to create a strategy paper for heating (until mid 2026 for larger, 2028 for smaller ones), which includes feasibility studies for district heating. So we might see them become more widespread.