🎓 Dr. Freemo 🇳🇱 @freemo@social.cleverlibre.org

@holybear

If I pay 100$ tax on 200$, and then make 800$ I will pay 400$ in tax... as my income increases my tax burden also increases. Assuming of course you are already in the highest tax bracket

@MisterWanko @rbreich

@hasmis No thats not what your saying.. the model is a static thing that can be proven wrong... its **prediction** is of a "rate of change" not of a value at a point of future... it claims "the rate of change will be X so long as the input is Y".

The model never changes, it always states a relationship between X and Y and that relationship doesnt change... X and Y (the real world) does change, and thus the prediction for the future does.

Do you not fathom how idiotic what your arguing sounds like to anyone with any degree of common sense here.

"The model predicts if we keep pumping CO2 levels at the rate we are temperature will raise by 5 degrees! .. (5 years later after reducing emissions by 20%)... See the model was wrong! Temperature didnt increase by 5 degrees"... that sounds like a level of idiocy I cant even argue with, OBVIOUSLY the temperature isnt what the model said 5 years ago, we reduced emissions and prevent the outcome it warned against! But guess what, if you plug in what the temperature should be into that same model with a 20% reduction in CO2 emissions all of a sudden the model is right, AS EXPECTED... to use that as an argument that model is wrong is just so idiotic I have to think it is intentional.

@Lunatech

Omit age, improve tests, monitor future and historic demonstration of capabilities.

Age is the leqast fair of all since it discreminates explicitly against age. Tests may be bias but atleast it is incidental and possible to improve it.

@Hyolobrika@berserker.town

As long as it is cold enough out that you have to run a heater you have 0 enviromental impact, 0 additional electricity cost, and income rather than expense from running a cryptominer.

Its really the only scenario where you can argue cryptomining has no negative enviromental impact.

Of course this doesnt count the impact of manufacturing the GPUs though.

@lore

@eris

By that logic doing anything that makes you money is a scam :)

@lore

@eris

How ya figure? The user doesnt invest anything, they get a free space heater and get paid money. How can that be a scam?

@lore

@MisterWanko

Umm thats exactly how taxes work already...

@rbreich

@lore Actually this rabbit hole goes much deeper than people think.

A cryptocurrency miner is almost exactly as efficient at heating a room as a space heater, except it pays you when it does. So if we replaced all space heaters with cryptominers then you would get payed to heat your home rather than paying to heat your home.

In fact there was a company that did exactly this, they sold space heaters that were cryptominers. I wish I could remember the name.

@hasmis See this is why you should be asking scientists and not just making assumptions. Everything you just said is wrong.

Models make predictions, the prediction of a model changes day by day as the inputs to the model changes. Lets makeup some arbitrary placeholder numbers to show what I mean..

Say some model claims Y = X*2. then Y will only be a straight line when X doesnt change.

Models about weather claim things like "If we keep pumping CO2 at the current rate into our atmosphere then the temperature will rise by X degrees a year"... that is a prediction the model makes at the **current** state of things... **obviously** if people actually make an effort to improve our CO2 emissions then temperature will rise by **less** than the model predicted back when we were producing more CO2.

So any model will change year by year based on how people change year by year.. as we improve our emissions the model will predict less, as we do worse it will predict more.

Again if you are showing a straight red line you are **not** showing a model, you are showing a fixed prediction of a model at a fixed point in time, the model itself is only claiming that prediction to be true if we do not change our behavior, which we do, and therefore you are blatantly wrong in your interpritation,.

@jaysonmassey Subtlety never won an election :)

@hasmis The redline is clearly **not** a model, it wouldnt be a straight line. Models are by definition adaptive, they define how temperature would change as the response of the public changes.. if its a straight line it is very clearly not a model you are showing.

@hasmis https://qoto.org/@freemo/106825140617586262

@hasmis Uhh no.. th3e models predict a number, anyone can check how close the real number is to what the model claims.. its hardly exclusive to scientists.

@Cosmo Ahh philly, I miss home... the chemical spill probably added some well needed flavor to the skukyll punch

@hasmis The average of of models accepted by scientists at the time 30 years ago have performed remarkably accurately.

@shuttersparks Agreed!

@georgetakei I admire the fact that you encourage others to make fun of you! As always you rock!

@skobkin fair, there is a noticable lag on gitlab, but reasonable id say.

@skobkin I wanted a git solution that includes CI as an integrated element, not an after thought. There are many reasons for this but mainly because the CI and GIT need to integrate tightly so third-party solutions tend to be convoluted or insufficient. For example we want to ensure merge requests dont show up and are auto-blocked if they dont pass CI, and that we have all the linting, anakysis and security warnings all in one place.

The issue of it being slow is legitimate, and im currently dealing with that issue.. ill see how fast it is after im done hacking it to fix that. Obviously if it will be slow on good hardware we may need to abandon it. but gitlab.com seems to run fast with a huge user base so should be possible.

@skobkin It is, admitidly, fairly heavy and slow. But we have more than enough money to afford the hardware to compensate for that, and we know it can scale linearly if done right.

The reason it was picked is mainly for its CI toolset which is the best i have ever come across. It also helps that we have a free ultimate license that extends to all our users (normally 100$ per user per month)