On the PS2 there was a very small memory area, called the scratchpad, that was very quick to access, the rough idea on the PS2 was to DMA data in and out of the scratch pad, and then do work in the data, without creating contention with everything else going on at the same time.
In general most developers struggled to do much with it, it was just too small (combined with the fiddlyness of using it).
PS2 programmer's were very used to thinking in this way as it's how the rendering had to be done. There is a couple of vector units, and one of them is connected to the GPU, so the general structure most developers followed was to have 4 buffers in the VU memory (I think it only had 16kb of memory or something pretty small), but essentially in parallel you'd have:
1. New data being DMAd in from main memory to VU memory (into say buffer 1/4).
2. Previous data in buffer 3/4 being transformed, lit, coloured, etc and output into buffer 4/4.
3. Data from buffer 2/4 being sent/rendered by the GPU.
Then once the above had finished it would flip, so you'd alternate like:
Data in: B1 (main memory to VU)
Data out: B2 (VU to GPU)
Data process from: B3 (VU processing)
Data process to: B4 (VU processing)
Data in: B3
Data out: B4
Data process from: B1
Data process to: B2
The VU has two pipelines running in parallel (float and integer), and every instruction had an exact number of cycles to process, if you read a result before it is ready you stall the pipeline, so you had to painstakingly interleave and order your instructions to process three verts at a time and be very clever about register pressure etc.
There is obviously some clever syncing logic to allow all of this to work, allowing the DMA to wait until the VU kicks off the next GPU batch etc.
It was complex to get your head around, set up all the moving parts and debug when it goes wrong. When it goes wrong it pretty much just hangs, so you had to write a lot of validators. On PS2 you basically spend the frame building up a huge DMA list, and then at the end of the frame kick it off and it renders everything, so the DMA will transfer VU programs to the VU, upload data to the VU, wait for it to process and upload next batch, at the end upload next program, upload settings to GPU registers, bacially everything. Once that DMA is kicked off no more CPU code is involved in rendering the frame, so you have a MB or so of pure memory transfer instructions firing off, if any of them are wrong you are in a world of pain.
Then throw in, just to keep things interesting, the fact that anything you write to memory is likely stuck in caches, and DMA doesn't seem caches, so extra care has to be taken to make sure caches are flushed before using DMA.
It was a magical, horrible, wonderful, painful, joyous, impossible, satisfying, sickening, amazing time.
Well, if you were traveling at light speed you could move anywhere in the universe instantly. If you are an observer on earth, watching an object move away from you at the speed of light, then it will take a very long time to traverse the tiniest regions of the universe.
even in a vaccum, light speed travel from the travelers POV still takes time, and said traveler would perceive time passing exactly as occurring in that local space.
But yes you're totally correct, the observer on earth would in this time see only the briefest part of my journey's trail due to light from my journey taking "exponentially" longer to travel back to the observer.
Actually, in special relativity, if you could somehow travel at light speed, your own proper time wouldn’t pass at all. The journey would be instant from your perspective. You’d experience zero time between departure and arrival.
That’s not just "relativistically instant", that’s literally instant in your frame. The time dilation becomes total at light speed, and length contraction collapses the entire distance in the direction of travel to zero.
Now, it’s true we can’t really assign a rest frame to a photon, so this is a thought experiment. But if you extrapolate the math, the conclusion holds: no time passes for something traveling at light speed.
> if i am not writing or editing a file, just doing nothing, then why FS breaks?
For a few reasons, but the most common is that when you write a file, you 'fwrite' operation doesn't typically immediately write blocks to the disk, instead it gets written to caches that will be written to disk in due course. Changes to the disk, tables with crucial data, etc again are just updated in memory, and written to the disk at a 'sensible' time. It's possible that you aren't actively performing operations to modify the disk, but various previous changes are sitting in memory waiting to be synced. If you kill the power then these changes aren't flushed so you can end up with corruption/filesystem problems. The main reason things work this way is for performance.
then I'm screwed but I don't think the model be that off from the actual question.
If I ask "copy a file from local to AWS S3" it won't return the delete command. But yeah you gotta be careful.
I show a warning after every command you run "Please don't run commands that you don't understand "especially destrictive ones")
I use Copy Me That (website/app), I am not sure if it has all the features you want, but I use it for storing recipes in a "no nonsense" format and can share recipes with friends/family. I get the impression if you pay you can do shopping lists etc but I haven't used that side of it to comment.
I also use this. It's the best recipe collector website/app i've found (there's a chrome extension and the parsing is excellent). It also supports all the features you want, like meal planning[0] and shopping lists[1] for free but you can get more powerful capabilities with a modest subscription fee.
I heard this £30 saving stuff on the radio this morning, and it made me shake my head a little. I remember when I got a socket adapter that measures power consumption, and I went around checking all my appliances, eager to find out which are drawing lots of power, excited at the money savings. Started with my old, huge, plasma TV, that is the best part of 15 years old - surely chugging loads of power in standby? Nope, next to nothing. Monitors, same. Network switches, same. Essentially I didn't find anything in my property that was doing anything unexpected. Having hot water (which I get via electricity as my flat doesn't have gas) for a single day will use up more energy than I'd save unplugging all my devices. Similar for heating (again, electricity based) on a colder day in winter would cost more than all the standby devices.
I remember a long time ago (25 years?) someone 'invented' a device where you plug your TV into it, and it has an IR receiver to go with it, you program it to respond to your TVs power button (on the remote control), and then when you go to put your TV into standby, the plug then turns off too, and when you switch it on it does the opposite. At the time it was being touted as the money saver, and I got the impression (although never actually tested!) TVs back then did draw a lot in standby, so this 'invention' was a legitimately good one. But it didn't seem to really take off, I am assume because standby was rapidly implemented in a way where there is such minimal electrical draw that it would be pointless.
A nice side effect of people doing this is that, depending on where you live, you might be able to obtain these older, less power efficient devices for relatively cheap on the used market. Of course if you can barely afford electricity buying them is probably not a bright idea, but for someone with e.g. their own PV system it can be a great way to save some money while also reducing the amount of e-waste going to landfills.
Even better, this is not limited to large power hungry appliances. You can build a surprisingly decent gaming handheld out of an old flagship phone with the help of rooting and some hardware mods [1], basically taking advantage of the low resale value of Android devices.
Thankfully, you are wrong about that! Just like you don't need a cooler source _from which to source cold_ in the summer with AC. Technology Connections has a fun (well, to me) video on how heat pumps work and are great. Literally just AC units running in reverse, thanks to having a valve.
At a certain temperature differential you need to source "heat" from elsewhere, which is why up north you either need a geothermal heat pump (underground piping) or you supplement with gas or electric heat.
Some very efficient pumps can work down to 0 degrees, but most start losing efficiency at 25 to 40 (Fahrenheit).
For some parts of the world, a heat pump is perfect - pump in heat in the winter and heat out in the summer, but when the outside temperature can hit -40º Fahrenheit (which is -40º Celsius), you're gonna need something else.
For the overwhelming majority of the world they are perfect. Even in high latitudes, they might only reach those extremes for a few days in a year, that's when you would require a backup.
Losing efficiency is fine since they are so ridiculously efficient to begin with.
100% efficiency is rather low for heating. Heat pumps easily give you 400%. Efficiency in this case calculated as: generated energy (in heat) / electrical energy. Efficiency goes beyond 100% because heat is extracted from air or water.
A good rule of thumb, and I hope someone more knowledgeable can correct me if I'm wrong, is that stuff that generate more heat will consume more power. So hot shower, iron, curling iron, hairdryer, rice cooker, this kind of stuff. Electronics that generate minimal heat will consume minimal power in general.
In general, all the electricity consumed by an electrical device gets converted into heat. If your GPU uses 250W of power, it generates 250W of heat. If your whole computer draws 1kW from the wall, it generates 1kW of heat. If your little raspberry pi and its power supply together draws 15W from the wall, it generates 5W of heat. A 1500W electrical space heater converts 1500W of electricity into 1500W of heat.
It's just basic energy physics. Making something hot or cold is real work. Same with moving something or producing light. But light is extremely "cheap" compared to the rest. Computing data or rendering a video game isn't real work and would consume no power if it wasn't for the inefficiency of current flowing through the chip making it hot.
If I remember my long ago physics classes correctly, irreversible computation inherently costs energy: you can't get its energy "consumption" down to zero because of entropy, no matter how efficient the technology. Our computer architectures are far from reversible in a physical sense. If they were, you could run programs in reverse. That said, the minimum amount of energy required is an absolutely tiny rounding error compared to the waste heat of today's technology.
And if the heat goes into water, it needs even more power because water is notoriously hard to heat up.
At one point I owned many appliances that leaked heat, and I think I learned to estimate how much power they drew simply by putting my hand on them and feeling how hot they were. I'm not sure I have that superpower anymore. (Obviously it was never that exact, because it depends on many other things like volume, material, isolation, etc. But you can get fairly close for common household things – between a finger and a few heads in size, surrounded by air, plastic case.)
Or to put it another way -- if hot water goes down the drain (shower, clothes or dish washing, etc.), then that's another place the waste energy is going.
Better down the drain then to radiate it back into your air. Then you gotta spend more energy pumping that heat outside (if you live in a hotter climate like I do).
Alternatively, use it to warm up the cold water coming into the shower (I'd guess the largest usage of hot water in a home). Warmer "cold" water mixing with the hot water means less hot water used out of the hot water heater. My understanding, placing this only on the cold supply to the shower is it'll only impact your showers. https://ecodrain.com/en/
Correct, and the difference is not even close - anything to do with temperature manipulation consumes orders of magnitude more. Literally, not figuratively
The laptop I'm writing this on consumes around 10W. Kettle 2100W
Making one coffee in the morning consumes enough electricity to power the laptop for the whole workday
The idle power consumption of my home is dominated by the fridge and freezer (around 150W combined). Idle mode of any other devices is a rounding error
My power company sent me a power strip that has a somewhat similar feature. It constantly powers the TV, but the rest of the strip is turned off when the TV is off. It's so you won't waste power having Roku, etc., always running.
> I remember a long time ago (25 years?) someone 'invented' a device where you plug your TV into it, and it has an IR receiver to go with it, you program it to respond to your TVs power button (on the remote control), and then when you go to put your TV into standby, the plug then turns off too, and when you switch it on it does the opposite. At the time it was being touted as the money saver, and I got the impression (although never actually tested!) TVs back then did draw a lot in standby, so this 'invention' was a legitimately good one. But it didn't seem to really take off, I am assume because standby was rapidly implemented in a way where there is such minimal electrical draw that it would be pointless.
I had a couple of these (in fact I probably still have them), from the 2000s - an Intelliplug and Intellipanel - the later being an 8-way plug. I used it for a TV, 5.1 amp/receiver, blu-ray player, Xbox360, Sky+, Wii, and Squeezebox Radio. The Sky+ box was in the "always on" socket so it could record stuff at all times, while the TV was in "if this comes on, turn on everything". I convinced myself I was doing a good thing for the planet and my finances, but I never measured it. Blindly believing it was far preferable to spending any effort on proving I'd been a fool, especially since I needed the sockets anyway - I don't recall the price being that much different to any other 8-way I looked at.
Eco-friendly or not, I'm impressed; detecting power draw and using that as a signal is actually a legitimately clever solution to the problem of wanting to conveniently switch multiple things at once. The only caveat I can think of is that some devices probably don't like being switched at the wall.
I've found these devices to be finicky. You have to set the threshold of current flow that matches your TV, and sometimes switching from one input device to another will spike power consumption low enough to turn things off in a way you don't want.
I thought the same, and my Smart Sony TV claims very low standby consumption, but when it is on the Wifi and sat there ready to Airplay too it consumes something close to 27W all night long.
I setup a smart plug to turn it off at night on a schedule.
Also seen that my Hifi Amp draws a good few watts on standby.
We have recurring argument in my house where my partner tells me I should have turned the LED lights off to save electricity and I tell them to close the curtains to keep the heat in.
In truth I have no idea which wastes more energy/money
I used Keepass + Dropbox (to sync database). This set up was suggested to me when I joined a company that requires complex unique passwords for all sign ups. At the time I didn't think too much about it, but this thread has made me thankful I was guided in this direction.
I don't have experience of any other password management software, so I certainly can't compare and contrast. But I will say Keepass + Dropbox has worked flawlessly for me across desktop, laptop and mobile. The biggest inconvenience I have had is things like manually typing in a Netflix password into a Smart TV when on holiday (just takes time with long passwords with capitals, lowercase, numbers and symbols).
Netflix is a problem for every password manager, since they insist on using their own onscreen keyboard. Even if you’re on Android TV, you can’t switch to the remote app to copy & paste a password. Very annoying, but I’m starting to think they’re doing this to make password sharing more of a bother. (Even if it’s just one sign-in being more of a chore)
My partner and I took a year out for all of 2018 and travelled for 365 days, aged 34. We went to 21 countries during that time, the route:
UK
India
Hong Kong
Vietnam
Cambodia
Laos
Thailand
Russia
Mongolia
China
Tibet
Japan
Malaysia
Indonesia
Australia
New Zealand
Fiji
Hawaii
North America
Chile
Bolivia
Peru
Colombia
Brazil
We saw many wonders of the world (from various lists) including Taj Mahal, Angkor Wot, Great Barrier Reef, Machu Picchu and Christ the Redeemer.
We drove coast to coast in the US (LA to New York via the South) over a month.
We took the Trans-Siberian Express train across Russia to Beijing (St Petersburg, Moscow, Kazaan, Yekaterinburg, Novosibirsk, Irkutsk, Lake Baikal, Ulaanbaatar, deeper into Mongolia and stayed in a Gerr and Beijing).
We also looped around China by train including going to Tibet via Xinning. Tibet is like nowhere else we have been in the world.
We spent a couple of weeks driving around New Zealand South Island in a campervan.
We did the classic backpacker route around South East Asia, traveling overnight on sleeper buses, visiting the tourist destinations, site seeing, going on a booze cruise party island in the middle of a site of natural beauty (Castaways), visiting temples, full moon party. All in all a bit of everything. We were at the absolute age limit of some activities and accomodation, where although they were willing to show some flexibility, you wouldn't want to be much older as the majority of people were about 20. At times we felt old and past it, but we were happy we went ahead and did it.
Whilst a lot of the trip we just planned and booked ourselves, small parts, like Australia, we went on an organised trip. It wasn't one where we follow a tour leader, instead we were given vouchers for accomodation and activities and transport that went from Cairns down to Melbourne and followed it ourselves. It included dolphin feeding, scuba diving at the great barrier reef, driving around Fraser Island on 4x4s and staying a couple of nights, going on a boat tour of the Whitsundays and seeing White Haven beach with the softest sand I have experienced.
We could have done a fraction of the amount of stuff we did, and seen more or each place. We could have recouperated more rather than end up exhausted. But it was one hell of a trip.
I worried about finding work when it was over, but had several job offers before I returned so that turned out to be a false worry!
It took a while to plan, not just the trip stuff but finishing a job, packing up a house, signing cars off the road, cancelling bills etc.
We got a round the world flight ticket, we had to commit to the first six month flight times then half way through commit to the rest. It was possible to change for a fee. But we found the flights gave us a schedule to work around.
Some parts we planned extensively (you can't go to Tibet without being on a tour the whole time you are there, Machu Picchu books up months in advance, etc) other parts (Vietnam, Cambodia, Laos, Thailand) there is somewhat of an established back packer route you can follow and just make sure you get through in time for the next flight.
By far the least safe place was South America, we met many travellers there who had been mugged, beaten up, etc. It was our last destination and we were not very adventurous as we were tired and also just wanted to get to the end without an incident.
Other than that, we generally felt safe, there were one off isolated incidents (drunk Russian man on train trying to make advances on my partner when he thought I was asleep) which were uncomfortable, or going to a train station in Jaipur at 3 in the morning and waiting for a train, with lots of people loitering around and bothering us, and no real security or staff to reassure us, but these were exceptions rather than the norm.
We typically went to places which had a reasonable number of travellers, so it didn't feel too crazy in that sense.
Budgeted £42k ended up costing £60k for us both. I know some people which did it for a lot less (like half) but we didn't want to miss out on various things as we went.
We stayed in a lot of hostels, along with some guest houses, Airbnbs, hotels but the majority of the time it was something in the cheaper end of the spectrum.
We initially budgeted £10 food, £5 drink and £5 activities per day, and kept quite a lot of records at time, but ultimately we just ended up feeling "it is what it is". We made the active decision to just make the most of the trip and overspend. For a reasonable number of days we could average out towards our budget but we found fairly regularly there were bigger expenses for activities which would blow the budget and we neither wanted to miss them or do nothing for the next fortnight! Once you have committed so much, it sort of feels like it is better to commit a bit more and make it awesome (£60k spent on awesome seemed better than £42k spent on okay).
Before leaving I ran a couple of companies, one which employed 12 people and another which was contractor based. I wasn't wealthy (at least not by the standards set in the UK if you speak to a financial advisor) and £60k was a large amount of money for me to spend (a very large amount). But all things going well with life, you earn it, you spend it, we decided we wanted to do some of that spending earlier in life whilst we could enjoy it.
Since then... I have had heart ops, a blood clot on the lung, my partner has developed ME/chronic fatigue syndrome... COVID19 has engulfed the world. Needless to say we are glad, thankful and lucky we did what we did when we did!
In general most developers struggled to do much with it, it was just too small (combined with the fiddlyness of using it).
PS2 programmer's were very used to thinking in this way as it's how the rendering had to be done. There is a couple of vector units, and one of them is connected to the GPU, so the general structure most developers followed was to have 4 buffers in the VU memory (I think it only had 16kb of memory or something pretty small), but essentially in parallel you'd have:
1. New data being DMAd in from main memory to VU memory (into say buffer 1/4). 2. Previous data in buffer 3/4 being transformed, lit, coloured, etc and output into buffer 4/4. 3. Data from buffer 2/4 being sent/rendered by the GPU.
Then once the above had finished it would flip, so you'd alternate like:
Data in: B1 (main memory to VU) Data out: B2 (VU to GPU) Data process from: B3 (VU processing) Data process to: B4 (VU processing)
Data in: B3 Data out: B4 Data process from: B1 Data process to: B2
The VU has two pipelines running in parallel (float and integer), and every instruction had an exact number of cycles to process, if you read a result before it is ready you stall the pipeline, so you had to painstakingly interleave and order your instructions to process three verts at a time and be very clever about register pressure etc.
There is obviously some clever syncing logic to allow all of this to work, allowing the DMA to wait until the VU kicks off the next GPU batch etc.
It was complex to get your head around, set up all the moving parts and debug when it goes wrong. When it goes wrong it pretty much just hangs, so you had to write a lot of validators. On PS2 you basically spend the frame building up a huge DMA list, and then at the end of the frame kick it off and it renders everything, so the DMA will transfer VU programs to the VU, upload data to the VU, wait for it to process and upload next batch, at the end upload next program, upload settings to GPU registers, bacially everything. Once that DMA is kicked off no more CPU code is involved in rendering the frame, so you have a MB or so of pure memory transfer instructions firing off, if any of them are wrong you are in a world of pain.
Then throw in, just to keep things interesting, the fact that anything you write to memory is likely stuck in caches, and DMA doesn't seem caches, so extra care has to be taken to make sure caches are flushed before using DMA.
It was a magical, horrible, wonderful, painful, joyous, impossible, satisfying, sickening, amazing time.