Video streaming is getting cheaper all the time. Bandwidth costs are dropping every year (substantially in most cases) and bitrates aren't keeping up (the few exceptions probably being AppleTV+ and BBC iPlayer which do 30-50mbit/sec 4K HDR streams).
You can do this for so much cheaper than AWS etc price for bandwidth. You can get 100gigE transit from he.net for list price $4500/month. Add probably the same again for colo + hardware (don't need much hardware these days to saturate 100gigE) and you can probably stream videos to 20,000 concurrent users at 5mbit/sec for ~$10k/month.
Another way would be to use someone like OVH who offers dedicated servers with 10gigE (supposedly 'guaranteed') for about $800/month each list price, without having to bother with colo and ip transit setup.
Obviously this is highly simplified as you will require encoding resource and storage, but again with someone like OVH you'd be able to spin up a lot of cheap boxes to do this. How much this will cost will depend on how many videos you get and how many views per video etc.
So IMO the actual bandwidth is a bit of a non issue. The far bigger issue is getting users to use your platform (marketing is MUCH more expensive than IP transit) and then having advertisers on your site. This is a much harder problem to solve and where the real barrier to entry is.
Yes exactly. In the UK power is becoming increasingly negative in pricing when there is high wind + solar output. This is actually increasingly alarmingly rapidly (and there is 3-5GW of offshore wind, plus loads of utility scale solar coming online).
Prices are then very high when wind and solar is low - which happens to be when demand is the highest (cold weather snaps in winter which tend to result in very low windspeeds).
National Grid is already paying £1bn/yr to turn off wind farms when supply is too high (plus paying a fortune for new nat gas peakers, which are limited by law to run for 10 days a year max). It's projected that curtailment payments to wind farms will reach £4bn/yr.
While some of this will be rectified with more transmission capacity (there is a 4GW offshore HVDC link being built between scotland and england), if the claims of terraform are true and hold up at scale, I think this is the actual breakthrough people have been looking for.
These could be connected to substations near wind farms (which also happen to be near major gas interconnectors from the north sea) and generate when power prices were low or negative, which will be a large amount of the time. They'd then get paid not only for the arbitrage in gas prices but also they would be able to take (most/all) of the curtailment payments national grid is paying the wind farms.
To be clear batteries do not work particularly well for a market like the UK. Batteries work well for overnight storage of solar, they do not work well for northern climates like the UK that require weeks of storage of power to cover low renewable output in winter. That's not to say there isn't loads of batteries being constructed right now, there is, but it's to cover very short term movements in supply and demand - the much harder problem is covering days or weeks of low output.
Because some of the windfarms still run on legacy contracts where they were incentivized by a guaranteed selling price for the power they produce. They are reimbursed for the loss of income they suffer due to curtailment.
Correct. The UK regulator (Ofgem) and thus UK consumers are being taken for a ride by windfarmers, who made these deals a precondition of building farms.
We will be paying them £2.5bn a year to not generate electricity by 2030
I like the theory someone else offered of other people "using up" the numbers o pre-existing stamps which would be feasible if the numbers were predictable.
They deduplicate and compress before storing, so it may end up only taking 1.5TB.
Still, that's $375/month. Nothing to shake a stick at.
I think the intended use case of tarsnap is for <100 GB raw data that is really important to you, for which you must have an offsite backup that is (according to them) insanely backed up and safe. I could see myself using them for a curated folder of personal documents I can't afford to lose in the event of a house fire. Then again, I could probably fit all my most important documents into my 1Password vault.
from my understanding of the tarsnap client, it performs the deduplication then encryption locally to achieve high compression. still the cost is too high for my exact use case.
This is just totally incorrect. Total rail subsidy in the UK is £11bn/yr. Ticket sales are another ~£8bn/yr Total TOC profits are £100m/yr. Rolling stock operating companies take maybe £200m/yr in profit but it varies.
So TOC/ROSOC profits 'take out' 1.5% of the money in the system. Saying they are 'mostly' a profit making enterprise is completely ridiculous.
Also, while the UK has privatised TOCs, Germany and other countries are also opening regional/long distance rail routes to franchising of sorts. National Express (a British company) operates a surprising amount of routes (and growing) in North Rhine-Westphalia (and probably other regions) for example. It's not just a one way thing.
After spending a lot of time in Europe, the grass definitely isn't greener.
Trains in France are not vastly cheaper than the UK, but the service pattern is awful outside the very main routes. Often huge gaps in service for parts of the day and non-clockface timetabling. Seems to be very poor utilisation of rolling stock, with a lot of the stations having trains sitting for hours doing nothing (which should really be operated more intensely to give a better service pattern).
DB has horrendous reliability problems, basically the entire network gets something similar to TPE on time performance (the worst performing TOC in the UK). ICE/IC trains are also not particularly cheap for on the day travel.
Spain's high speed lines are excellent and very cheap. But outside that the network is incredibly limited and slow, so much so that buses will almost always beat the train in journey time.
Netherlands is good, affordable, frequent services and reliable. I think NL is the only country I've been to where the system is noticeably better than the UK across a lot of dimensions.
As a regular train traveller in Germany, it averages out. It's almost exclusively the intercity connections which are delayed, but those usually run on an hourly clock. It's actually not that uncommon that I can save half an hour of travel time because I can catch a delayed ICE from the previous time slot instead of waiting 45 minutes for the next train ;)
You cannot possible rate the Netherlands good on comfort. It's cattle class, even on 'long' (for Dutch standards) haul routes. They _just_ started to operate carriages with power sockets! I rode better trains in Poland 10 years ago than in NL.
I take el cheapo French TGV (Ouigo) over any Dutch rolling stock any day.
Intercity trains have had sockets for a while, but YMMV if this wasn't on the most used tracks.
Short distance or trains that stop at towns (called sprinters) didn't. Haven't ridden a train in quite a few years so that might have changed.
The trains are mostly on time though, which is something most countries fail at. Granted Dutch people still complain about public transport, me included. Most people think it's too expensive and it takes too long to really get anywhere that isn't a direct connection. These days it is often cheaper to go by car too, especially if you are going somewhere with more than one person.
Depending on the article the Netherlands seems to be either high in the top 10 or low in the top 20.
Checked the latest WEF report I could find (2021), and it's ranked 14 on there. But that includes much more than just public transport. So not that useful.
(Travel & Tourism Development or T&T index).
NS uses a wilde variety of materiel for intercities, only a handful of new ones now that finally come with sockets. Never had a single one here at Maastricht. I did have a sprinter on intercity duty last time I went up north.
from the handfull of times I've taken Ouigo, it's just normal TGV, just without first class and sometime less convenient train stations (like outside of Paris instead of close to the center)...
Ouigo comfort is far less than any Dutch IC-class stock. Maybe the Ouigo is on par with Sprinter-class stock, but even then it’s a stretch.
Compared to German stock — yes, the German stock is nicer (mainly has to do with different usage patterns, though — Netherlands train network has been said to resemble a country-sized metro network).
Outlets — who cares, just bring a power bank — again, compare to networks like London Overground or Paris RER.
That doesn't make any sense. The UK runs 2x the services on a network of less than 50% of the size of France - so 4x the services per route km. Believe me the UK network is riddled with horrendous chokepoints too.
Most of these routes have about 1-2tph running on them max. The UK (and I'm sure many other countries) manages to run clockface timetabling (for the most part) with a lot of routes on 10tph++.
It's not the lack of clockface timetabling that is the problem per se, it's the constant huge gaps in service. Eg approx 1tph, then a big gap of 3 hours, then 1tph. It's all over the place.
But France is on the "continent" so has to manage transcontinental traffic, especially freight trains - which are very long, very heavy and very slow.
But no, it's so easy for you to just shout "it doesn't make any sense" without first wondering if you had the complete picture. The French are incompetent and wasteful, that's obvious isn't it?
Two wrongs don't make a right. If a particular taxpayer-funded road was one cent per mile for the end user as opposed to other higher-priced roads, you wouldn't say, "wow, this road is reasonably priced!" Unless you're literally just doing a travel blog
There's a lot of externalities around transit that aren't directly priced into costs vs ticket revenue.
If France decides "We want high speed national rail connectivity between cities", I'd look at it more of an entitlement / service than a profitable enterprise.
Nobody expects national healthcare services to be profitable.
(Also, both France and Germany's relatively recent experience with their national rail networks being the reason their countries are still independent sovereign states)
They want it so much that they're unwilling to pay the actual cost of a ticket lol. They have to launder the money through an intermediary and offload the cost to people who don't use the train that everybody supposedly wants
Germany and France were on the verge of collapse or being conquered but government trains saved them? Do tell.
But also, they're maybe not very sovereign when a foreign body drafts, ratifies, and enforces their laws
If everyone generally agreed that injecting messenger RNA into the eyeballs of infants provided a social benefit you would go along with it.
So you're saying that if the Nazis didn't have subsidized train rides for tourists they would not have been able to collapse the Soviet Union in 1991 and regain East Germany? Interesting. Very interesting, sir.
Coal is not really very suitable for replacing renewables, natgas is much better suited.
The UK is building/planning a lot of natgas peaker plants at the moment, seemingly without anyone realising. One I read about is limited (for emissions reasons) to 240hours/10 days operation per year, which suggests that the price of power when renewables aren't going is going to be extremely high to make it profitable to build a plant that only runs 3% of the time.
> is going to be extremely high to make it profitable to build a plant that only runs 3% of the time.
It's probably fine, as the cost of renewable electricity is now lower than the fuel cost (aka the variable cost) of these plants.
In a lot of places, it's now cheaper to run a gas plant with renewables than without, as the build costs of the renewables is lower than the cost of the fuel they prevent to get burned.
It's definitely not true to say they are the drugs of last resort in the UK. They may not be the drug of 'first resort' like they were in the US, but there is still an awful lot of people getting Oxycodone etc via the NHS, especially in certain areas (tend to be rural and deprived... very similar to the US): https://ichef.bbci.co.uk/news/976/cpsprodpb/166DF/production.... The top regions on that graph is actually higher than the US rate (https://www.cdc.gov/drugoverdose/rxrate-maps/index.html)
You can do this for so much cheaper than AWS etc price for bandwidth. You can get 100gigE transit from he.net for list price $4500/month. Add probably the same again for colo + hardware (don't need much hardware these days to saturate 100gigE) and you can probably stream videos to 20,000 concurrent users at 5mbit/sec for ~$10k/month.
Another way would be to use someone like OVH who offers dedicated servers with 10gigE (supposedly 'guaranteed') for about $800/month each list price, without having to bother with colo and ip transit setup.
Obviously this is highly simplified as you will require encoding resource and storage, but again with someone like OVH you'd be able to spin up a lot of cheap boxes to do this. How much this will cost will depend on how many videos you get and how many views per video etc.
So IMO the actual bandwidth is a bit of a non issue. The far bigger issue is getting users to use your platform (marketing is MUCH more expensive than IP transit) and then having advertisers on your site. This is a much harder problem to solve and where the real barrier to entry is.
reply