https://launchpad.net/ubuntu/p...

Source packages are on that page, follow the links to the builds for the binaries.

IIRC at one stage it specified vorbis/theora as a baseline which every implementation should support but under pressure from apple and MS they took that out.

Depends where in the UK you are.

Some areas have competion between virgin meda cable and openreach FTTC. At the other extreme some areas have no cable and phone lines so long/shitty that they can't handle DSL at all. Many providers have caps or "fair usage policies", especially in areas that don't have LLU.

Maybe maybe not. Once a format is deemed "good enough" it can stick around for a long time. See mp3, jpeg png etc. Furthermore bandwidth prices have dropped through the floor in recent years,

Yes we can, the combination of simplicitly and ubiquity mean it is highly unlikely we will lose the ability to read it.

UTF-8 is a little more complex but the encoding method can still be described in less than a page, the harder bit is what to do with the code point sequence you get from decoding but for most widely used languages* that is a simple table lookup.

Well you might end up with swapped case and block-drawing would be a mess but you could read the actual text by just treating the file as ASCII.

* The exceptions being languages like hebrew, arabic and some indian languages.

How well does word with odf actually work? does saving from word as odf and opening it in libreoffice have a higher or lower chance of screwing up the document than saving as doc and opening it in libreoffice?

Which is WHY it's important that big guys are doing this.

When you are the little guy there is a lot of pressure on you to conform to the standards set by those you work with (and that may mean not just using MS office but using a specific version of MS office), when you are the big guy you SET the standards and require other people to conform to them.

There are legitimate reasons for asymetry on DSL and cable

On DSL upstream and downstream have to be given seperate frequency slices out of the limited bandwidth available on a typical phone pair (which lets not forget was only designed to carry voiceband). So you have to tradeoff upstream speed and downstream speed and for most users it makes more sense to tradeoff towards downstream. Having said that I do think it's scandalous that symetric services are insanely expensive compared to asymetric ones of comparable total bandwidth.

On cable the technical reasons are even greater, cable networks are designed for broadcasting TV with a high power transmitter broadcasting through the high-loss (due to the splitting/padding) network to a lot of receivers. Upstream traffic is going against the flow which means it has a lower acceptable transmit power and a lot more interference present at the receiver.

On the other hand with fiber the only reason for the asymetry is artifical crippling (making it harder to use P2P, run servers etc)

Depends on what you see as "the whole point of using a VPN".

Afaict there are three main reasons to use a VPN

1: you don't trust the provider of your internet connection
2: you need to access IP-locked resources on the internet
3: you need to access resources on a private network that is not directly reachable from the internet.

"Split tunnel" kills reason 1 and probablly also reason 2 (unless there is some complex routing configuration in place). It certainly does not kill reason 3 which is often the main reason for using a VPN.

On the other hand forcing everything down the VPN kills the ability to use resources on your local network (a PITA if you use a network printer) and means traffic to the internet is wastefully forced to take a roundabout route to it's destination.

From what I can gather both comcast and verizon bullied netflix into paid peering by refusing to expand peering with any carrier netflix used or tried to use as an upstream.

When netfllix paid up to comcast they got massive improvments in connectivity to comcast customers, when they paid up to verizon they didn't.

http://hardforum.com/showthrea...

I don't know about the amiga specifically but some computers have a memory backup battery mounted on the main PCB. If they leak and the leakage is not noticed and dealt with quickly it can cause severe damage to PCB traces which can be a nightmare to repair.

Until you have to deal with an old machine that has one of the buggy SATA1 controllers that falls over if you plug in a SATA2 drive.

Some drives had a jumper to force SATA1 mode but that has dissappeared on more modern drives.

AFAICT oil isn't such a big issue because it's routinely shipped around the world, so unless there is noone for russia to sell it's oil to oil sanctions between russia and europe won't change things much. Europe will pay slightly more, russia will get slightly less. Other countries and trasnportation companies will profit.

Gas is the big issue because it has traditionally been moved by pipeline. Moving it by ship requires special terminals to purify and liquify it and special ships to carry the cryogenic liquid. The US currently has a glut of gas but moving that gas to europe will mean the building of more LNG terminals and ships which takes time.

IIRC 10 years is the max on com/net/org

How it could happen is pretty simple, someone is working on a new service, they are in a hurry and just buy the domain with a company credit card or a small one time PO or whatever putting their individual work email address as the contact info. They register it for a few years, maybe even the maximum of 10. Maybe they set a reminder for themselves to renew it, maybe they don't bother as they think it unlikely the domain will stay in use that long.

The project grows in importance but noone notices that the domain behind it is associated with one employee, then that employee becomes an ex-employee and their email is shut down

The problem is what the customers purchased is generally a connection to the internet with no particular gaurantees about performance. If you want connections with service level agreements coverting performance to defined locations (e.g. major peering connections) you can get them but expect to pay a hell of a lot more than you would pay for a regular "broadband" connection.

Since they never agreed to provide any particular ammount of bandwidth in the first place there is little to stop them taking away some of the bandwidth they currently give to "best effort IP" to reallocate it to premium services. Whether they do that statically by creating fixed bandwidth channels or dynamically through prioritisation doesn't really make a fundamental difference.

When the "best effort IP" service is the entire service it's in the provider's interest to make it not suck so they retain customers. OTOH when they offer both "best effort IP" and premium services it's in their interests to make the "best effort IP" service suck so they can sell more premium services (which may or may not be IP based).