You have no idea what langauge your customer speaks, or if they speak many languages including the language your software supports.
Well, I'm going to stop you right there, because marketing is a thing and finding out what your market wants is a rather large part of that thing. So in fact you often have a very good idea of what language(s) your customers speak.
What makes you think you're required to support more than one language?
Erm... The fact that I have done this, multiple times, with large scale projects, and I have seen detailed data on both market research before the fact and the effects that offering different translations had on sales in different locations after the fact?
Some people will be happy using some products or services supplied in a language other than their native tongue. In particular, many people in Europe speak English as a second language and will consider buying some things that are only available in English. However, not everyone does, and depending on the nature of your product or service, you might still increase sales significantly if you can offer it in your target audience's native language(s) as well.
Now, if you go ahead and offer said translations available for sale, offering differing price points for different locales most definitely should be illegal. If you cant set your price at a point that encompasses all of the labour involved in creating it, then you simply misunderstand business, where the rule is always "Charge everyone more".
No, it isn't. In particular, it may make economic sense to offer your product for sale in, say, English, at a high price point that is accessible to western European economies but not to most customers in less wealthy eastern Europe. Given that you can't viably offer it at the same price point in eastern Europe, it may still make commercial sense to offer it to that market at a lower price point that they can afford, possibly translated to increase the size of the market, rather than giving up on sales to that region altogether.
This is what the EU is now trying to prohibit. Unfortunately, the EU is doing so in the hopes that western Europe will wind up paying the same lower price that eastern Europe might be offered. In reality, the reverse is far more likely, resulting in both the businesses offering useful products/services and the potential customers for those products/services in less wealthy EU member states losing out and absolutely no-one gaining.
When people criticise the EU for not understanding business, this is exactly the kind of ideologically driven loss they are talking about.
I'm not unsure about it at all. I've worked on a variety of professional projects, including software/UIs, written content and even video material, where translation has been an issue. Sometimes it's a significant expense. Sometimes it is relatively modest compared to the overall cost of the product or service. That's why I wrote "could".
Your scheme of charging different amounts for different languages would likely attract some legal action from the EU.
On what basis? Translating a work from one language to another could be a very expensive undertaking. This EU policy is already economically naive, but expecting all translations of works to be provided at the same cost would be economically absurd.
It will presumably still be possible for content providers to supply content to EU nations, unless they have existing exclusive licensing deals with intermediaries from some of those nations already that would now be in competition with their own offering intended for another nation but now applicable EU-wide.
What this will do is mean that selling at a lower price in the less well-off nations is no longer acceptable, so the result is likely to be that some content will now only be legally available to those in the richer EU member states (and piracy will presumably rise in the others for content that was popular there).
As so often happens, the EU is putting its idealised principles ahead of pragmatism and the consequence will be shafting businesses and/or a significant part of its own population.
I don't see why. I mean, just set some basic laws that say AIs can't harm humans, have to do what they're told, and shouldn't harm themselves either, and then how bad could the results be, really?
The only things I have seen failing were custom software jobs
Then you have been very lucky. Unfortunately, not everyone is. I've had to reinstall entire machines because of things as stupid as bad updates to malware signatures for the security software that wound up quarantining/removing critical files so the system would no longer boot, for example.
It was huge organisations that are widely reported as being hit. It's more obvious when a big organisation takes a hit. But small organisations have been hit as well, and in any case the advice about whether or not to install updates is being repeated all over the place without reference to organisation size.
There are plenty of IT consultancy businesses that will stand in for an in-house IT group.
There are also plenty of small businesses who aren't tech experts and have no idea why they would need such a service or how to judge who can competently provide it. Most people have absolutely no idea how crazily bad most software is.
You "need" a lot of things even in a small business. Plenty of problems can kill a young business before it becomes established at all. The reality is that you almost certainly won't be able to deal with some of the issues for a while, and you have to prioritise and do the best job you can in the meantime.
On unmanaged systems, we install the security-only rollups, not the all-in ones that you get through Windows Update. As far as we're aware, the security-only bundles don't include the telemetry malware. If you know better, please cite, because finding detailed information about exactly what each of the new monthlies includes is often a pain.
That's strange. I'm sure I've worked in several different organisations with 25-50 people and no dedicated IT staff, yet they all managed to keep their systems working just fine.
Oh, wait, that was before the modern updates-every-ten-minutes junk. Never mind.
If you mean how could two people have the same email address, obviously they shouldn't, but many mail address assignment schemes are vulnerable to name/initials clashes and in practice result in one person with a similar name getting sent someone else's mail from time to time. There are also problems like domains being given up and then subsequently reregistered by someone else, for another example.
If you mean how could one person's email address be represented in multiple ways that map onto the same underlying mailbox, case sensitivity and non-Western character sets are two common examples.
His arguments against using biometrics as identifiers were the birthday paradox and fuzzy matching, which absolutely don't apply at all to email addresses.
I'm sorry, but they most certainly do apply to email addresses.
For the birthday paradox, haven't you ever worked in an office that had addresses of the form j.smith@example.com, until both John and Jane Smith joined, and then suddenly their scheme broke down? Technically speaking everyone presumably has their own address according to some alternative scheme invented to avoid that problem, but in reality anyone who is familiar with the original scheme is quite likely to send mail to the wrong address.
For fuzzing, you might consider something as simple as upper vs. lowercase: technically, domains aren't case-sensitive but the local part before the @ may or may not be depending on the rules of the local system. As another example, local systems may allow multiple representations in regions with non-Western character sets.
Good identifiers should have uniqueness guarantees, biometrics don't. Good identifiers should always either match or not match, biometric matching is fuzzy, every match is a judgement call.
You make good points, here and throughout your post. However, I don't think the above undermines the biometrics-as-IDs analogy to the extent that you're arguing here. A great many online systems today use an email address as an ID, yet email addresses can suffer from exactly the same problems. We use a person's name and mailing address to send them post, but again the same problems can arise. In practice, not many IDs that we use are good identifiers by your definition -- and again, I'm not disputing that definition -- but we continue to use them, mostly effectively. In most contexts, biometrics can serve the same purposes for the same reasons.
Treating them as passwords is a popular idea but will inevitably lead to disaster. Who would choose a password they could never change and then give that same password to countless other parties? Even if we did that, what would be the equivalent to good practices like storing password hashes instead of the originals in case of compromise?
Don't worry, thanks to the higher security in modern Windows versions and the vulnerability being discussed elsewhere today, a hacker can now install a virus automatically without even troubling your users to open their mail!
Exactly. Email is a nice, stable standard that mostly Just Works. An email client can be the same, and that will do just fine if the alternative is everything breaking every five minutes because someone couldn't follow a standard or didn't want to bother maintaining the ancient code they wrote last month.
Once all the major browsers restrict what you can do with plugins, and require all plugins be signed so there is top-down control of the whole ecosystem, you may as well give up.
Or we could stop just accepting self-updating software that doesn't understand the difference between an important security patch, expendable changes in functionality or UI, and changes that actually reduce functionality and make things not work when they worked before. The latter is an area where browsers have been particularly awful for some time.
There are already projects attempting to move in this direction based on the Firefox code. Perhaps we should do more to support them and raise their profile.
I suspect that would be true if it were only Netflix, as there are other video streaming services. But if it started to be, say, Netflix and a couple of the other big streaming services and one or two social networks a household uses (which still work fine on everyone's phone) then it might be a different story.
The curious thing about this stand-off is that if the content providers are right, and net neutrality really is necessary to ensure a level playing field in markets where competition between ISPs is insufficient, then they might not have to spend their money lobbying to the same degree. If heavy traffic users like Facebook and Netflix call the ISPs' bluff and make a public statement that under the new arrangements they will no longer offer access to their services via US ISPs that require additional fees to transport the traffic, who is going to get the blame?
There are ways to set up DRM that don't rely on single providers like that. In fact, you pretty much have to if you're distributing content on the Web, because no single DRM scheme works reliably on all popular platforms.
Of course, if you actually are a large creator/distributor like Netflix, it certainly would be viable to implement your own scheme if you wanted to as well.
Slashdot Mirror
You have no idea what langauge your customer speaks, or if they speak many languages including the language your software supports.
Well, I'm going to stop you right there, because marketing is a thing and finding out what your market wants is a rather large part of that thing. So in fact you often have a very good idea of what language(s) your customers speak.
What makes you think you're required to support more than one language?
Erm... The fact that I have done this, multiple times, with large scale projects, and I have seen detailed data on both market research before the fact and the effects that offering different translations had on sales in different locations after the fact?
Some people will be happy using some products or services supplied in a language other than their native tongue. In particular, many people in Europe speak English as a second language and will consider buying some things that are only available in English. However, not everyone does, and depending on the nature of your product or service, you might still increase sales significantly if you can offer it in your target audience's native language(s) as well.
Now, if you go ahead and offer said translations available for sale, offering differing price points for different locales most definitely should be illegal. If you cant set your price at a point that encompasses all of the labour involved in creating it, then you simply misunderstand business, where the rule is always "Charge everyone more".
No, it isn't. In particular, it may make economic sense to offer your product for sale in, say, English, at a high price point that is accessible to western European economies but not to most customers in less wealthy eastern Europe. Given that you can't viably offer it at the same price point in eastern Europe, it may still make commercial sense to offer it to that market at a lower price point that they can afford, possibly translated to increase the size of the market, rather than giving up on sales to that region altogether.
This is what the EU is now trying to prohibit. Unfortunately, the EU is doing so in the hopes that western Europe will wind up paying the same lower price that eastern Europe might be offered. In reality, the reverse is far more likely, resulting in both the businesses offering useful products/services and the potential customers for those products/services in less wealthy EU member states losing out and absolutely no-one gaining.
When people criticise the EU for not understanding business, this is exactly the kind of ideologically driven loss they are talking about.
I'm not unsure about it at all. I've worked on a variety of professional projects, including software/UIs, written content and even video material, where translation has been an issue. Sometimes it's a significant expense. Sometimes it is relatively modest compared to the overall cost of the product or service. That's why I wrote "could".
Your scheme of charging different amounts for different languages would likely attract some legal action from the EU.
On what basis? Translating a work from one language to another could be a very expensive undertaking. This EU policy is already economically naive, but expecting all translations of works to be provided at the same cost would be economically absurd.
It will presumably still be possible for content providers to supply content to EU nations, unless they have existing exclusive licensing deals with intermediaries from some of those nations already that would now be in competition with their own offering intended for another nation but now applicable EU-wide.
What this will do is mean that selling at a lower price in the less well-off nations is no longer acceptable, so the result is likely to be that some content will now only be legally available to those in the richer EU member states (and piracy will presumably rise in the others for content that was popular there).
As so often happens, the EU is putting its idealised principles ahead of pragmatism and the consequence will be shafting businesses and/or a significant part of its own population.
I don't see why. I mean, just set some basic laws that say AIs can't harm humans, have to do what they're told, and shouldn't harm themselves either, and then how bad could the results be, really?
The only things I have seen failing were custom software jobs
Then you have been very lucky. Unfortunately, not everyone is. I've had to reinstall entire machines because of things as stupid as bad updates to malware signatures for the security software that wound up quarantining/removing critical files so the system would no longer boot, for example.
It was huge organisations that are widely reported as being hit. It's more obvious when a big organisation takes a hit. But small organisations have been hit as well, and in any case the advice about whether or not to install updates is being repeated all over the place without reference to organisation size.
There are plenty of IT consultancy businesses that will stand in for an in-house IT group.
There are also plenty of small businesses who aren't tech experts and have no idea why they would need such a service or how to judge who can competently provide it. Most people have absolutely no idea how crazily bad most software is.
You "need" a lot of things even in a small business. Plenty of problems can kill a young business before it becomes established at all. The reality is that you almost certainly won't be able to deal with some of the issues for a while, and you have to prioritise and do the best job you can in the meantime.
On unmanaged systems, we install the security-only rollups, not the all-in ones that you get through Windows Update. As far as we're aware, the security-only bundles don't include the telemetry malware. If you know better, please cite, because finding detailed information about exactly what each of the new monthlies includes is often a pain.
Windows 7 and 8 also include essentially the same telemetry now.
None of my systems do. Oh, wait, that's because with previous Windows versions I could just choose not to install that crap in the first place.
You do understand that the majority of professional work is done by small businesses, and most of those don't have dedicated IT teams at all, right?
Enterprise IT is actually the exception, not the norm.
You need dedicated people or else shit breaks.
That's strange. I'm sure I've worked in several different organisations with 25-50 people and no dedicated IT staff, yet they all managed to keep their systems working just fine.
Oh, wait, that was before the modern updates-every-ten-minutes junk. Never mind.
(You posted that text, or something very like it, in another comment here on Slashdot, and I addressed your point in a reply there.)
If you mean how could two people have the same email address, obviously they shouldn't, but many mail address assignment schemes are vulnerable to name/initials clashes and in practice result in one person with a similar name getting sent someone else's mail from time to time. There are also problems like domains being given up and then subsequently reregistered by someone else, for another example.
If you mean how could one person's email address be represented in multiple ways that map onto the same underlying mailbox, case sensitivity and non-Western character sets are two common examples.
His arguments against using biometrics as identifiers were the birthday paradox and fuzzy matching, which absolutely don't apply at all to email addresses.
I'm sorry, but they most certainly do apply to email addresses.
For the birthday paradox, haven't you ever worked in an office that had addresses of the form j.smith@example.com, until both John and Jane Smith joined, and then suddenly their scheme broke down? Technically speaking everyone presumably has their own address according to some alternative scheme invented to avoid that problem, but in reality anyone who is familiar with the original scheme is quite likely to send mail to the wrong address.
For fuzzing, you might consider something as simple as upper vs. lowercase: technically, domains aren't case-sensitive but the local part before the @ may or may not be depending on the rules of the local system. As another example, local systems may allow multiple representations in regions with non-Western character sets.
Good identifiers should have uniqueness guarantees, biometrics don't. Good identifiers should always either match or not match, biometric matching is fuzzy, every match is a judgement call.
You make good points, here and throughout your post. However, I don't think the above undermines the biometrics-as-IDs analogy to the extent that you're arguing here. A great many online systems today use an email address as an ID, yet email addresses can suffer from exactly the same problems. We use a person's name and mailing address to send them post, but again the same problems can arise. In practice, not many IDs that we use are good identifiers by your definition -- and again, I'm not disputing that definition -- but we continue to use them, mostly effectively. In most contexts, biometrics can serve the same purposes for the same reasons.
Biometrics aren't passwords, they are user IDs.
Treating them as passwords is a popular idea but will inevitably lead to disaster. Who would choose a password they could never change and then give that same password to countless other parties? Even if we did that, what would be the equivalent to good practices like storing password hashes instead of the originals in case of compromise?
I got Windows Defender ... installed.
Oops. :-(
(Seriously, though, you should probably read the news today and make sure anything relevant is updated on your system.)
Don't worry, thanks to the higher security in modern Windows versions and the vulnerability being discussed elsewhere today, a hacker can now install a virus automatically without even troubling your users to open their mail!
Exactly. Email is a nice, stable standard that mostly Just Works. An email client can be the same, and that will do just fine if the alternative is everything breaking every five minutes because someone couldn't follow a standard or didn't want to bother maintaining the ancient code they wrote last month.
Once all the major browsers restrict what you can do with plugins, and require all plugins be signed so there is top-down control of the whole ecosystem, you may as well give up.
Or we could stop just accepting self-updating software that doesn't understand the difference between an important security patch, expendable changes in functionality or UI, and changes that actually reduce functionality and make things not work when they worked before. The latter is an area where browsers have been particularly awful for some time.
There are already projects attempting to move in this direction based on the Firefox code. Perhaps we should do more to support them and raise their profile.
I suspect that would be true if it were only Netflix, as there are other video streaming services. But if it started to be, say, Netflix and a couple of the other big streaming services and one or two social networks a household uses (which still work fine on everyone's phone) then it might be a different story.
The curious thing about this stand-off is that if the content providers are right, and net neutrality really is necessary to ensure a level playing field in markets where competition between ISPs is insufficient, then they might not have to spend their money lobbying to the same degree. If heavy traffic users like Facebook and Netflix call the ISPs' bluff and make a public statement that under the new arrangements they will no longer offer access to their services via US ISPs that require additional fees to transport the traffic, who is going to get the blame?
There are ways to set up DRM that don't rely on single providers like that. In fact, you pretty much have to if you're distributing content on the Web, because no single DRM scheme works reliably on all popular platforms.
Of course, if you actually are a large creator/distributor like Netflix, it certainly would be viable to implement your own scheme if you wanted to as well.