The average news reader couldn't tell an iPhone photo from one taken with a full frame camera and 1000$ glass.
Last I checked you couldn't get massive zoom lenses for an iphone, which rather precludes doing "journalism" with them. (Where "journalism" == "publish nudie photos of $celebrity").
If you make a big line in the airport for bomb-checks, you've moved the killing location from the plane to the airport waiting line.
Likewise, if you move the Friday surgeries to Thursday such that Thursday is the new day before the end of the work week, then Thursday might exhibit the same death rates.
Also, craming the same number of surgeries into fewer days is going to increase pressures, which tends to increase mistakes.
Possibly. But before we get into consumer demographics and user behavior statistics I think it's best we just agree that there's a huge variance in the types of users.
Which is why I was pointing out that your argument that "these aren't consumer grade ISPs because they don't allow blazingly fast connections that allow terabytes a month to be downloaded" is bogus - *most* people don't want to download that much, the people who do are in the minority.
Reliability is of course somewhat less tangible. I don't know of any statistics that would let me evaluate different companies performance. My experience has usually been between 0.5-1.5% downtime. I can see how this might be a dealbreaker for some businesses, but I can live with it. The link you posted certainly doesn't advertise guaranteed fault-correction, which is typically part of a business grade contract.
0.5-1.5% downtime sounds extremely bad to me - 1% downtime equates to 7 hours per month or 3.5 days a year. Conversely, I have had no downtime at all in the past 4 years (other than the occasional resync, which is on the order of a couple of seconds). If I was experiencing outages amounting to days per year then I would be changing ISP, whether or not that connection was used for business.
"Uncompetetive" is what I'd label it. Whether or not you want to call it outrageous I'll leave up to you.
And yet the ISPs in question seem to do pretty well, even from their home users, so it can't be that uncompetetive.
I guess you can compare it to the supermarkets - you can go to Tesco and pick up a packet of "Value" meat for under half the price of the equivalent "Finest" meat; does that mean that the "Finest" brand is uncompetetive? Seemingly not - lots of people buy that (I have no statistics, but anecdontally I would suggest that more people probably buy the "Finest" brand meat than the "Value" equivalent). Same goes for drinks - you can buy a supermarket own-brand cola, or you can pay several times that price for Coca Cola; is coke uncompetetive? it seems not - they're still in business and doing pretty well by all accounts.
Someone offering a product that is at a higher price than the competition doesn't make it "uncompetetive" unless the customers perceive the more expensive product to not be worth the extra money. A lot of people are willing to pay a small amount extra for what they perceive as a higher quality service (and from personal experience dealing with the useless customer support departments at Virgin, Talk Talk and a few of the other low-cost ISPs, I would say that the improved support you get by paying a bit extra is definitely worth the money).
So let's reiterate: For just short of £30/month (don't forget the "line rental") I get an outdated ADSL connection with a speed of anything southward of 24Mb/s
Errm, I'm not sure how having a speed of under 24Mbps makes an ADSL connection "outdated" since that's the top speed of *any* ADSL connection...
Meanwhile at Virgin Media, £30/month will get you 60Mb/s maximum speed (and because they don't use copper wires you do actually get this speed on pretty much every connection) and a traffic usage policy that's more like 2GB/hour.
Firstly, despite Virgin's insistance that they offer "fibre optic broadband", they use copper cables just like everyone else (only they use coax instead of twisted pair, but its still copper).
Secondly, some of us don't consider speed to be the primary factor when buying a service. I don't actually use my internet connection for anything that would benefit from 60Mb/s, whilst I do want decent support, reliability, etc.
But the take-home message here is that for average consumers who are interested in watching youtube, movies, video chat, and maybe a little bit of naughty file sharing, *these are not great deals*.
I think you overestimate what the "average consumer" does with their internet connection. For example, my fiancée does about 20GB per month - she doesn't have a TV (uses iplayer exclusively), a good chunk of that bandwidth is actually used by me working from her home when I visit and we use video chat quite a lot (often just leaving it going the whole day). My parents probably use way below 10GB/month. The vast vast majority of internet users don't do huge amounts of file sharing and buying a service that allows them to download terabytes a month (which they don't want to do) instead of buying a service thats reliable and has good customer support would be silly.
In fact, personally I actively avoid services that are promoting high usage (by offering cheap "unlimited" accounts) because I know the economics of running an ISP doesn't allow you to do "cheap", "unlimited" and "reliable" at the same time as paying for staff who have a clue. I *want* my ISP to limit the heavy users so that I'm not subsidising them.
Now, if every package they offer is much more expensive than otherwise comparable products from the big providers, the "free" service can't reasonably be considered free. Any reasonable person would conclude that you went for a specialist ISP and are paying a premium for a service you require (IP addresses).
The prices I see for the ISPs you are claiming are "specialist" just don't seem that high to me. Yes they aren't absolutely the rock-bottom prices, but they're not outrageous.
Oh so much pain and suffering! You didn't have a cheap mass produced century old replacement available over night; poor you. Get over yourself.
I'm not sure what your point is... you seem to be agreeing that the "green" replacements are frequently not as good at the job. Are you saying that I should "get over myself" and ignore these failings rather than taking them into account when deciding if the "green" replacement will do the job I need?
Don't get hung up on that 1 issue you happened to "suffer" so heavily over for some years...
I wasn't - I was using CFLs as an example of why "green" has often become synonimous with "not as good". You can take small engine "eco-cars" as another example if you like - they can similarly be considered "not as good" by virtue of being small (less carrying capacity/leg room/etc.) and having poor acceleration.
I'm also not saying there's necessarilly anything you can actually do about this - green technologies *are* often going to have limitations that weren't present in the non-green versions; all I'm saying is that people are wary of "green" stuff, and probably quite rightly so.
There is a school of thought where most significant progress is attributed to WAR and the conclusion is that we need wars in order to provide the motivation to innovate; otherwise, progress is SLOW, stagnant, or doesn't even happen. I'm not saying I'm one of those, but the grain of truth they build their hypothesis upon exists and shouldn't be ignored.
One thing that war helps with is directing vast sums of government money into research projects. This produces great strides in research at the expense of driving a country into debt (which isn't necessarilly a bad thing - the economic benefits from all that money moving around may well outweigh the detriment of the debt).
His statements seem to be very confused. He wants Google and others to do more to block material depicting child abuse. Well that's already blocked in the UK, and it's done at the ISP level
No its not. *Some* ISPs have signed up to censor content the IWF(*) tells them to (notably Virgin and BT). The majority of ISPs haven't.
(* The IWF mandates blocks on whatever content they feel like with no kind of oversight to prevent abuses - they have a history of being a tad overzealous, so sensible people don't use an ISP that signs up to the IWF. The IWF also requires a relatively large mandatory "voluntary donation" from the ISPs, which is another reason why the smaller ISPs aren't interested).
He wants Google to use 'safe search' as their default search setting. I thought they already did?
The default is moderate - it sounds like he wants it to default to strict, which would piss people off by being overzealous. I suppose its also one step away from automatically incriminating people who have opted to turn safesearch off...
He seems to think people will have to register to be able to search for porn. Register where?
Previous plans have suggested that ISPs should block all porn unless the customer has explicitly requested for it to be unblocked. This, of course, is a terrible idea for various reasons, not least the fact that this means the ISPs are going to have an (arguably) rather sensitive list of people who watch porn that would be embarrassing if it leaked.
This is where I'm utterly confused by what he's assuming, what he's proposing and how he thinks it will work.
As with most kneejerk reactions, its not been thought through at all and appears to just be a brain-dump from someone who wants to fix a perceived problem with the internet, but doesn't actually understand anything about the internet and hasn't even formalated his goals.
The good news is that in another ten years or so the politicians will start to be replaced by people that grew up with the Internet, that understand it better and that will at least have a grasp of the pragmatic realities involved.
Have you actually talked to someone nontechnical who grew up with the internet? Most seem to treat it like magic (and therefore anything is trivially possible) as much as the current politicians.
None of the companies you listed is a consumer-grade ISP. Your Entanet doesn't even advertise that they have residential services.
Entanet doesn't sell directly - you have to go through one of their many resellers, who *do* offer residential connections. For example, I use UKFSN - oh look, right there on their website it says "family broadband". If you buy a "family broadband" connection from UKFSN you'll get an Entanet account which can do IPv6, static IPv4 subnets, etc.
The front page of A&A advertises "Home::1 for the modern family" as a product. Again, you get IPv6, etc. with that.
I'll admit that Bogons is more aimed at business, but in what way are the two ISPs I just mentioned, which both advertise internet connections "for the family" not "consumer grade"?
The few "home" services I found from your recommended companies were also all much more expensive.
Much more expensive than what? Yes, you can certainly get much much cheaper services with quality to match, but the ISPs I mentioned aren't in any way outrageously priced - UKFSN, for example, starts at under £15/month.
Or are you saying that "consumer grade" equates to "cheap shit" and therefore if I point at any ISP that isn't right at the bottom of the barrel then that doesn't qualify?
And your static IP isn't really free if you're paying for a business-class service at a business-class price.
I had static IPs at no extra cost on the three non-business-class services I used before switching to a business account (when I started my business). The first one was Demon, second one was Plusnet, the third was a non-business UKFSN account. Of course, your tortalogical argument of "it isn't consumer grade if it isn't shit so there are no non-shit consumer grade accounts" argument somehow suggests that these consumer-grade accounts aren't actually consumer-grade.
FWIW I've never gotten the option of a "free" static IP with any of the ISPs I've been with.
Maybe you just didn't look?
But then again I don't shop for business-grade broadband for my home anyway.
How much do you pay for your service?
As mentioned above, whilst my current internet connection is a low cost business class ADSL (£22/month + VAT), my previous 3 internet connections were just plain home-user accounts. I can't recall the prices off the top of my head, but I think it was around £20 inc VAT for UKFSN around the 2007-2008 era, £20-or-so for Plusnet around the 2002-2007 era and probably a similar price for Demon in the pre-2002 era. None of these are outrageous prices.
There are studies that show "conservatives" here in the USA will buy CFL bulbs on their own (if they think) but as soon as you label them "green" or with other labels and slogans that have been associated as belonging to the enemy tribe, they will fuck themselves just to not have anything to do with the opposing tribe.
First, let me state that I'm certainly in favour of protecting the environment, although I frequently disagree with the environmental nuts about how to do it (the environmentalists often seem to reject anything they see as "not perfect" even if it is better than the status quo. For example, the greenie political campaigns usually seem to be anti-nuclear and claim we should be investing exclusively in "renewables" instead, whereas I believe that whilst nuclear isn't perfect, its better than what we have already and don't believe that the renewables can do the same job).
IMHO, "green" has frequently come to mean "not as good", probably because the green technologies are new and get pushed into the market place to compete with the established technologies before they are ready.
CFLs are a pretty good example of this: people are used to instant-on incandescents with an excellent CRI; then CFLs were introduced as the "green" equivalent and they were, quite frankly, crap in comparison - terrible CRI, took forever to warm up and due to frequent mis-labelling by the vendors often never got as bright as the incandescent bulbs they were claimed to be "equivalent" to. This has got better over recent years, but CFLs still don't come close to being as instant-on as incandescents.
The anti-green backlash is probably made worse when governments start outlawing the "non-green" alternatives - for example, here in Europe, a wide range of tungsten incandescent bulbs are no longer allowed to be sold. Whilst I do want to use energy efficient lighting in most places, there are some areas where a good old tungsten light would be more suitable and it pisses me off that I'm not allowed to use it. This is less of an issue now LED bulbs are easy and relatively cheap to get hold of, but at the time the legislation was passed that wasn't the case.
So whilst I *do* want to buy environmentally sound products, I also know that "green" has often become synonymous with "not as good" and "not quite ready for the real world yet", so I look at such products cautiously before buying them - nothing to do with tribalism, everything to do with having been burnt a few times by "green" products that just didn't do the job I bought them for.
Jesus. People keep talking about this, and reliability / maintenance cost (eg: "expensive battery replacements") as a huge unknown. The Prius has been available in the US since 2000. That's 13 years. If they were being junked sooner than non-hybrids, we'd know by now.
The Prius used NiMh batteries; Tesla is using Li-ion. Comparing them is an apples/oranges thing - in particular, Li-ion batteries *do* have a finite life that starts ticking down as soon as they roll off the production line, no matter how you treat them. Given the cost of batteries, I too would be somewhat dubious unless the car came with a warranty offering free (or extremely reduced cost) battery replacements for the first 10-15 years. My current (petrol) car is a bit over 10 years old at the moment, I have no plans to replace it any time soon, so I would want an electric car to have a similar life without majorly expensive maintenance.
Not really a banks fault though - why is the browser hanging on to post'd data after it's been post'd??
So that when you hit refresh on the page, the browser can pop up its usual "you'll need to repost to refresh this page, are you sure?" and do the repost if you tell it to.
So maybe you shouldn't try to control web access from your network if you allow it at all, but rather deal with people browsing Slashdot or porn sites all day long when and if it becomes a problem?
1. You're assuming this is just about stopping people wasting their time browsing porn all day, whereas filtering web traffic is very useful for security purposes (e.g. blocking phishing sites, malware, etc). 2. You're assuming I'm dealing with adults. I'm not.
A proxy never works for HTTPS, you're always going to be doing MITM.
You're wrong. Without MITMing the encrypted traffic you will, of course, never get access to much of the information being transferred. But you do get access to the host name being connected to, what useragent is being used, etc.
You're using the wrong tools, you should be using Kerberos for service authentication
Yes, HTTP proxies support Kerberos just fine.
802.1x for device authentication if that's what you really want
802.1x would certainly be ideal; but frequently not feasible due to the amount of integration required within a network - you're going to need to replace all your switches with 802.1x capable ones, integrate it with DHCP and firewalling, etc. and the overhead of managing the certificates on hundreds of devices is quite extreme.
to block access to certain sites you could be using transparent proxies
As mentioned elsewhere, transparent proxying isn't possible with HTTPS unless you're going to MITM the encrypted traffic.
or block the DNS queries.
That's a very blunt instrument - you're talking about blocking all access to a host for all client software, rather than blocking specific parts of a website.
Even so, proxies are dead with the amount of user-specific content the Internet generates these days.
I don't see how "user-specific content" is at all relevant to the conversation. Sure it reduces that amount of content that a proxy can cache, but not greatly so - there is a *lot* of static content around still (javascript, css, images, etc) and caching it certainly has a big impact on the performance of a network. Its generally only the HTML and JSON traffic that is user-specific and uncachable, which tends to be a small proportion.
Proxy servers are relic of a time before NAT. Please, please, please stop using this old hack to "share" your office Internet connection.
Thats not the purpose of a proxy server in a modern environment. A great many large organisations use web proxies to control web access; this involves stuff like anti-virus/anti-phishing (by examining the http traffic); accellerating a busy internet connection using a cache is also a big performance boost, especially in certain environmnet where you can expect a large number of people to simultaneously access some specific resources. You may consider them a relic, many organisations don't and have actual legitimate use for them beyond sharing a connection (just a look at the traffic on the Squid mailing list will show you that it is still extremely popular).
If you want to prevent SMTP/FTP/IRC/etc traffic on your network, set up a proper firewall that blocks those port ranges.
What on earth have SMTP/FTP/IRC got to do with a conversation about http proxy servers?
As you pointed out, using a proxy server in 2013 is going to give grief to anybody that has to touch it.
Its funny, Windows and OS-X, and the applications that run on them largely handle proxy servers without any problems. Its basically Android and iOS (mostly iOS) that causes problems - Apple's implementation is so utterly half-arsed and bugridden I'm often left wondering why they bothered implementing it at all.
Both devices have plenty of support for HTTP proxies.
Android Gingerbread lets you set a single HTTP proxy which applies to all networks. That means device owners have to manually enter and clear the proxy settings as they move between the office network and their home network. Not that it matters - almost all apps ignore the proxy settings anyway.
Android ICS and Jellybean let you set an HTTP proxy per wifi network, which at least means the user isn't expected to reconfigure the phone all the time. Most apps still ignore the proxy settings. Most of the apps that do pay attention to the proxy settings don't support authenticated proxy servers.
All recent versions of iOS allow the proxy and authentication credentials to be set on a per wifi network basis. That's excellent. Except that most apps (including a good chunk of the stock iOS apps that Apple ship with the phone) either ignore the proxy settings entirely or fail to support authenticated proxy servers. (Yes, Apple is aware of these problems - there are bug reports in their bug tracking system that have been open for several years, they aren't interested in fixing them).
Even then, Squid has a transparent proxy option.
Transparent proxying only works for HTTP, not HTTPS unless you are going to MITM all the sessions (which involves installing certificates on all the clients). And even then, you can't authenticate the users if you're proxying transparently.
You shouldn't trust your own network to begin with. How do you make sure no-one plugs in whatever they want?
Yep, I've had customers insist they don't need to worry about antivirus, etc. on their workstations because they have a company policy that no one plugs unauthorised kit into the network. A few weeks later they invariably get an infection because one of the directors ignored policy and plugged his personal laptop in - afterall, who's going to tell the director off?
BYOD is not just about someone saving money. It's about people expecting to have their devices work and IT in organizations being too slow or not having enough funding to give everybody their device of choice.
I've found BYOD is actually a big PITA for large organisations because the devices people are bringing are almost universally Android or iOS, and in both cases the OS and apps have terrible support for HTTP proxies; and many large organisations use proxies to control web access from within their networks.
I have to disagree with you here. I think Apple created their own timing. The only thing they waited on was broadband internet access. When it was introduced, the iPod blew people away;
There were plenty of MP3 players around before the ipod (and good ones too!). What Apple seems to have is the ability to create a religious following for their products. I don't think this is necessarilly down to the product itself - the original ipod really was nothing special compared to the competition - the user interface wasn't significantly better, the size wasn't significantly different, the battery life wasn't signficantly different, yet somehow they managed to become the "defacto" MP3 player.
The iPhone is another good example - when people go on about how the iphone took the market because it was so much better than everything else, I think they forget what the _original_ iphone was like - it was marketted as a smartphone, but it lacked most of the features that other smartphones had - there was no tethering, no 3G(*), no apps. About the only thing that beat the competition was the web browser they integrated into it.
(* I know the US-centric crowd say that no 3G was never a big deal because there were no 3G networks anywhere in the world at that time, but this is patently untrue - across europe 3G was commonplace and the introduction of a fantastic new smartphone that didn't have 3G struck a lot of us as a complete WTF).
the iPad did something that Microsoft failed to do for over 10 years: get a tablet to be accepted and used by the general public.
There are 3 factors here I think, in addition to the aforementioned way that Apple seems to be able to sell anything, irrespective of how it compares to the competition: 1. Technology. 10 years before the ipad, it wasn't possible to make a device that small and light. It simply wasn't. So the MS tablets were always big and heavy - that wasn't MS's fault, the technology simply wasn't there and Apple would've had the same problems at that time. 2. User interface. MS has always tried to shoe-horn their existing software into new markets instead of developing whole new systems where necessarilly. MS's tablets always had their desktop UI shoehorned onto them, whereas Apple realised that wasn't going to work and built a new UI. We can see the continuation of MS's problems with the way they've now produced a UI for tablets and tried to shoehorn it inappropriately onto the desktop. If MS had started off on the tablet market and moved into the desktop market they would've probably never managed to make the desktop take off because they seem to be unable to see that their existing software isn't always suitable for all situations. This has nothing to to with timing and everything to do with MS just being generally terrible at this stuff. 3. Smartphones. At the point that Apple launched their tablet, people had had several years to get used to browsing the web on their phones with decent browsers using mobile apps, etc., so a move towards a tablet for content consumption was reasonable. Without that experience I think people would be a lot less inclined to ditch the keyboard and compatibility with their normal desktop software.
That all being said, whilst Apple was about the first to market with something the size and performance of an ipad, the other manufacturers certainly weren't far behind - if apple hadn't launched the ipad I'm pretty sure someone else would've done the equivalent at around that time.
Google wouldn't have even considered creating Android. MS and Google were happy with the status quo in terms of computing devices. Apple really introduced the computing appliance.
Android was in development for a long time before Apple released the iPhone, as were various other similar projects (for example, OpenMoko; which was never taken seriously by the industry, but basically got quite a long way towards producing somethi
It can be unethical when the manufacturer or a group of manufacturer makes sure that the products you can buy are only available with certain limitations and at a fixed price.
Conversely, even though the hardware may be capable of doing many things with the right software, those software features cost money to create. So the vendor has a choice: 1. Give everyone those software features, raise the price for everyone to cover the cost of creating them. 2. Give those software features only to the people willing to pay for them, therefore keeping the price down for the people who aren't.
(2) seems like a better option for everyone - the consumers who aren't interested in paying for a feature get to keep the cheap price they desire; the consumers who are interested in paying for a feature gets that feature; the vendor recoups the cost of (and profits from) development of that feature.
The slashdot crowd seem to think that just because software distribution is essentially free, software creation is too.
I don't call RHEL an experimental workstation distro. Unless they go with Gnome 3 and/or Poettering's reinventions in RHEL 7, in which case I may have to revise my view.
Actually, more to the point: why would you be running an X desktop environment on a server at all? Running X applications remotely (tunnelled over SSH) is occasionally useful, but why on earth would you want a full desktop environment?
Gnome3's interface... let's not speak of it, I prefer to not use words it deserves among civilised people.
I migrated to Gnome 3 from E17 because it's actually about the best DE I've found. Yes, it has problems; there are parts of it that make me wonder WTF the developers were thinking (mostly the bits they've ripped off from Apple, which frequently seem ill thought out even on OS X), but it generally works better for me than any other DE I've tried. If you don't like Gnome 3 then that's fine - there's plenty of choice, but don't shoot down the whole distro because it happens to default to a DE that you personally don't like, but which many other people find to be excellent.
As for Network Manager, try running it with any USB networking (direct connect, like with a phone, rather than an USB-connected ethernet card): it will kill the interface every roughly 30 seconds. Its upstream refused to fix that saying they don't aim to support every possible device. Or, bridged setups. Or, basically anything more complex than a plain ethernet or wifi interface.
The USB thing is a shame - I can't really comment on that as I've not tried using USB NICs with it.
As for "you can't do anything complex with it", IMHO it isn't intended for that use - network manager is intended as a "plug and play networking for dummies" system; if you want something complex then set NM_CONTROLLED=no in the network config and configure it yourself. Adding lots of support for very complex setups to NetworkManager itself, when that's already supported via other mechanisms, would seem to defeat its purpose of offering a *simple* network configurator.
But not, it insists it has the complete view of the system's network, everything else is wrong, and even if you blacklist a device it knows (not possible for ones it doesn't), it still says you're in "offline mode" when you use programs that made the mistake of querying NM.
That certainly doesn't seem to agree with my experiences. I frequently set systems up with NM_CONTROLLED=no in the NIC configuration and NetworkManager handles that just fine (in fact, on servers I make a point of doing this; which is fine - IMHO NetworkManager is neither intended nor suited to server environments so turning it off and using more traditional configurations (which are still supported) is a good idea).
If a single line, "apt-get purge network-manager", instantly fixes all problems of this kind, I'm kind of disinclined to believe that "it works pretty well".
If you're using apt-get then you're not using Fedora, so your comments seem a bit irrelevant to a discussion about the latest Fedora release. I can't comment on how well NetworkManager works in other distros, but under both Fedora and Scientific Linux it seems to work well and is trivial to bypass if you need lots of complexities in your network configuration.
Gnome 3 works well in a VM or with remote X on headless servers? That's news to me.
Why on earth would you be running Fedora of all things on a server? Yes, things don't work so well when you use software that's inappropriate for the job - why not get a LTS server distro instead of a experimental workstation distro if you want to run servers?
1,2 and3) fat chance of getting that with a consumer grade ISP. Which national ISPs do IPv6 anyway?
I've been using EntaNet for years - I get a free static IPv4/29 subnet and a free native IPv6/56 subnet. The monthly cost is quite reasonable.
You'll get similar from A&A and Bogons, albeit at a fairly high price.. Static IPv4 IP addresses are pretty common - most of the not-dirt-cheap ISPs are happy to offer them, usually at no extra cost. Even Plusnet used to do free static IPv4 subnets (not sure if they still do).
I think Claranet and a couple of others also do native IPv6 as standard.
I was paying 25 GBP a month for a 30 Mb connection, which seemed quite reasonable compared to the DSL providers who had low usage caps, no speed guarantees, and wanted an extra 15 GBP on top of their advertised price.
No static IP, no chance of an IPv4 subnet, no IPv6 at all, terrible customer service.
Which is a reason not to go for the DSL. You have to compare the full price for the whole package. The cable providers always offer double or triple play and give you a telephone adapter to use you POTS phone.
There's only one cable provider in the UK and the are extremely crap. There's a great many very good DSL providers to choose between.
Slashdot Mirror
The average news reader couldn't tell an iPhone photo from one taken with a full frame camera and 1000$ glass.
Last I checked you couldn't get massive zoom lenses for an iphone, which rather precludes doing "journalism" with them. (Where "journalism" == "publish nudie photos of $celebrity").
If you make a big line in the airport for bomb-checks, you've moved the killing location from the plane to the airport waiting line.
Likewise, if you move the Friday surgeries to Thursday such that Thursday is the new day before the end of the work week, then Thursday might exhibit the same death rates.
Also, craming the same number of surgeries into fewer days is going to increase pressures, which tends to increase mistakes.
Possibly. But before we get into consumer demographics and user behavior statistics I think it's best we just agree that there's a huge variance in the types of users.
Which is why I was pointing out that your argument that "these aren't consumer grade ISPs because they don't allow blazingly fast connections that allow terabytes a month to be downloaded" is bogus - *most* people don't want to download that much, the people who do are in the minority.
Reliability is of course somewhat less tangible. I don't know of any statistics that would let me evaluate different companies performance. My experience has usually been between 0.5-1.5% downtime. I can see how this might be a dealbreaker for some businesses, but I can live with it. The link you posted certainly doesn't advertise guaranteed fault-correction, which is typically part of a business grade contract.
0.5-1.5% downtime sounds extremely bad to me - 1% downtime equates to 7 hours per month or 3.5 days a year. Conversely, I have had no downtime at all in the past 4 years (other than the occasional resync, which is on the order of a couple of seconds). If I was experiencing outages amounting to days per year then I would be changing ISP, whether or not that connection was used for business.
"Uncompetetive" is what I'd label it. Whether or not you want to call it outrageous I'll leave up to you.
And yet the ISPs in question seem to do pretty well, even from their home users, so it can't be that uncompetetive.
I guess you can compare it to the supermarkets - you can go to Tesco and pick up a packet of "Value" meat for under half the price of the equivalent "Finest" meat; does that mean that the "Finest" brand is uncompetetive? Seemingly not - lots of people buy that (I have no statistics, but anecdontally I would suggest that more people probably buy the "Finest" brand meat than the "Value" equivalent). Same goes for drinks - you can buy a supermarket own-brand cola, or you can pay several times that price for Coca Cola; is coke uncompetetive? it seems not - they're still in business and doing pretty well by all accounts.
Someone offering a product that is at a higher price than the competition doesn't make it "uncompetetive" unless the customers perceive the more expensive product to not be worth the extra money. A lot of people are willing to pay a small amount extra for what they perceive as a higher quality service (and from personal experience dealing with the useless customer support departments at Virgin, Talk Talk and a few of the other low-cost ISPs, I would say that the improved support you get by paying a bit extra is definitely worth the money).
So let's reiterate: For just short of £30/month (don't forget the "line rental") I get an outdated ADSL connection with a speed of anything southward of 24Mb/s
Errm, I'm not sure how having a speed of under 24Mbps makes an ADSL connection "outdated" since that's the top speed of *any* ADSL connection...
Meanwhile at Virgin Media, £30/month will get you 60Mb/s maximum speed (and because they don't use copper wires you do actually get this speed on pretty much every connection) and a traffic usage policy that's more like 2GB/hour.
Firstly, despite Virgin's insistance that they offer "fibre optic broadband", they use copper cables just like everyone else (only they use coax instead of twisted pair, but its still copper).
Secondly, some of us don't consider speed to be the primary factor when buying a service. I don't actually use my internet connection for anything that would benefit from 60Mb/s, whilst I do want decent support, reliability, etc.
But the take-home message here is that for average consumers who are interested in watching youtube, movies, video chat, and maybe a little bit of naughty file sharing, *these are not great deals*.
I think you overestimate what the "average consumer" does with their internet connection. For example, my fiancée does about 20GB per month - she doesn't have a TV (uses iplayer exclusively), a good chunk of that bandwidth is actually used by me working from her home when I visit and we use video chat quite a lot (often just leaving it going the whole day). My parents probably use way below 10GB/month. The vast vast majority of internet users don't do huge amounts of file sharing and buying a service that allows them to download terabytes a month (which they don't want to do) instead of buying a service thats reliable and has good customer support would be silly.
In fact, personally I actively avoid services that are promoting high usage (by offering cheap "unlimited" accounts) because I know the economics of running an ISP doesn't allow you to do "cheap", "unlimited" and "reliable" at the same time as paying for staff who have a clue. I *want* my ISP to limit the heavy users so that I'm not subsidising them.
Now, if every package they offer is much more expensive than otherwise comparable products from the big providers, the "free" service can't reasonably be considered free.
Any reasonable person would conclude that you went for a specialist ISP and are paying a premium for a service you require (IP addresses).
The prices I see for the ISPs you are claiming are "specialist" just don't seem that high to me. Yes they aren't absolutely the rock-bottom prices, but they're not outrageous.
Oh so much pain and suffering! You didn't have a cheap mass produced century old replacement available over night; poor you. Get over yourself.
I'm not sure what your point is... you seem to be agreeing that the "green" replacements are frequently not as good at the job. Are you saying that I should "get over myself" and ignore these failings rather than taking them into account when deciding if the "green" replacement will do the job I need?
Don't get hung up on that 1 issue you happened to "suffer" so heavily over for some years...
I wasn't - I was using CFLs as an example of why "green" has often become synonimous with "not as good". You can take small engine "eco-cars" as another example if you like - they can similarly be considered "not as good" by virtue of being small (less carrying capacity/leg room/etc.) and having poor acceleration.
I'm also not saying there's necessarilly anything you can actually do about this - green technologies *are* often going to have limitations that weren't present in the non-green versions; all I'm saying is that people are wary of "green" stuff, and probably quite rightly so.
There is a school of thought where most significant progress is attributed to WAR and the conclusion is that we need wars in order to provide the motivation to innovate; otherwise, progress is SLOW, stagnant, or doesn't even happen. I'm not saying I'm one of those, but the grain of truth they build their hypothesis upon exists and shouldn't be ignored.
One thing that war helps with is directing vast sums of government money into research projects. This produces great strides in research at the expense of driving a country into debt (which isn't necessarilly a bad thing - the economic benefits from all that money moving around may well outweigh the detriment of the debt).
His statements seem to be very confused. He wants Google and others to do more to block material depicting child abuse. Well that's already blocked in the UK, and it's done at the ISP level
No its not. *Some* ISPs have signed up to censor content the IWF(*) tells them to (notably Virgin and BT). The majority of ISPs haven't.
(* The IWF mandates blocks on whatever content they feel like with no kind of oversight to prevent abuses - they have a history of being a tad overzealous, so sensible people don't use an ISP that signs up to the IWF. The IWF also requires a relatively large mandatory "voluntary donation" from the ISPs, which is another reason why the smaller ISPs aren't interested).
He wants Google to use 'safe search' as their default search setting. I thought they already did?
The default is moderate - it sounds like he wants it to default to strict, which would piss people off by being overzealous. I suppose its also one step away from automatically incriminating people who have opted to turn safesearch off...
He seems to think people will have to register to be able to search for porn. Register where?
Previous plans have suggested that ISPs should block all porn unless the customer has explicitly requested for it to be unblocked. This, of course, is a terrible idea for various reasons, not least the fact that this means the ISPs are going to have an (arguably) rather sensitive list of people who watch porn that would be embarrassing if it leaked.
This is where I'm utterly confused by what he's assuming, what he's proposing and how he thinks it will work.
As with most kneejerk reactions, its not been thought through at all and appears to just be a brain-dump from someone who wants to fix a perceived problem with the internet, but doesn't actually understand anything about the internet and hasn't even formalated his goals.
The good news is that in another ten years or so the politicians will start to be replaced by people that grew up with the Internet, that understand it better and that will at least have a grasp of the pragmatic realities involved.
Have you actually talked to someone nontechnical who grew up with the internet? Most seem to treat it like magic (and therefore anything is trivially possible) as much as the current politicians.
None of the companies you listed is a consumer-grade ISP. Your Entanet doesn't even advertise that they have residential services.
Entanet doesn't sell directly - you have to go through one of their many resellers, who *do* offer residential connections. For example, I use UKFSN - oh look, right there on their website it says "family broadband". If you buy a "family broadband" connection from UKFSN you'll get an Entanet account which can do IPv6, static IPv4 subnets, etc.
The front page of A&A advertises "Home::1 for the modern family" as a product. Again, you get IPv6, etc. with that.
I'll admit that Bogons is more aimed at business, but in what way are the two ISPs I just mentioned, which both advertise internet connections "for the family" not "consumer grade"?
The few "home" services I found from your recommended companies were also all much more expensive.
Much more expensive than what? Yes, you can certainly get much much cheaper services with quality to match, but the ISPs I mentioned aren't in any way outrageously priced - UKFSN, for example, starts at under £15/month.
Or are you saying that "consumer grade" equates to "cheap shit" and therefore if I point at any ISP that isn't right at the bottom of the barrel then that doesn't qualify?
And your static IP isn't really free if you're paying for a business-class service at a business-class price.
I had static IPs at no extra cost on the three non-business-class services I used before switching to a business account (when I started my business). The first one was Demon, second one was Plusnet, the third was a non-business UKFSN account. Of course, your tortalogical argument of "it isn't consumer grade if it isn't shit so there are no non-shit consumer grade accounts" argument somehow suggests that these consumer-grade accounts aren't actually consumer-grade.
FWIW I've never gotten the option of a "free" static IP with any of the ISPs I've been with.
Maybe you just didn't look?
But then again I don't shop for business-grade broadband for my home anyway.
How much do you pay for your service?
As mentioned above, whilst my current internet connection is a low cost business class ADSL (£22/month + VAT), my previous 3 internet connections were just plain home-user accounts. I can't recall the prices off the top of my head, but I think it was around £20 inc VAT for UKFSN around the 2007-2008 era, £20-or-so for Plusnet around the 2002-2007 era and probably a similar price for Demon in the pre-2002 era. None of these are outrageous prices.
There are studies that show "conservatives" here in the USA will buy CFL bulbs on their own (if they think) but as soon as you label them "green" or with other labels and slogans that have been associated as belonging to the enemy tribe, they will fuck themselves just to not have anything to do with the opposing tribe.
First, let me state that I'm certainly in favour of protecting the environment, although I frequently disagree with the environmental nuts about how to do it (the environmentalists often seem to reject anything they see as "not perfect" even if it is better than the status quo. For example, the greenie political campaigns usually seem to be anti-nuclear and claim we should be investing exclusively in "renewables" instead, whereas I believe that whilst nuclear isn't perfect, its better than what we have already and don't believe that the renewables can do the same job).
IMHO, "green" has frequently come to mean "not as good", probably because the green technologies are new and get pushed into the market place to compete with the established technologies before they are ready.
CFLs are a pretty good example of this: people are used to instant-on incandescents with an excellent CRI; then CFLs were introduced as the "green" equivalent and they were, quite frankly, crap in comparison - terrible CRI, took forever to warm up and due to frequent mis-labelling by the vendors often never got as bright as the incandescent bulbs they were claimed to be "equivalent" to. This has got better over recent years, but CFLs still don't come close to being as instant-on as incandescents.
The anti-green backlash is probably made worse when governments start outlawing the "non-green" alternatives - for example, here in Europe, a wide range of tungsten incandescent bulbs are no longer allowed to be sold. Whilst I do want to use energy efficient lighting in most places, there are some areas where a good old tungsten light would be more suitable and it pisses me off that I'm not allowed to use it. This is less of an issue now LED bulbs are easy and relatively cheap to get hold of, but at the time the legislation was passed that wasn't the case.
So whilst I *do* want to buy environmentally sound products, I also know that "green" has often become synonymous with "not as good" and "not quite ready for the real world yet", so I look at such products cautiously before buying them - nothing to do with tribalism, everything to do with having been burnt a few times by "green" products that just didn't do the job I bought them for.
Jesus. People keep talking about this, and reliability / maintenance cost (eg: "expensive battery replacements") as a huge unknown. The Prius has been available in the US since 2000. That's 13 years. If they were being junked sooner than non-hybrids, we'd know by now.
The Prius used NiMh batteries; Tesla is using Li-ion. Comparing them is an apples/oranges thing - in particular, Li-ion batteries *do* have a finite life that starts ticking down as soon as they roll off the production line, no matter how you treat them. Given the cost of batteries, I too would be somewhat dubious unless the car came with a warranty offering free (or extremely reduced cost) battery replacements for the first 10-15 years. My current (petrol) car is a bit over 10 years old at the moment, I have no plans to replace it any time soon, so I would want an electric car to have a similar life without majorly expensive maintenance.
Not really a banks fault though - why is the browser hanging on to post'd data after it's been post'd??
So that when you hit refresh on the page, the browser can pop up its usual "you'll need to repost to refresh this page, are you sure?" and do the repost if you tell it to.
So maybe you shouldn't try to control web access from your network if you allow it at all, but rather deal with people browsing Slashdot or porn sites all day long when and if it becomes a problem?
1. You're assuming this is just about stopping people wasting their time browsing porn all day, whereas filtering web traffic is very useful for security purposes (e.g. blocking phishing sites, malware, etc).
2. You're assuming I'm dealing with adults. I'm not.
A proxy never works for HTTPS, you're always going to be doing MITM.
You're wrong. Without MITMing the encrypted traffic you will, of course, never get access to much of the information being transferred. But you do get access to the host name being connected to, what useragent is being used, etc.
You're using the wrong tools, you should be using Kerberos for service authentication
Yes, HTTP proxies support Kerberos just fine.
802.1x for device authentication if that's what you really want
802.1x would certainly be ideal; but frequently not feasible due to the amount of integration required within a network - you're going to need to replace all your switches with 802.1x capable ones, integrate it with DHCP and firewalling, etc. and the overhead of managing the certificates on hundreds of devices is quite extreme.
to block access to certain sites you could be using transparent proxies
As mentioned elsewhere, transparent proxying isn't possible with HTTPS unless you're going to MITM the encrypted traffic.
or block the DNS queries.
That's a very blunt instrument - you're talking about blocking all access to a host for all client software, rather than blocking specific parts of a website.
Even so, proxies are dead with the amount of user-specific content the Internet generates these days.
I don't see how "user-specific content" is at all relevant to the conversation. Sure it reduces that amount of content that a proxy can cache, but not greatly so - there is a *lot* of static content around still (javascript, css, images, etc) and caching it certainly has a big impact on the performance of a network. Its generally only the HTML and JSON traffic that is user-specific and uncachable, which tends to be a small proportion.
Proxy servers are relic of a time before NAT. Please, please, please stop using this old hack to "share" your office Internet connection.
Thats not the purpose of a proxy server in a modern environment. A great many large organisations use web proxies to control web access; this involves stuff like anti-virus/anti-phishing (by examining the http traffic); accellerating a busy internet connection using a cache is also a big performance boost, especially in certain environmnet where you can expect a large number of people to simultaneously access some specific resources. You may consider them a relic, many organisations don't and have actual legitimate use for them beyond sharing a connection (just a look at the traffic on the Squid mailing list will show you that it is still extremely popular).
If you want to prevent SMTP/FTP/IRC/etc traffic on your network, set up a proper firewall that blocks those port ranges.
What on earth have SMTP/FTP/IRC got to do with a conversation about http proxy servers?
As you pointed out, using a proxy server in 2013 is going to give grief to anybody that has to touch it.
Its funny, Windows and OS-X, and the applications that run on them largely handle proxy servers without any problems. Its basically Android and iOS (mostly iOS) that causes problems - Apple's implementation is so utterly half-arsed and bugridden I'm often left wondering why they bothered implementing it at all.
Both devices have plenty of support for HTTP proxies.
Android Gingerbread lets you set a single HTTP proxy which applies to all networks. That means device owners have to manually enter and clear the proxy settings as they move between the office network and their home network. Not that it matters - almost all apps ignore the proxy settings anyway.
Android ICS and Jellybean let you set an HTTP proxy per wifi network, which at least means the user isn't expected to reconfigure the phone all the time. Most apps still ignore the proxy settings. Most of the apps that do pay attention to the proxy settings don't support authenticated proxy servers.
All recent versions of iOS allow the proxy and authentication credentials to be set on a per wifi network basis. That's excellent. Except that most apps (including a good chunk of the stock iOS apps that Apple ship with the phone) either ignore the proxy settings entirely or fail to support authenticated proxy servers. (Yes, Apple is aware of these problems - there are bug reports in their bug tracking system that have been open for several years, they aren't interested in fixing them).
Even then, Squid has a transparent proxy option.
Transparent proxying only works for HTTP, not HTTPS unless you are going to MITM all the sessions (which involves installing certificates on all the clients). And even then, you can't authenticate the users if you're proxying transparently.
You shouldn't trust your own network to begin with. How do you make sure no-one plugs in whatever they want?
Yep, I've had customers insist they don't need to worry about antivirus, etc. on their workstations because they have a company policy that no one plugs unauthorised kit into the network. A few weeks later they invariably get an infection because one of the directors ignored policy and plugged his personal laptop in - afterall, who's going to tell the director off?
BYOD is not just about someone saving money. It's about people expecting to have their devices work and IT in organizations being too slow or not having enough funding to give everybody their device of choice.
I've found BYOD is actually a big PITA for large organisations because the devices people are bringing are almost universally Android or iOS, and in both cases the OS and apps have terrible support for HTTP proxies; and many large organisations use proxies to control web access from within their networks.
I have to disagree with you here. I think Apple created their own timing. The only thing they waited on was broadband internet access. When it was introduced, the iPod blew people away;
There were plenty of MP3 players around before the ipod (and good ones too!). What Apple seems to have is the ability to create a religious following for their products. I don't think this is necessarilly down to the product itself - the original ipod really was nothing special compared to the competition - the user interface wasn't significantly better, the size wasn't significantly different, the battery life wasn't signficantly different, yet somehow they managed to become the "defacto" MP3 player.
The iPhone is another good example - when people go on about how the iphone took the market because it was so much better than everything else, I think they forget what the _original_ iphone was like - it was marketted as a smartphone, but it lacked most of the features that other smartphones had - there was no tethering, no 3G(*), no apps. About the only thing that beat the competition was the web browser they integrated into it.
(* I know the US-centric crowd say that no 3G was never a big deal because there were no 3G networks anywhere in the world at that time, but this is patently untrue - across europe 3G was commonplace and the introduction of a fantastic new smartphone that didn't have 3G struck a lot of us as a complete WTF).
the iPad did something that Microsoft failed to do for over 10 years: get a tablet to be accepted and used by the general public.
There are 3 factors here I think, in addition to the aforementioned way that Apple seems to be able to sell anything, irrespective of how it compares to the competition:
1. Technology. 10 years before the ipad, it wasn't possible to make a device that small and light. It simply wasn't. So the MS tablets were always big and heavy - that wasn't MS's fault, the technology simply wasn't there and Apple would've had the same problems at that time.
2. User interface. MS has always tried to shoe-horn their existing software into new markets instead of developing whole new systems where necessarilly. MS's tablets always had their desktop UI shoehorned onto them, whereas Apple realised that wasn't going to work and built a new UI. We can see the continuation of MS's problems with the way they've now produced a UI for tablets and tried to shoehorn it inappropriately onto the desktop. If MS had started off on the tablet market and moved into the desktop market they would've probably never managed to make the desktop take off because they seem to be unable to see that their existing software isn't always suitable for all situations. This has nothing to to with timing and everything to do with MS just being generally terrible at this stuff.
3. Smartphones. At the point that Apple launched their tablet, people had had several years to get used to browsing the web on their phones with decent browsers using mobile apps, etc., so a move towards a tablet for content consumption was reasonable. Without that experience I think people would be a lot less inclined to ditch the keyboard and compatibility with their normal desktop software.
That all being said, whilst Apple was about the first to market with something the size and performance of an ipad, the other manufacturers certainly weren't far behind - if apple hadn't launched the ipad I'm pretty sure someone else would've done the equivalent at around that time.
Google wouldn't have even considered creating Android. MS and Google were happy with the status quo in terms of computing devices. Apple really introduced the computing appliance.
Android was in development for a long time before Apple released the iPhone, as were various other similar projects (for example, OpenMoko; which was never taken seriously by the industry, but basically got quite a long way towards producing somethi
It can be unethical when the manufacturer or a group of manufacturer makes sure that the products you can buy are only available with certain limitations and at a fixed price.
Conversely, even though the hardware may be capable of doing many things with the right software, those software features cost money to create. So the vendor has a choice:
1. Give everyone those software features, raise the price for everyone to cover the cost of creating them.
2. Give those software features only to the people willing to pay for them, therefore keeping the price down for the people who aren't.
(2) seems like a better option for everyone - the consumers who aren't interested in paying for a feature get to keep the cheap price they desire; the consumers who are interested in paying for a feature gets that feature; the vendor recoups the cost of (and profits from) development of that feature.
The slashdot crowd seem to think that just because software distribution is essentially free, software creation is too.
Why on earth not?
Because most servers don't have a monitor connected to them?
I don't call RHEL an experimental workstation distro. Unless they go with Gnome 3 and/or Poettering's reinventions in RHEL 7, in which case I may have to revise my view.
Actually, more to the point: why would you be running an X desktop environment on a server at all? Running X applications remotely (tunnelled over SSH) is occasionally useful, but why on earth would you want a full desktop environment?
Gnome3's interface... let's not speak of it, I prefer to not use words it deserves among civilised people.
I migrated to Gnome 3 from E17 because it's actually about the best DE I've found. Yes, it has problems; there are parts of it that make me wonder WTF the developers were thinking (mostly the bits they've ripped off from Apple, which frequently seem ill thought out even on OS X), but it generally works better for me than any other DE I've tried. If you don't like Gnome 3 then that's fine - there's plenty of choice, but don't shoot down the whole distro because it happens to default to a DE that you personally don't like, but which many other people find to be excellent.
As for Network Manager, try running it with any USB networking (direct connect, like with a phone, rather than an USB-connected ethernet card): it will kill the interface every roughly 30 seconds. Its upstream refused to fix that saying they don't aim to support every possible device.
Or, bridged setups. Or, basically anything more complex than a plain ethernet or wifi interface.
The USB thing is a shame - I can't really comment on that as I've not tried using USB NICs with it.
As for "you can't do anything complex with it", IMHO it isn't intended for that use - network manager is intended as a "plug and play networking for dummies" system; if you want something complex then set NM_CONTROLLED=no in the network config and configure it yourself. Adding lots of support for very complex setups to NetworkManager itself, when that's already supported via other mechanisms, would seem to defeat its purpose of offering a *simple* network configurator.
But not, it insists it has the complete view of the system's network, everything else is wrong, and even if you blacklist a device it knows (not possible for ones it doesn't), it still says you're in "offline mode" when you use programs that made the mistake of querying NM.
That certainly doesn't seem to agree with my experiences. I frequently set systems up with NM_CONTROLLED=no in the NIC configuration and NetworkManager handles that just fine (in fact, on servers I make a point of doing this; which is fine - IMHO NetworkManager is neither intended nor suited to server environments so turning it off and using more traditional configurations (which are still supported) is a good idea).
If a single line, "apt-get purge network-manager", instantly fixes all problems of this kind, I'm kind of disinclined to believe that "it works pretty well".
If you're using apt-get then you're not using Fedora, so your comments seem a bit irrelevant to a discussion about the latest Fedora release. I can't comment on how well NetworkManager works in other distros, but under both Fedora and Scientific Linux it seems to work well and is trivial to bypass if you need lots of complexities in your network configuration.
Gnome 3 works well in a VM or with remote X on headless servers?
That's news to me.
Why on earth would you be running Fedora of all things on a server? Yes, things don't work so well when you use software that's inappropriate for the job - why not get a LTS server distro instead of a experimental workstation distro if you want to run servers?
Did they upgrade away from Gnome3, network-manager and systemd? If not, why should we even look at it?
Because Gnome 3 and Network Manager happen to work pretty well these days?
1,2 and3) fat chance of getting that with a consumer grade ISP. Which national ISPs do IPv6 anyway?
I've been using EntaNet for years - I get a free static IPv4 /29 subnet and a free native IPv6 /56 subnet. The monthly cost is quite reasonable.
You'll get similar from A&A and Bogons, albeit at a fairly high price.. Static IPv4 IP addresses are pretty common - most of the not-dirt-cheap ISPs are happy to offer them, usually at no extra cost. Even Plusnet used to do free static IPv4 subnets (not sure if they still do).
I think Claranet and a couple of others also do native IPv6 as standard.
I was paying 25 GBP a month for a 30 Mb connection, which seemed quite reasonable compared to the DSL providers who had low usage caps, no speed guarantees, and wanted an extra 15 GBP on top of their advertised price.
No static IP, no chance of an IPv4 subnet, no IPv6 at all, terrible customer service.
Which is a reason not to go for the DSL. You have to compare the full price for the whole package. The cable providers always offer double or triple play and give you a telephone adapter to use you POTS phone.
There's only one cable provider in the UK and the are extremely crap. There's a great many very good DSL providers to choose between.