The only one I can remember was when the server that responds to WISPr probes went down, rendering everyone's ipad unable to connect to a network...
Background: When an iOS device associates with a wifi network, it makes a web request to apple's server to see if its behind a captive portal. It expects to get back "SUCCESS" (returned by Apple's server) or a captive portal login page (returned by the wifi hotspot). If it doesn't get "SUCCESS" it displays the captive portal page so the user can log in. Unfortunately, Apple's software is unable to cope with the web request not being answered at all, and you end up with a blank "login" page and a non-funcational network connection. Yes, this is the usual quality I've come to expect from Apple, given the numerous problems I've had to deal with since iOS devices started to get popular in the workplace...
If you let others insert scripts into your pages they can steal your visitors.
Maybe it'll make sites think about who they script src from.
One of the bad things I've noticed recently is that HSBC is including objects from third party organisations in their ebanking login pages. I do wonder if any thought has gone into the security of such things, or if HSBC simply don't care (my experience of banks tells me that none of them have a single clue when it comes to internet security).
I was hoping someone would pick up on this. I live in the UK, but I'm vegetarian, so I get to laugh and point at all my friends who delight in eating mystery meat. Personally, I don't see much difference between horse and cow; why eat one and not the other?
I don't see a problem with eating horse... To my mind, the problem seems to be that the suppliers obviously aren't capable of keeping track of what goes into the products, so it could have been *anything*... we just got lucky it was horse.
On the other hand, vegitarians need to be careful - uniquorn has been found in some veggy meals:)
Something with a decent sized pair of screens, a good keyboard, a good mouse, plenty of memory, fast wired ethernet and a fast CPU. And yes, I know you can plug keyboards, monitors, mice, etc into a tablet, but once you've done that you've bought everything you need for a workstation except the cheap bit (the computer itself); and rather than having a cheap but powerful machine to run it all, you have an expensive and slow tablet.
A stand alone tablet would be handy, a tablet that you can't take with you because you have to leave it docked while it processes something that should've been done on a desktop machine is not useful.
What do you mean by actual work?
I'm a software developer, but the same would seem to apply to many disciplines.
I have a dual hex core (24 hardware threads) 26GB dev box in my office, a new Mac Mini, and a Dell Touch screen all-in-one - I develop on all of them.
Ok, so would you be happy to have your dev box replaced by a tablet (since that's what you proposed - "people no longer have PCs at the office")? I didn't say people wouldn't have tablets, I said they wouldn't have them _instead_ of a PC, rather, as well as it.
The majority of people at a company (that isn't an ISV) don't need anything more powerful than an iPad to do everything they've got to do.
I'm not sure I'll agree there - software developers have a great ability to squander hardware upgrades in inefficencies (for example, the current trend of doing *everything* in javascript is nutty and requires more powerful hardware to do the same job that we were already doing in native compiled code years ago).
However, even if you're right, why would you give the secretary an ipad instead of a PC? Sure, it might be useful to have a tablet as well, but when I worked at a company that had a policy of giving everyone only a laptop (rather than a desktop as well) I frequently found myself unable to take the laptop to meetings because I would be in the middle of doing something that would preclude undocking it.
The Surface Pro is going to give them an iPad-ish form factor except it will run Windows 7/8 software that already exists - That's a pretty huge win for IT departments, especially given that they can provision the devices with group policies/AD.
Don't get me wrong, I thought that Windows RT seemed like a crazy idea because the only possible reason I can think for people buying a Windows tablet is so it can be provisioned with AD (which RT can't do). However, the requirement for such meaty hardware just to get AD support is crazyness - having such a heavy weight OS on a tablet gives you the worst of both worlds (for the record, I don't see _why_ windows 8 needs such powerful hardware. My 6 year old laptop runs a perfectly capable linux desktop, and that's the same kind of performance as you could get out of top-end Android hardware these days, so I think MS have completely lost the plot by making Windows 8 so heavy it needs heavy hardware to go with it.
and you'll find that people no longer have PCs at the office, they've got 'surface pro 3' with full blown M$ Office on it - and by that time it will run 10 hours on a charge.
From what I've seen, I'm sure some technology execs are smoking the same thing you are. However, I see no point in the future where a tablet is going to replace my workstation. I can see myself having a tablet to augment my workstation (e.g. having manuals on a tablet instead of on paper), but the actual work is always going to be done on a proper computer.
My concern are the small business with a server or 2. those need a migration path but have no dedicated it guy and refuse to pay someone to fix something that might still work but scream when the mail does not arrive anymore through 3 layers of NAT.
I tend to find that mail not arriving is a pretty good motivation for those sorts of people to pay me to make it work again...
Well - the main question that comes to my mind is: how are you going to migrate users from iv4 to ipv6?
Well you don't migrate them *from* IPv4 - you keep IPv4 running, you just add IPv6 too.
- Start by dual-stacking the ISP. At this point, everything works as it always did, all but the geeks are still limited to IPv4. - Start supplying dual-stacked routers to your new customers. Your old customers will carry on as before, your new customers will be using both protocols, but favour IPv6 where possible. Devices like the Windows 7 laptop, or Android tablet will Just Work with the IPv6 router, no configuration needed, the user probably won't even notice. Old IPv4-only devices will continue to work with IPv4.
At this point, you can leave it alone and let the natural turnover of hardware replace the old v4-only routers (old router breaks, replacement supports v6. customer wants to upgrade their wifi network, replaces router, new one does v6, etc). Or the ISP can be a bit proactive and ship out replacement routers to their old customers.
At some point, the v4 addresses will start to migrate from being real addresses to CGNATted addresses. This might be a big switch over for all the ISP's users, or they might just decide that it'll only affect new users.
Either way, at some point you will get a critical mass of IPv6-capable users, and that will allow server operators to entertain the idea of IPv6-only servers and IPv6-only services. You can see parallels with web browsers - for a long time, IE6 was very popular and all websites had to support it (even though this was very expensive to do), but eventually there were enough users using other browsers that websites said "no more IE6, you need to upgrade". The same would happen for the IPv4 to IPv6 transition - eventually there are enough users with IPv6 for service providers to say "we're not supporting IPv4 any more, if you want our service then upgrade your router". That will be what pushes the remaining people to upgrade their routers.
The remaining IPv4-only hardware is going to be more specialist stuff that doesn't need to talk to the internet anyway. We're talking things like printers, etc. Workstations and printers can talk to each other over the LAN forever more - that requires no support from the ISP.
Eventually, many many years in the future, ISPs will pull the plug on IPv4 entirely. Routers will still do DHCP and hand out IPv4 addresses to machines on the LAN but there will be no IPv4 internet connectivity. Your PC may still talk to your printer over IPv4, no one really cares.
Later, router vendors might decide no one uses IPv4 at all any more and remove it entirely - at that point, all the old legacy IPv4-only devices on the network will cease to work (without manual configuration).
As you can see, none of this requires any kind of knowledge for the home users and no home visits - it just requires a little bit of proactivity from the ISPs and router vendors. Large corporate networks will, of course, require more than this, but that's why they employ net admins.
Virgin Media say they should have ipv6 running before they run out of addresses.
Virgin Media don't seem to know what's going on, as far as I can tell. Last summer they said "by the end of 2012" - that clearly didn't happen. Now they just keep saying "we don't need IPv6, we've got plenty of IPv4 addresses" - completely missing the point that it doesn't matter how many spare IPv4 addresses they have if their customer needs to talk to someone who hasn't got any spare themselves.
Other ISPs offer ipv6 today and are available to anyone with broadband via a BT land line.
I'm not aware of any ADSL ISPs offering IPv6 other than the 3 mentioned in the article - 3 that are several times the price of the ISPs that the average person uses. Until the average joe gets IPv6 as standard from their cheapy ISP, it won't have the critical mass needed to aleviate the impending problems.
>I'd have to imagine the percentage of UK homes that have internet access they would care about but who cannot get it via either BT or Virgin Media is very small.
The issue isn't whether the *home user* cares about it - the vast majority of home users don't have a clue about networking and don't know how the internet works. The issue is that until the vast majority of users have IPv6 connectivity, no one can entertain the idea of running a server without an IPv4 address, and that's going to be a big problem when the server operators can't *get* and IPv4 address.
There are ISPs which offer IPv6 over DSL to all areas of the UK. So, at least in UK, IPv6 is available for anyone who can get DSL.
The only ISPs I'm aware of that do this are the 3 mentioned (Entanet (who I use), A&A and Fluidata (who I've never heard of before)) - none of them are exactly mainstream ISPs, and their prices are far above what the mainstream ISPs charge. I guess you can say you get what you pay for, but until the mainstream ISPs start rolling out v6 it really isn't going to get much of a foothold.
There are several groups to consider:
- Hosting centres. They can see the writing on the wall and most have already rolled out IPv6 connectivity. - Server operators. I suspect most of these haven't considered IPv6. However, these are going to be the first people to be hit with IPv4 shortages since this group of people *can't* exist behind NAT. For the most part though, there's still a lot of "what's the point? almost none of our customers can do IPv6 and they can all do IPv4" going on, and I suspect that will continue going on until server operators really can't get an IPv4 address for their new server (and possibly a bit beyond!). - Transit networks. These seem to have got the message and have been running IPv6 for many years. - ISPs. Mostly they seem to be in no hurry to roll out IPv6 because most of their customers aren't asking for it and for the time being, almost everything is accessible over IPv4 anyway. They also say there are no consumer grade routers around that do IPv6. - Home users. Most have no knowledge of networking at all and just plug things together and expect them to work. They largely aren't in any position to be demanding IPv6 because, frankly, they don't know anything about it. If someone like Google launched a big new service that was v6 only and told people "you need IPv6 to use this, talk to your ISP about turning it on" then things may change here though. - Corporates. Largely IPv6 isn't really on the horizon at all for most, as far as I can see. All my company's products have been IPv6 compatible for half a decade, but to date not one of my customers is actually using IPv6 anywhere. And even if they did want to, none of their ISPs will supply it (and these tend to be big leased lines, not your average cheapy ADSL provider). - Hardware vendors. This is a mixed bag - the big names in networking like Cisco do IPv6, but there's still an aweful lot of kit that just plain doesn't do IPv6. If you're buying brand new hardware _now_ that doesn't do IPv6 then you're crazy, but thats exactly what everyone seems to be doing. Things have got a little better in the consumer grade router space - you can now get IPv6 capable ADSL routers, albeit for 3 or 4 times the price of IPv4-only ones. You do have to look carefully though - some of the stuff sold as "IPv6 ready" is only "ready" in that it doesn't support IPv6 at all but if you're very lucky the vendor might eventually make a firmware update available in a few years, if they can be arsed. - Application vendors. Many applications still have no concept of IPv6 support, although this is getting better. The fear here is that applications that don't play well with NAT (such as VoIP, etc.) are going to be locked out, and even if the vendors build IPv6 support into them to circumvent the problem, it'll be next to useless until the network operators roll out IPv6.
Unfortunately, there are a lot of interdependencies, and almost everyone is being very conservative and spending as little as possible to keep things running, rather than actually investing in the future. My fear is that the long run of "spend as little as possible every so often" will be much more expensive than biting the bullet and doing a big upgrade now, and we'll be left with an internet thats kludged to hell and very inflexible.
I certainly see CGNAT as a requirement going forward, but it needs to be done along side an IPv6 rollout so there is a workable option for applications that w
I presume the "special" pricing you get if you're a large organisation and say to MS, "we're going to switch to linux to save money and then talk to the press about it"
Power off is now visible by default in 3.6, so this niggle has been fixed.
I'm aware of that, it was just one example of several similar WTFs - I just can't figure out what they were thinking to implement that kind of misfeature in the first place.
You can re-enable the minimise button with GNOME Tweak Tool. However, because there is no taskbar work-alike, minimised windows don't visibly "go anywhere" and you have to alt-tab or use the Activities view to switch back to them.
As you point out, minimised windows are essentially lost as far as the UI goes, which makes enabling it a bit pointless - adding a panel to the overview screen with all your minimised windows in would've been a good idea. I don't subscribe to the whole "move them to another workspace" idea that the G3 devs were pushing as the replacement for minimise - moving a load of windows to another workspace, and then finding them again later is a lot more effort than minimising them.
However, since we still have window shade mode, I'm not that bothered by this one.
4. Focus stealing is a problem - Compiz/Beryl did a much better job at stopping this.
Really? I have no issues in this regard.
Stuff focus-steals (by means of popping up a new window that automagically gets focus even though it wasn't the focussed application to begin with) with reasonable regularity for me.
In fact, if you liked E17 because it got out of your way, I'm surprised you're so happy with GNOME 3, where the shell forcibly takes over whenever you use the Activities view to switch windows/launch things. That doesn't fit with my definition of getting out of the way.
Well yes and no. The Gnome shell largely _is_ "out of the way" - the only Gnome stuff visible most of the time is the bar at the top of the screen, which is much smaller than the old panel. I have to actively ask for the activities screen, so it doesn't get in my way when I don't want it (yes, I know the old panel was hideable). I guess the main thing is that the in-your-face bits of Gnome 3, such as activities, tend to do more or less exactly what I want so I don't mind them being in my face, whereas with many other DEs I've tried the in-your-face bits seemed to mostly just get in the way.
I suppose the one bit of Gnome 3 that does get in the way for me is the notification bar. It pops up and covers things I'm working on (working primarilly in terminal windows means that the most important bit of the window is often right at the bottom, which is the bit that gets covered by the notification bar). Also I tend to find that Empathy's IM notifications pop up when I'm not looking and then vanish without trace so I don't actually see them until I manually bring up the notification bar. That may have been fixed though - as mentioned I'm running Fedora 16, so not up with the latest version of Gnome 3. I think its hard to get the right balance for notifications, to some extent it would be nice to have eye tracking hardware on the computer that can figure out when you've seen a notification and can automagically get rid of it.
What *does* bother me in 3.6, and has me worried for the future, is the stupid application menu in the top panel.
I sincerely hope this isn't going the way of OS X - detatching an application's menus from the app's window itself is one of the worst things you can do. 1. It is completely incompatible with focus-follows-mouse. 2. It means you have to move the mouse much further - this is especially a big deal with large/multiple screens - I don't want to have to move my mouse from the bottom-right of the right-most 24" screen to the top-left of the left-most 24" screen in order to perform some menu action. 3. You have to be doubley sure which window is focussed before you start poking around at the menu.
To be completely honest, the bits that Gnome 3 see
but now that they're going off on the same kind of tangent that brought us Gnome 3 (which nobody is really very hot on, but they tolerate it)
I'm always curious what people find bad about Gnome 3 - it's about the best desktop environment I've used IMHO. There are a few niggles: 1. They have removed some options for the sake of simplicity which really shouldn't have been removed, and wouldn't have made things more complex if they were there. For example, not having a "disabled" option in the DPMS timeout is a WTF, as is hiding the "Power off" option until you press the magic "Alt" key (with no visual cues that you could press Alt to get at it). 2. They have copied apple's unintuitive mode-switching for the launcher icons. When I click on a launcher icon I always want it to launch a new window for that application; changing to a "raise all the existing windows that application has open" if it has at least one window open is idiotic, I can't think of a time I ever want to raise all 15 open terminal windows at the same time instead of openning a new one. 3. They've abolished minimise. However, shading windows is still possible and that's almost as good. 4. Focus stealing is a problem - Compiz/Beryl did a much better job at stopping this.
I should mention that I thought Gnome 2 was pretty much the worst DE I've ever used - that seemed to be an exercise in copying the worst bits of Windows and removing any options that might make it tollerable. Up until Gnome 3, I was using E17, largely because it was light, fast and just got out of my way when I wanted to get stuff done.
Well, given the immaturity surrounding Fedora 18's name and the questionability of the installer, we'll probably be moving everything to RHEL and keeping the entire office as a split Windows | RHEL setup.
I'd say that making a policy decision based on the name that has been given to a release is a bit daft - I care not what things are called, only how well they do the job I need them to do. However, I haven't tried F18 so can't comment on the quality of the software - I'm currently still running F16 on my workstations, and will likely skip F18, given the poor reviews.
Given that you said you've been happy with every release until now, why not wait until F19 and see if the problems have been fixed? This certainly isn't the first time Fedora have jumped the shark by pushing a technology that was nowhere near ready (can you say "SElinux" and "Pulse Audio"?), but generally when they've done it, its been largely fixed up within 1 or 2 releases.
We're largely running Fedora workstations and Scientific Linux servers here (occasionally Debian for non-x86 hardware), and it seems to work well.
In the UK for example, it is illegal to access a computer system without the owner's permission. You could argue that the TOS set out the terms under which the owner is prepared to give permission to access their system, and if you violate them, you don't have permission to access the site, and are therefore breaking the law.
The computer misuse act, as it stands, doesn't really apply very well to modern networks. It says you need permission to access a computer system - well when was the last time you asked google for written permission to use their website? The act come from a time when computer systems didn't tend to be open to the public, so demanding that you get permission before accessing it was sane; now it isn't.
Also, I will perpetually argue that one-time click-through licences aren't worth anything - you're not collecting a signature so how do you prove that a specific person agreed to the licence?
With NAT, an outside system can't initiate connectivity with any machine inside the NAT boundary without some kind of prior arrangement
That's untrue. Most consumer NAT routers (at least the ones I tested about 3 years ago - doubt its really changed) don't bother to include a stateful firewall and with appropriate ISP-side routing, will happilly let connections into the private network. What you need is a stateful firewall, not NAT - that will protect you, and also doesn't completely fuck up loads of protocols at the same time.
The depressing thing (other than idiots claiming that NAT is good for security) is that Plusnet *were* trialling IPv6, but pulled the plug on the trial last year. When I asked them a month or so ago, they informed me that they had no plans to roll out IPv6 at all. Time to switch to a competent ISP if you're with Plusnet, I suspect (EntaNet and AAISP both offer v6 connections over DSL).
If we ALL paid flood insurance, it would cost us 2p each a year. If only those who live in flood plains pay it, they might as well just put it in the bank and pay costs of each flood as it happens because it's only the high-risk people who are subsidising the majority of the insurance anyway.
I agree with everything you said right up to this point. I certainly think that discriminating against people based on something that is out of their control is bad and defeats the point of insurance. However, where something is within your control, there should be a penalty for making the more risky choice in order to incentivise people not to do it. If you're stupid enough to *choose* to buy a house that's built on a flood plain, you shouldn't expect those of us who picked a less risky location to pick up the tab for your bad choices. Building on flood plains is a really stupid idea (not only because you're likely to get flooded, but because it also produces new flood plains downstream, flooding previously safe properties), and the only way its going to stop is if there's a big penalty for doing it (being unable to insure a property would probably make people think twice before buying it, which would make builders think twice before building potentially unsalable properties in a crazy location).
And then people wonder why there are areas of London, say, where you cannot get insurance for your car because NOBODY there has insurance (Tottenham was in the news just last year for this - it's so hard to get insurance, because nobody else has it in the local area and it costs the insurer's money to pursue them when there's an accident, that nobody has insurance - something like 40% of drivers registered to Tottenham addresses are uninsured!).
The state of the UK car insurance industry is insane. The insurers claim that they need to increase premiums to cover the number of uninsured drivers, which only encourages more people to avoid insuring because they can no longer justify the increased price, so the premiums go up again, rinse-repeat. Add to that that the legal penalties if you're caught without insurance are often a fraction of what you would've paid for the insurance in the first place. And this is a huge problem for new drivers - a couple of years ago, out of curiosity, I looked at the price of insurance for new drivers with a 10 year old 1.0 litre Corsa and it was around £2000/year for third party only. What kind of new driver can afford that kind of money? Introducing new drivers to the idea that insurance is unaffordably expensive just trains them to avoid insuring, and they will likely continue for many years.
Notice, then, that car insurance rising because women have the pay the same as men now (i.e. closer to "real" insurance), is red-tape and bullshit and not related to the legislation at all.
I was interested that SwiftCover recently sent round a mailshot demanding that "due to recent antidiscrimination legislation" they require all their customers to provide additional information about themselves. Which I read as "because the legislators have stopped us discriminating against men, please provide more information about yourself so we can find something else to base our discrimination on."
Yet another example of a law taking away your opportunity. Both for the customer and the producer.
Yep, yet another example of a law taking away a manufacturer's opportunity to sell expensive and badly constructed crap that won't last more than a few months. Honestly, I would be very wary of buying from any manufacturer who has enough concerns about their hardware failing in short-order that they will publically flout the law to avoid having to take on any of the financial risk of it doing so.
If only more people believed this. Then you couldn't accuse Google of having a monopoly since an alternative is merely a free click away.
I'm not accusing them of having a monopoly, nor is the EU. They are saying that Google is in a dominant position in the search market, which I don't think anyone is really disputing - they *are* in a dominant position for whatever reason, and that doesn't mean there are no/few alternatives, it means that not many people are using the alternatives.
Furthermore, the EU is investigating the possibility that Google is leveraging their dominant position in the search market to give them an unfair advantage in other markets, which is illegal. This isn't completely cut & dry, which is why the EU has been investigating for some time.
If I wanted a Bing map to tell me where the business was I wouldn't have gone to Google's website.
Where would you go if you just wanted the most relevant result instead of having the results filtered based on who owns the website they are pointing at?
That's like complaining that a pub serves beer they brewed themselves instead of the beer brewed next door.
Only if the pub happens to be the world's most popular place to drink - if it is then you might have a point (and they might be being investigated for similar reasons - using their position as the most popular pub to muscle in on the brewing market, at the expense of other brewers who don't own any pubs).
When the EU starts trying to prevent me from getting the results I want from Google by making some pissant mappy result come up first rather than the market leader which is Google Maps, then No, the EU does not understand!.
Who said they were going to do that? Do you want the most relevant result, no matter where its found on the internet, or do you only ever want to see content hosted by Google?
The browser selection menu imposed on MS shows just what the EU is searching for: "fairness" not for best results.
Are they not the same thing (at least in this case)? If all sites were treated by Google in a fair way then that would seem to satisfy both "fair" and "best" unless your definition of "best" is just "must be owned by Google".
They want a system where search results will come up in a random order weighted by market share.
[Citation needed]
The EU just want things to be fair across the board here. That simply means that Google shouldn't give their own products special treatment compared to similar products from other vendor. That doesn't mean that everyone must have an equal change of appearing at the top, it just means that google can't use the "stick it at the top if its ours" criteria.
Slashdot Mirror
What was the apple one? I don't recollect it
The only one I can remember was when the server that responds to WISPr probes went down, rendering everyone's ipad unable to connect to a network...
Background:
When an iOS device associates with a wifi network, it makes a web request to apple's server to see if its behind a captive portal. It expects to get back "SUCCESS" (returned by Apple's server) or a captive portal login page (returned by the wifi hotspot). If it doesn't get "SUCCESS" it displays the captive portal page so the user can log in. Unfortunately, Apple's software is unable to cope with the web request not being answered at all, and you end up with a blank "login" page and a non-funcational network connection. Yes, this is the usual quality I've come to expect from Apple, given the numerous problems I've had to deal with since iOS devices started to get popular in the workplace...
If you let others insert scripts into your pages they can steal your visitors.
Maybe it'll make sites think about who they script src from.
One of the bad things I've noticed recently is that HSBC is including objects from third party organisations in their ebanking login pages. I do wonder if any thought has gone into the security of such things, or if HSBC simply don't care (my experience of banks tells me that none of them have a single clue when it comes to internet security).
I was hoping someone would pick up on this. I live in the UK, but I'm vegetarian, so I get to laugh and point at all my friends who delight in eating mystery meat. Personally, I don't see much difference between horse and cow; why eat one and not the other?
I don't see a problem with eating horse... To my mind, the problem seems to be that the suppliers obviously aren't capable of keeping track of what goes into the products, so it could have been *anything*... we just got lucky it was horse.
On the other hand, vegitarians need to be careful - uniquorn has been found in some veggy meals :)
Unlike US, trespass in UK applies *only* to physical locations.
Also, trespass only applies if you fail to leave the location after being asked to. If no one asks you to leave, you're not trespassing.
However, I see no point in the future where a tablet is going to replace my workstation. I can see myself having a tablet to augment my workstation
No such future is being suggested by the Surface.
Can you please go back and read the email I was replying to, which said:
you'll find that people no longer have PCs at the office, they've got 'surface pro 3' with full blown M$ Office on it
Clearly the poster was saying that the tablet would replace the workstation.
What exactly do you consider a proper computer?
Something with a decent sized pair of screens, a good keyboard, a good mouse, plenty of memory, fast wired ethernet and a fast CPU. And yes, I know you can plug keyboards, monitors, mice, etc into a tablet, but once you've done that you've bought everything you need for a workstation except the cheap bit (the computer itself); and rather than having a cheap but powerful machine to run it all, you have an expensive and slow tablet.
A stand alone tablet would be handy, a tablet that you can't take with you because you have to leave it docked while it processes something that should've been done on a desktop machine is not useful.
What do you mean by actual work?
I'm a software developer, but the same would seem to apply to many disciplines.
I have a dual hex core (24 hardware threads) 26GB dev box in my office, a new Mac Mini, and a Dell Touch screen all-in-one - I develop on all of them.
Ok, so would you be happy to have your dev box replaced by a tablet (since that's what you proposed - "people no longer have PCs at the office")? I didn't say people wouldn't have tablets, I said they wouldn't have them _instead_ of a PC, rather, as well as it.
The majority of people at a company (that isn't an ISV) don't need anything more powerful than an iPad to do everything they've got to do.
I'm not sure I'll agree there - software developers have a great ability to squander hardware upgrades in inefficencies (for example, the current trend of doing *everything* in javascript is nutty and requires more powerful hardware to do the same job that we were already doing in native compiled code years ago).
However, even if you're right, why would you give the secretary an ipad instead of a PC? Sure, it might be useful to have a tablet as well, but when I worked at a company that had a policy of giving everyone only a laptop (rather than a desktop as well) I frequently found myself unable to take the laptop to meetings because I would be in the middle of doing something that would preclude undocking it.
The Surface Pro is going to give them an iPad-ish form factor except it will run Windows 7/8 software that already exists - That's a pretty huge win for IT departments, especially given that they can provision the devices with group policies/AD.
Don't get me wrong, I thought that Windows RT seemed like a crazy idea because the only possible reason I can think for people buying a Windows tablet is so it can be provisioned with AD (which RT can't do). However, the requirement for such meaty hardware just to get AD support is crazyness - having such a heavy weight OS on a tablet gives you the worst of both worlds (for the record, I don't see _why_ windows 8 needs such powerful hardware. My 6 year old laptop runs a perfectly capable linux desktop, and that's the same kind of performance as you could get out of top-end Android hardware these days, so I think MS have completely lost the plot by making Windows 8 so heavy it needs heavy hardware to go with it.
and you'll find that people no longer have PCs at the office, they've got 'surface pro 3' with full blown M$ Office on it - and by that time it will run 10 hours on a charge.
From what I've seen, I'm sure some technology execs are smoking the same thing you are. However, I see no point in the future where a tablet is going to replace my workstation. I can see myself having a tablet to augment my workstation (e.g. having manuals on a tablet instead of on paper), but the actual work is always going to be done on a proper computer.
My concern are the small business with a server or 2. those need a migration path but have no dedicated it guy and refuse to pay someone to fix something that might still work but scream when the mail does not arrive anymore through 3 layers of NAT.
I tend to find that mail not arriving is a pretty good motivation for those sorts of people to pay me to make it work again...
The simulated "fuel" is ethanol. So no one, particularly the Russians, are complaining.
I dunno, in an accident losing that much vodka to the vacuum of space would be a disaster!
Well - the main question that comes to my mind is: how are you going to migrate users from iv4 to ipv6?
Well you don't migrate them *from* IPv4 - you keep IPv4 running, you just add IPv6 too.
- Start by dual-stacking the ISP. At this point, everything works as it always did, all but the geeks are still limited to IPv4.
- Start supplying dual-stacked routers to your new customers. Your old customers will carry on as before, your new customers will be using both protocols, but favour IPv6 where possible. Devices like the Windows 7 laptop, or Android tablet will Just Work with the IPv6 router, no configuration needed, the user probably won't even notice. Old IPv4-only devices will continue to work with IPv4.
At this point, you can leave it alone and let the natural turnover of hardware replace the old v4-only routers (old router breaks, replacement supports v6. customer wants to upgrade their wifi network, replaces router, new one does v6, etc). Or the ISP can be a bit proactive and ship out replacement routers to their old customers.
At some point, the v4 addresses will start to migrate from being real addresses to CGNATted addresses. This might be a big switch over for all the ISP's users, or they might just decide that it'll only affect new users.
Either way, at some point you will get a critical mass of IPv6-capable users, and that will allow server operators to entertain the idea of IPv6-only servers and IPv6-only services. You can see parallels with web browsers - for a long time, IE6 was very popular and all websites had to support it (even though this was very expensive to do), but eventually there were enough users using other browsers that websites said "no more IE6, you need to upgrade". The same would happen for the IPv4 to IPv6 transition - eventually there are enough users with IPv6 for service providers to say "we're not supporting IPv4 any more, if you want our service then upgrade your router". That will be what pushes the remaining people to upgrade their routers.
The remaining IPv4-only hardware is going to be more specialist stuff that doesn't need to talk to the internet anyway. We're talking things like printers, etc. Workstations and printers can talk to each other over the LAN forever more - that requires no support from the ISP.
Eventually, many many years in the future, ISPs will pull the plug on IPv4 entirely. Routers will still do DHCP and hand out IPv4 addresses to machines on the LAN but there will be no IPv4 internet connectivity. Your PC may still talk to your printer over IPv4, no one really cares.
Later, router vendors might decide no one uses IPv4 at all any more and remove it entirely - at that point, all the old legacy IPv4-only devices on the network will cease to work (without manual configuration).
As you can see, none of this requires any kind of knowledge for the home users and no home visits - it just requires a little bit of proactivity from the ISPs and router vendors. Large corporate networks will, of course, require more than this, but that's why they employ net admins.
Virgin Media say they should have ipv6 running before they run out of addresses.
Virgin Media don't seem to know what's going on, as far as I can tell. Last summer they said "by the end of 2012" - that clearly didn't happen. Now they just keep saying "we don't need IPv6, we've got plenty of IPv4 addresses" - completely missing the point that it doesn't matter how many spare IPv4 addresses they have if their customer needs to talk to someone who hasn't got any spare themselves.
Other ISPs offer ipv6 today and are available to anyone with broadband via a BT land line.
I'm not aware of any ADSL ISPs offering IPv6 other than the 3 mentioned in the article - 3 that are several times the price of the ISPs that the average person uses. Until the average joe gets IPv6 as standard from their cheapy ISP, it won't have the critical mass needed to aleviate the impending problems.
>I'd have to imagine the percentage of UK homes that have internet access they would care about but who cannot get it via either BT or Virgin Media is very small.
The issue isn't whether the *home user* cares about it - the vast majority of home users don't have a clue about networking and don't know how the internet works. The issue is that until the vast majority of users have IPv6 connectivity, no one can entertain the idea of running a server without an IPv4 address, and that's going to be a big problem when the server operators can't *get* and IPv4 address.
There are ISPs which offer IPv6 over DSL to all areas of the UK. So, at least in UK, IPv6 is available for anyone who can get DSL.
The only ISPs I'm aware of that do this are the 3 mentioned (Entanet (who I use), A&A and Fluidata (who I've never heard of before)) - none of them are exactly mainstream ISPs, and their prices are far above what the mainstream ISPs charge. I guess you can say you get what you pay for, but until the mainstream ISPs start rolling out v6 it really isn't going to get much of a foothold.
There are several groups to consider:
- Hosting centres. They can see the writing on the wall and most have already rolled out IPv6 connectivity.
- Server operators. I suspect most of these haven't considered IPv6. However, these are going to be the first people to be hit with IPv4 shortages since this group of people *can't* exist behind NAT. For the most part though, there's still a lot of "what's the point? almost none of our customers can do IPv6 and they can all do IPv4" going on, and I suspect that will continue going on until server operators really can't get an IPv4 address for their new server (and possibly a bit beyond!).
- Transit networks. These seem to have got the message and have been running IPv6 for many years.
- ISPs. Mostly they seem to be in no hurry to roll out IPv6 because most of their customers aren't asking for it and for the time being, almost everything is accessible over IPv4 anyway. They also say there are no consumer grade routers around that do IPv6.
- Home users. Most have no knowledge of networking at all and just plug things together and expect them to work. They largely aren't in any position to be demanding IPv6 because, frankly, they don't know anything about it. If someone like Google launched a big new service that was v6 only and told people "you need IPv6 to use this, talk to your ISP about turning it on" then things may change here though.
- Corporates. Largely IPv6 isn't really on the horizon at all for most, as far as I can see. All my company's products have been IPv6 compatible for half a decade, but to date not one of my customers is actually using IPv6 anywhere. And even if they did want to, none of their ISPs will supply it (and these tend to be big leased lines, not your average cheapy ADSL provider).
- Hardware vendors. This is a mixed bag - the big names in networking like Cisco do IPv6, but there's still an aweful lot of kit that just plain doesn't do IPv6. If you're buying brand new hardware _now_ that doesn't do IPv6 then you're crazy, but thats exactly what everyone seems to be doing. Things have got a little better in the consumer grade router space - you can now get IPv6 capable ADSL routers, albeit for 3 or 4 times the price of IPv4-only ones. You do have to look carefully though - some of the stuff sold as "IPv6 ready" is only "ready" in that it doesn't support IPv6 at all but if you're very lucky the vendor might eventually make a firmware update available in a few years, if they can be arsed.
- Application vendors. Many applications still have no concept of IPv6 support, although this is getting better. The fear here is that applications that don't play well with NAT (such as VoIP, etc.) are going to be locked out, and even if the vendors build IPv6 support into them to circumvent the problem, it'll be next to useless until the network operators roll out IPv6.
Unfortunately, there are a lot of interdependencies, and almost everyone is being very conservative and spending as little as possible to keep things running, rather than actually investing in the future. My fear is that the long run of "spend as little as possible every so often" will be much more expensive than biting the bullet and doing a big upgrade now, and we'll be left with an internet thats kludged to hell and very inflexible.
I certainly see CGNAT as a requirement going forward, but it needs to be done along side an IPv6 rollout so there is a workable option for applications that w
You mean made up pricing?
I presume the "special" pricing you get if you're a large organisation and say to MS, "we're going to switch to linux to save money and then talk to the press about it"
Power off is now visible by default in 3.6, so this niggle has been fixed.
I'm aware of that, it was just one example of several similar WTFs - I just can't figure out what they were thinking to implement that kind of misfeature in the first place.
You can re-enable the minimise button with GNOME Tweak Tool. However, because there is no taskbar work-alike, minimised windows don't visibly "go anywhere" and you have to alt-tab or use the Activities view to switch back to them.
As you point out, minimised windows are essentially lost as far as the UI goes, which makes enabling it a bit pointless - adding a panel to the overview screen with all your minimised windows in would've been a good idea. I don't subscribe to the whole "move them to another workspace" idea that the G3 devs were pushing as the replacement for minimise - moving a load of windows to another workspace, and then finding them again later is a lot more effort than minimising them.
However, since we still have window shade mode, I'm not that bothered by this one.
4. Focus stealing is a problem - Compiz/Beryl did a much better job at stopping this.
Really? I have no issues in this regard.
Stuff focus-steals (by means of popping up a new window that automagically gets focus even though it wasn't the focussed application to begin with) with reasonable regularity for me.
In fact, if you liked E17 because it got out of your way, I'm surprised you're so happy with GNOME 3, where the shell forcibly takes over whenever you use the Activities view to switch windows/launch things. That doesn't fit with my definition of getting out of the way.
Well yes and no. The Gnome shell largely _is_ "out of the way" - the only Gnome stuff visible most of the time is the bar at the top of the screen, which is much smaller than the old panel. I have to actively ask for the activities screen, so it doesn't get in my way when I don't want it (yes, I know the old panel was hideable). I guess the main thing is that the in-your-face bits of Gnome 3, such as activities, tend to do more or less exactly what I want so I don't mind them being in my face, whereas with many other DEs I've tried the in-your-face bits seemed to mostly just get in the way.
I suppose the one bit of Gnome 3 that does get in the way for me is the notification bar. It pops up and covers things I'm working on (working primarilly in terminal windows means that the most important bit of the window is often right at the bottom, which is the bit that gets covered by the notification bar). Also I tend to find that Empathy's IM notifications pop up when I'm not looking and then vanish without trace so I don't actually see them until I manually bring up the notification bar. That may have been fixed though - as mentioned I'm running Fedora 16, so not up with the latest version of Gnome 3. I think its hard to get the right balance for notifications, to some extent it would be nice to have eye tracking hardware on the computer that can figure out when you've seen a notification and can automagically get rid of it.
What *does* bother me in 3.6, and has me worried for the future, is the stupid application menu in the top panel.
I sincerely hope this isn't going the way of OS X - detatching an application's menus from the app's window itself is one of the worst things you can do.
1. It is completely incompatible with focus-follows-mouse.
2. It means you have to move the mouse much further - this is especially a big deal with large/multiple screens - I don't want to have to move my mouse from the bottom-right of the right-most 24" screen to the top-left of the left-most 24" screen in order to perform some menu action.
3. You have to be doubley sure which window is focussed before you start poking around at the menu.
To be completely honest, the bits that Gnome 3 see
but now that they're going off on the same kind of tangent that brought us Gnome 3 (which nobody is really very hot on, but they tolerate it)
I'm always curious what people find bad about Gnome 3 - it's about the best desktop environment I've used IMHO. There are a few niggles:
1. They have removed some options for the sake of simplicity which really shouldn't have been removed, and wouldn't have made things more complex if they were there. For example, not having a "disabled" option in the DPMS timeout is a WTF, as is hiding the "Power off" option until you press the magic "Alt" key (with no visual cues that you could press Alt to get at it).
2. They have copied apple's unintuitive mode-switching for the launcher icons. When I click on a launcher icon I always want it to launch a new window for that application; changing to a "raise all the existing windows that application has open" if it has at least one window open is idiotic, I can't think of a time I ever want to raise all 15 open terminal windows at the same time instead of openning a new one.
3. They've abolished minimise. However, shading windows is still possible and that's almost as good.
4. Focus stealing is a problem - Compiz/Beryl did a much better job at stopping this.
I should mention that I thought Gnome 2 was pretty much the worst DE I've ever used - that seemed to be an exercise in copying the worst bits of Windows and removing any options that might make it tollerable. Up until Gnome 3, I was using E17, largely because it was light, fast and just got out of my way when I wanted to get stuff done.
Well, given the immaturity surrounding Fedora 18's name and the questionability of the installer, we'll probably be moving everything to RHEL and keeping the entire office as a split Windows | RHEL setup.
I'd say that making a policy decision based on the name that has been given to a release is a bit daft - I care not what things are called, only how well they do the job I need them to do. However, I haven't tried F18 so can't comment on the quality of the software - I'm currently still running F16 on my workstations, and will likely skip F18, given the poor reviews.
Given that you said you've been happy with every release until now, why not wait until F19 and see if the problems have been fixed? This certainly isn't the first time Fedora have jumped the shark by pushing a technology that was nowhere near ready (can you say "SElinux" and "Pulse Audio"?), but generally when they've done it, its been largely fixed up within 1 or 2 releases.
We're largely running Fedora workstations and Scientific Linux servers here (occasionally Debian for non-x86 hardware), and it seems to work well.
Previously, installing SELinux couldn't be prevented and disabling it caused the boot process to fail.
Untrue - I've had SELinux turned off on my workstation on every version of Fedora since it was introduced. Works fine.
In the UK for example, it is illegal to access a computer system without the owner's permission. You could argue that the TOS set out the terms under which the owner is prepared to give permission to access their system, and if you violate them, you don't have permission to access the site, and are therefore breaking the law.
The computer misuse act, as it stands, doesn't really apply very well to modern networks. It says you need permission to access a computer system - well when was the last time you asked google for written permission to use their website? The act come from a time when computer systems didn't tend to be open to the public, so demanding that you get permission before accessing it was sane; now it isn't.
Also, I will perpetually argue that one-time click-through licences aren't worth anything - you're not collecting a signature so how do you prove that a specific person agreed to the licence?
With NAT, an outside system can't initiate connectivity with any machine inside the NAT boundary without some kind of prior arrangement
That's untrue. Most consumer NAT routers (at least the ones I tested about 3 years ago - doubt its really changed) don't bother to include a stateful firewall and with appropriate ISP-side routing, will happilly let connections into the private network. What you need is a stateful firewall, not NAT - that will protect you, and also doesn't completely fuck up loads of protocols at the same time.
The depressing thing (other than idiots claiming that NAT is good for security) is that Plusnet *were* trialling IPv6, but pulled the plug on the trial last year. When I asked them a month or so ago, they informed me that they had no plans to roll out IPv6 at all. Time to switch to a competent ISP if you're with Plusnet, I suspect (EntaNet and AAISP both offer v6 connections over DSL).
If we ALL paid flood insurance, it would cost us 2p each a year. If only those who live in flood plains pay it, they might as well just put it in the bank and pay costs of each flood as it happens because it's only the high-risk people who are subsidising the majority of the insurance anyway.
I agree with everything you said right up to this point. I certainly think that discriminating against people based on something that is out of their control is bad and defeats the point of insurance. However, where something is within your control, there should be a penalty for making the more risky choice in order to incentivise people not to do it. If you're stupid enough to *choose* to buy a house that's built on a flood plain, you shouldn't expect those of us who picked a less risky location to pick up the tab for your bad choices. Building on flood plains is a really stupid idea (not only because you're likely to get flooded, but because it also produces new flood plains downstream, flooding previously safe properties), and the only way its going to stop is if there's a big penalty for doing it (being unable to insure a property would probably make people think twice before buying it, which would make builders think twice before building potentially unsalable properties in a crazy location).
And then people wonder why there are areas of London, say, where you cannot get insurance for your car because NOBODY there has insurance (Tottenham was in the news just last year for this - it's so hard to get insurance, because nobody else has it in the local area and it costs the insurer's money to pursue them when there's an accident, that nobody has insurance - something like 40% of drivers registered to Tottenham addresses are uninsured!).
The state of the UK car insurance industry is insane. The insurers claim that they need to increase premiums to cover the number of uninsured drivers, which only encourages more people to avoid insuring because they can no longer justify the increased price, so the premiums go up again, rinse-repeat. Add to that that the legal penalties if you're caught without insurance are often a fraction of what you would've paid for the insurance in the first place. And this is a huge problem for new drivers - a couple of years ago, out of curiosity, I looked at the price of insurance for new drivers with a 10 year old 1.0 litre Corsa and it was around £2000/year for third party only. What kind of new driver can afford that kind of money? Introducing new drivers to the idea that insurance is unaffordably expensive just trains them to avoid insuring, and they will likely continue for many years.
Notice, then, that car insurance rising because women have the pay the same as men now (i.e. closer to "real" insurance), is red-tape and bullshit and not related to the legislation at all.
I was interested that SwiftCover recently sent round a mailshot demanding that "due to recent antidiscrimination legislation" they require all their customers to provide additional information about themselves. Which I read as "because the legislators have stopped us discriminating against men, please provide more information about yourself so we can find something else to base our discrimination on."
Yet another example of a law taking away your opportunity. Both for the customer and the producer.
Yep, yet another example of a law taking away a manufacturer's opportunity to sell expensive and badly constructed crap that won't last more than a few months. Honestly, I would be very wary of buying from any manufacturer who has enough concerns about their hardware failing in short-order that they will publically flout the law to avoid having to take on any of the financial risk of it doing so.
Maybe you just want to watch the film instead of battling with the DRM that all the legit formats have...
If only more people believed this. Then you couldn't accuse Google of having a monopoly since an alternative is merely a free click away.
I'm not accusing them of having a monopoly, nor is the EU. They are saying that Google is in a dominant position in the search market, which I don't think anyone is really disputing - they *are* in a dominant position for whatever reason, and that doesn't mean there are no/few alternatives, it means that not many people are using the alternatives.
Furthermore, the EU is investigating the possibility that Google is leveraging their dominant position in the search market to give them an unfair advantage in other markets, which is illegal. This isn't completely cut & dry, which is why the EU has been investigating for some time.
That's like complaining that a pub serves beer they brewed themselves instead of the beer brewed next door.
Oh, and for the record I do tend to stick to free houses because I value choice in beer.
If I wanted a Bing map to tell me where the business was I wouldn't have gone to Google's website.
Where would you go if you just wanted the most relevant result instead of having the results filtered based on who owns the website they are pointing at?
That's like complaining that a pub serves beer they brewed themselves instead of the beer brewed next door.
Only if the pub happens to be the world's most popular place to drink - if it is then you might have a point (and they might be being investigated for similar reasons - using their position as the most popular pub to muscle in on the brewing market, at the expense of other brewers who don't own any pubs).
When the EU starts trying to prevent me from getting the results I want from Google by making some pissant mappy result come up first rather than the market leader which is Google Maps, then No, the EU does not understand!.
Who said they were going to do that? Do you want the most relevant result, no matter where its found on the internet, or do you only ever want to see content hosted by Google?
The browser selection menu imposed on MS shows just what the EU is searching for: "fairness" not for best results.
Are they not the same thing (at least in this case)? If all sites were treated by Google in a fair way then that would seem to satisfy both "fair" and "best" unless your definition of "best" is just "must be owned by Google".
They want a system where search results will come up in a random order weighted by market share.
[Citation needed]
The EU just want things to be fair across the board here. That simply means that Google shouldn't give their own products special treatment compared to similar products from other vendor. That doesn't mean that everyone must have an equal change of appearing at the top, it just means that google can't use the "stick it at the top if its ours" criteria.