Your argument seems baseless and has nothing to do with the text of the person you're attacking. Lucractius said that your artist daughter should work harder, perfect her craft and market the hell out of herself. How is this a "typical engineer's attitude towards art" ? That's a sane attitude to have in a world where anything can be made cheaper and faster elsewhere. Do you honestly believe that an artist has some magical income forever job? Most artists I know don't live very high.
You clearly have no idea what you're talking about. UBB was never about paying for what you use. It was a way for Bell/Rogers to force the wholesalers into offering packages that were no better than what Bell/Rogers were providing. The incumbents were using their monopoly on the infrastructure for the last mile to try and "level" the playing field of the competitive carriers.
I have absolutely no problem with paying for what I use. The problem is that the pricing they're trying to force is not reasonable. It doesn't cost even $0.10 for 1MB delivered in a second, yet they were trying to charge well over that per KILObit. I kind of wish that the CRTC gave them exactly what they asked for. I wish the CRTC gave them usage based billing, but enforced pricing based on legitimate data costs plus a legitimate profit. If they wanted to make more money than that, they'd have to find a way to cut costs without cutting service.
Connectivity in Canada is terribly, terribly out of touch with reality. There are a few monopolies who still own a huge chunk of the last mile, and short of cutting all ties between the last-mile provider and the bandwidth/content provider, nothing will fix this.
I spent 15 years in industrial power, working with the exact communications and control networks you complain about. I know what you're talking about and sympathize to a large extent.
Your boss' request, however, does not compromize security. There is nothing insecure about providing read-only access to the network. A secure device sitting on the industrial network could export/expose the system status through a variety of means, none of which would accept requests to alter the system state. At a very extreme case, a unidirectional optical link could stream system state from the critical side of the device to the noncritical one, and the insecure side could then pick out what it wants to report on. There's enough bandwidth available in gigabit ethernet and the filtering problem is largely solved already (and has been for well over a decade with CAN), which is where DeviceNet and Ethernet/IP get their roots from.
It'd be an uphill battle, but I'd like to design a PLC module that works exactly this way. I miss the industrial work.:-)
You know, I used to do what you are doing... I hated the idea of getting rid of old hardware. It's perfectly functional and let's face it, the computational power required to run a printer or file server or even a light mail server is very, very low.
What changed my opinion on this was coming to the realization that the power consumption of these older devices is not small. That old P90 running idle can't compare to a modern core2 or iwhatever system. Atom processors are useless; their idle power is slightly lower than a core2, but any time you have to make it do something it ends up taking more power to do it than it takes the core2 to spin up, do the task and get back to idle.
It still bothers me a bit when I decide to recycle an otherwise functional piece of old equipment, but in the mid/long run I'm actually saving money by consolidating equipment (mail+print+fileserver instead of 3 small servers) and making conscious decisions about the power consumption of the parts in the servers.
You seem to think nobody is even considering it. That's certainly not true.
I'm just really concerned that the decision to "design in network transparency" will be what is said, but when everyone's salivating over how cool and fast and slick it's going to be that design decision will be marginalized. I'd love to get some kind of involvement in Wayland but I know that I haven't even the slightest amount of time to be useful in such a project.
I guess that's why I'm posting really concerned comments on a random web forum.:-/
Oh I see. my mistake for not even clicking the links. Sorry.
I installed the 4.7 ppa for Ubuntu 11.04 last night. You are right, 4.7 cleaned up a lot of little things that were bugging me in 4.6. I have not moved to the 4.7 kdepim yet (that's in the experimental ppa) but 4.7 so far offers considerable relief from 4.6 (at least on Ubuntu).
I want to thank you for managing all the comments in this story. I really, really do love KDE and I want to use it for a good long time to come. I think it's obvious from the comments that there are a LOT of other people like me who want to see KDE succeed but we're very concerned over the way development seems to be more concerned over bells and whistles instead of speed and stability.
um... kmail/kontact has had this feature for a LONG time... I was complaining that I am going to miss it if I really do have to move off of kontact for my email/calendaring app.
The devs repeatedly said it wasn't for everyday use for everyone, and that it was mainly for developers to have a base to build from.
As both a developer and a picky user, perhaps the KDE team should NOT have called it a 4.0 release. Perhaps a 4.0-RTD or something (release to developers) -- make it obvious from the name that it's not for users. I know that KDE tried to do that with all the warnings and everything, but let's be honest... most users are stupid. Kubuntu made a hideous mistake by including it. Ubuntu's got this problem at a systemic level (ripping out perfectly good code/apps to try to establish new paradigms before they're ready to be used) so I guess it was inevitable, but perhaps if KDE 4.0 were named KDE-4DEVRELEASE or something...
I don't know. I hope that what happened with 4.x is never again repeated. 4.7 is almost out and it's still not where it was in 3.5. Lord knows software development ain't easy, and trying to write a desktop environment for a large and varied userbase is difficult to put it mildly, but I don't think KDE really helped themselves at all with 4.x. I'm your most loyal fan but even I'm starting to check out alternatives.
As long as there are idiots like you who say "performance will not be a problem since even embedded platforms will run on multi-core chips" there will be slow, bloated interfaces. Not caring about performance because we have cheap fast hardware doesn't mean you'll get decent performance. It just means you'll suck more juice to get an approximation of what the hardware can actually do.
I'm with you here. I really, really want to love KDE. I've been with it since the early 3.x if not earlier (hard to remember). Gnome's got everything ass-backward and, to coin a phrase, idiocized. I feel you have to think like an idiot to "get" Gnome. KDE 3.5 was VERY well integrated, I loved dcop. Shit just worked.
I too am slowly moving off of KDE technologies. I hate to do it, but I'm tired of the bugs. Konqueror was really good, both as a web browser and a file manager. Dolphin is awful. Rekonq feels broken whenever I try to use it. I've been using Chrome for the last year. Kontact is an app I have married to for a long, long time. When the basket thing came out I really got into it. Now however kontact can't deal with an 11k-mail maildir without choking most of the time. viewing a new message takes sometimes 5-10s to actually show up (and it's not an HTML email, either). Akgregator can't handle large databases. Calendaring is so brittle that it has a 70% chance of crashing after editing any entry in my remote (caldav) calendar. Accepting a meeting from email is guaranteed to crash kontact within 10 minutes. I've actually got Evolution running right now and am trying it out. Thunderbird is like Firefox: bloated, heavy and slow. I guess I'll file a feature request for Evolution to allow me to edit the From: line (a feature I've used for close to a decade now for mailing lists and any kind of registration system so I can track where my spam comes from).
Plasma-desktop was probably one of the bigger mistakes KDE brought in. I admire what it is capable of doing, but at the cost of stability of the ENTIRE DESKTOP? sorry, that was a bad choice.
Konsole, Kate and Kopete are my mainstays, but Kopete will be the next to go. Whose idiot idea was it to store Kopete history in XML files? I keep history of all my IM just like I do with email and some private newsfeeds. No seamless, searchable archive support means that these things *chug*.
I really, really want to keep using KDE. I love the integration and I love the way it's SUPPOSED to work. KDE 4.x was a real let down, even in 4.6.
Actually it's not the PHY that's special for IEEE1588, it's the MAC. It has "fast path" hardware which can accurately timestamp/send out IEEE1588 frames.
The point is, 9x% of users use X locally now. Focus on that, and when that is good, then you can turn to network transparency. And you know what? That would allow them to focus the attention on network transparency it deserves, rather than shoehorning it in as an all of nothing proposition.
The problem with that is that when you decide to "turn on" network transparency you find it breaks everything because people haven't written Wayland nor the apps running Wayland to work properly with a network. Then you end up with disgusting hacks like RDP and VNC. I've used both and neither can hold a candle to "ssh -CX appname" let alone something a hundred times better like NoMachine NX.
No, network transparency has to be part of the system from the very get-go, or just don't bother mentioning it. You can't build in a feature like that at a later date.
I am in love with NoMachine NX; it's remote X11 done right. The free version is more than sufficient for my needs, and there's FreeNX too if you care about running OSS-only.
Let's see... korganizer reminder daemon generating a gazillion popup errors when the network is disconnected and it can't get to a remote calendar? kopete's history window being terribly slow and clunky? general input window slowness with large blocks of text (kmail generally, but other apps do it too)? general fucktardery with kontact's calendaring (timezone issues, slowness/laggy response, crashing)? akgregator becoming horribly, horribly slow when left idle "too long"? amarok being a steaming pile of shit since 4.x? the kde power daemon profiles not actually doing what they're set to do? plasma-desktop crashing out on a regular basis?
Is that enough to get you started? I've been using KDE since before 3.x and let's be honest: 4.x, even 4.7, is still an obese, steaming pile of bug-ridden feces. I try to be diligent and file the bug reports but very little seems to get done outside of the traditional games of hot potato between the respective Ubuntu and KDE teams.
I've been running KDE since the 3.x days, stuck through the early 4.x, and am currently running the latest and greatest that comes with Kubuntu 11.04.
You've got your head up your arse if you think KDE4 is even close to flawless. It is *still* a HUGE memory hog. It is *still* a HUGE CPU hog. It's leaps and bounds better than Gnome 3 and now Gnome 4, but it's still very big, slow and unstable compared to the 3.5 days. Almost all the old Enlightenment jokes now apply to KDE, except the hardware is a bajillion times faster.
What's funny is that I've never had a problem with how KDE 4's taskbar works. I just have issues with its stability. Plasma crashes far too often for a "point 7" release. Maybe you guys should focus on stability and trimming the fat rather than racing Gnome and seeing how many flashy glossy things you can add.
So what happens when you have to write a long email, and you're not in your car?
I wait, just like I would if the battery was dead or I had no cell signal or any of the other reasons someone would have to wait to send a long email. Tapping out a long email even on a physical but tiny keyboard is still tedious, albeit marginally less so than on an onscreen keyboard.
Anyway, you should understand that the need for physical keyboard depends on one's usage scenarios. For most people, one is not, generally speaking, a requirement, and then they either do what you do, or - the majority - do not bother with BT at all. But there is a reason why e.g. all Blackberries used to have one (and all business ones still do). For people who do use the device for email and other written communication a lot, having an integrated keyboard is indispensable.
Yes, you're absolutely right, it depends on use cases. However I think it would be hard to argue that 70% of those who have blackberries (or smartphones to a larger extent) actually have a real need for the device. They're status symbols to a large extent, much like "I need an SUV.":-)
You know, I've never given a shit how many cores there are on my phone. I want it to be reasonably responsive and if it needs one or 12 cores to do this, I don't care. I don't need or even WANT a zillion choices. I want things that work well, and that's why even today, I will choose an iPhone 4 over anything Android. I've seen them and I've used them and Android is exactly like Linux on the desktop. Full of choices, but nothing's quite finished. I'm fine with that on my desktop, but not on my phone.
I disagree very strongly with your assertion that an on-device keyboard is vastly superior to an external one. The number of times I actually have to use a keyboard with my iOS device is low enough that I am happy to keep the BT keyboard in the car and not have the extra weight/bulk on me all the time. When I have to ssh in somewhere and it's for more than a dozen commands, or when I want to have a real conversation on IM/IRC, or I have to write a long email, I bust out the BT keyboard. For everything else the onscreen one is sufficient, and I don't have any slider/hinge mechanisms to break or wear out.
I've used gogo's internet several times. It's exactly the same speed as a modern CDMA (Verizon) mobile internet connection. No drops, consistently 50ms latency, no apparent port blocking, although forced web proxy. I really don't know what kind of shit hardware you're using, because my 4yo macbook with original battery still gets 4h of battery life doing the kinds of things you'll be using internet for on a flight (i.e. not compiling or gaming). Hell, my 4yo Toshiba U300 with a new battery running Ubuntu gets between 2 and 3h doing the same thing.
I like their service, and $10 really isn't a bad price to pay if you want to chat on IRC or talk to your friends on MSN on an otherwise boring flight. It's about the same price you'd pay for a frou-frou coffee to do the same thing in a coffee shop. Sure, I'd prefer free, but let's be realistic here. All this "need internet badly" and "attention problems" bullshit comments I see here are missing the point entirely. I get by just fine without internet, but sometimes it's just nice to virtually hang out with your friends when you're flying over them.
I call bullshit. I have long preferred laser but my 3 year old $200 Canon Pixma MP530 not only does duplexing and has ADF for scanning but the ink tanks are refillable. I put new tanks in it maybe once every 3-4 months, at a total cost of $45 for all 5 tanks. A colour laser that did duplexing would have cost me a good deal more than $400 today, let alone 3 years ago, and I'd still have to buy a scanner with ADF for what, $50 used off of ebay.
You can tell by my use (new tanks 3-4 times a year) that I don't print a whole lot, yet the heads have never needed cleaning. I just wish it did hole punching and binding.:-)
Nonsense. It's a rule that you are obligated to abide by when working for the company. "Don't piss in the boardroom" is another rule. Are you suggesting that breaking that rule shouldn't be a firing offense, or are things somehow different because it's an IT policy?
I bought a little APC battery about 5 or 6 years ago; it's got a USB and microUSB port on it. You plug it into a USB port on a computer or charger to charge it, and you plug your phone into it to charge/power the phone. Works great, stores way more than any secondary phone battery I've ever used, can be charged without a proprietary connector and works for more than one device. Also works great for watching movies on long-haul flights without power sockets.
Seems a lot better solution to me than carrying around a second phone battery.
Sure, with all the hidden node problems and access problems a busy wifi network gives as well. Standard 802.11abg wifi is not suited for this kind of application. You really want to use TDMA, not CSMA/CD.
Re:What functionality are we BSD users ...
on
Xfce 4.8 Released
·
· Score: 1
You're seriously timing even a hundred file opens and closes when you're talking about ANY operation centered around a human clicking a button or using a keyboard? You're a fool. Registries are hideous single points of failure with no redeeming value. If you're doing some high speed repetitive operation and need to access data you MIGHT have an argument, but then again the VFS takes care of keeping that data in the filesystem cache already and you're not going to be hitting disk more than once.
Honestly... Do you optimize your code as you write it too?
Re:Making it just as heavy as Gnome and KDE now?
on
Xfce 4.8 Released
·
· Score: 1
I have no problems with Dropbox in KDE (Kubuntu). I did install the nautilis plugin, but my Dropbox folder works, the systray icon works... *shrug*
Slashdot Mirror
Your argument seems baseless and has nothing to do with the text of the person you're attacking. Lucractius said that your artist daughter should work harder, perfect her craft and market the hell out of herself. How is this a "typical engineer's attitude towards art" ? That's a sane attitude to have in a world where anything can be made cheaper and faster elsewhere. Do you honestly believe that an artist has some magical income forever job? Most artists I know don't live very high.
You clearly have no idea what you're talking about. UBB was never about paying for what you use. It was a way for Bell/Rogers to force the wholesalers into offering packages that were no better than what Bell/Rogers were providing. The incumbents were using their monopoly on the infrastructure for the last mile to try and "level" the playing field of the competitive carriers.
I have absolutely no problem with paying for what I use. The problem is that the pricing they're trying to force is not reasonable. It doesn't cost even $0.10 for 1MB delivered in a second, yet they were trying to charge well over that per KILObit. I kind of wish that the CRTC gave them exactly what they asked for. I wish the CRTC gave them usage based billing, but enforced pricing based on legitimate data costs plus a legitimate profit. If they wanted to make more money than that, they'd have to find a way to cut costs without cutting service.
Connectivity in Canada is terribly, terribly out of touch with reality. There are a few monopolies who still own a huge chunk of the last mile, and short of cutting all ties between the last-mile provider and the bandwidth/content provider, nothing will fix this.
I spent 15 years in industrial power, working with the exact communications and control networks you complain about. I know what you're talking about and sympathize to a large extent.
Your boss' request, however, does not compromize security. There is nothing insecure about providing read-only access to the network. A secure device sitting on the industrial network could export/expose the system status through a variety of means, none of which would accept requests to alter the system state. At a very extreme case, a unidirectional optical link could stream system state from the critical side of the device to the noncritical one, and the insecure side could then pick out what it wants to report on. There's enough bandwidth available in gigabit ethernet and the filtering problem is largely solved already (and has been for well over a decade with CAN), which is where DeviceNet and Ethernet/IP get their roots from.
It'd be an uphill battle, but I'd like to design a PLC module that works exactly this way. I miss the industrial work. :-)
You know, I used to do what you are doing... I hated the idea of getting rid of old hardware. It's perfectly functional and let's face it, the computational power required to run a printer or file server or even a light mail server is very, very low.
What changed my opinion on this was coming to the realization that the power consumption of these older devices is not small. That old P90 running idle can't compare to a modern core2 or iwhatever system. Atom processors are useless; their idle power is slightly lower than a core2, but any time you have to make it do something it ends up taking more power to do it than it takes the core2 to spin up, do the task and get back to idle.
It still bothers me a bit when I decide to recycle an otherwise functional piece of old equipment, but in the mid/long run I'm actually saving money by consolidating equipment (mail+print+fileserver instead of 3 small servers) and making conscious decisions about the power consumption of the parts in the servers.
You seem to think nobody is even considering it. That's certainly not true.
I'm just really concerned that the decision to "design in network transparency" will be what is said, but when everyone's salivating over how cool and fast and slick it's going to be that design decision will be marginalized. I'd love to get some kind of involvement in Wayland but I know that I haven't even the slightest amount of time to be useful in such a project.
I guess that's why I'm posting really concerned comments on a random web forum. :-/
Oh I see. my mistake for not even clicking the links. Sorry.
I installed the 4.7 ppa for Ubuntu 11.04 last night. You are right, 4.7 cleaned up a lot of little things that were bugging me in 4.6. I have not moved to the 4.7 kdepim yet (that's in the experimental ppa) but 4.7 so far offers considerable relief from 4.6 (at least on Ubuntu).
I want to thank you for managing all the comments in this story. I really, really do love KDE and I want to use it for a good long time to come. I think it's obvious from the comments that there are a LOT of other people like me who want to see KDE succeed but we're very concerned over the way development seems to be more concerned over bells and whistles instead of speed and stability.
um... kmail/kontact has had this feature for a LONG time... I was complaining that I am going to miss it if I really do have to move off of kontact for my email/calendaring app.
The devs repeatedly said it wasn't for everyday use for everyone, and that it was mainly for developers to have a base to build from.
As both a developer and a picky user, perhaps the KDE team should NOT have called it a 4.0 release. Perhaps a 4.0-RTD or something (release to developers) -- make it obvious from the name that it's not for users. I know that KDE tried to do that with all the warnings and everything, but let's be honest... most users are stupid. Kubuntu made a hideous mistake by including it. Ubuntu's got this problem at a systemic level (ripping out perfectly good code/apps to try to establish new paradigms before they're ready to be used) so I guess it was inevitable, but perhaps if KDE 4.0 were named KDE-4DEVRELEASE or something...
I don't know. I hope that what happened with 4.x is never again repeated. 4.7 is almost out and it's still not where it was in 3.5. Lord knows software development ain't easy, and trying to write a desktop environment for a large and varied userbase is difficult to put it mildly, but I don't think KDE really helped themselves at all with 4.x. I'm your most loyal fan but even I'm starting to check out alternatives.
As long as there are idiots like you who say "performance will not be a problem since even embedded platforms will run on multi-core chips" there will be slow, bloated interfaces. Not caring about performance because we have cheap fast hardware doesn't mean you'll get decent performance. It just means you'll suck more juice to get an approximation of what the hardware can actually do.
I'm with you here. I really, really want to love KDE. I've been with it since the early 3.x if not earlier (hard to remember). Gnome's got everything ass-backward and, to coin a phrase, idiocized. I feel you have to think like an idiot to "get" Gnome. KDE 3.5 was VERY well integrated, I loved dcop. Shit just worked.
I too am slowly moving off of KDE technologies. I hate to do it, but I'm tired of the bugs. Konqueror was really good, both as a web browser and a file manager. Dolphin is awful. Rekonq feels broken whenever I try to use it. I've been using Chrome for the last year. Kontact is an app I have married to for a long, long time. When the basket thing came out I really got into it. Now however kontact can't deal with an 11k-mail maildir without choking most of the time. viewing a new message takes sometimes 5-10s to actually show up (and it's not an HTML email, either). Akgregator can't handle large databases. Calendaring is so brittle that it has a 70% chance of crashing after editing any entry in my remote (caldav) calendar. Accepting a meeting from email is guaranteed to crash kontact within 10 minutes. I've actually got Evolution running right now and am trying it out. Thunderbird is like Firefox: bloated, heavy and slow. I guess I'll file a feature request for Evolution to allow me to edit the From: line (a feature I've used for close to a decade now for mailing lists and any kind of registration system so I can track where my spam comes from).
Plasma-desktop was probably one of the bigger mistakes KDE brought in. I admire what it is capable of doing, but at the cost of stability of the ENTIRE DESKTOP? sorry, that was a bad choice.
Konsole, Kate and Kopete are my mainstays, but Kopete will be the next to go. Whose idiot idea was it to store Kopete history in XML files? I keep history of all my IM just like I do with email and some private newsfeeds. No seamless, searchable archive support means that these things *chug*.
I really, really want to keep using KDE. I love the integration and I love the way it's SUPPOSED to work. KDE 4.x was a real let down, even in 4.6.
Actually it's not the PHY that's special for IEEE1588, it's the MAC. It has "fast path" hardware which can accurately timestamp/send out IEEE1588 frames.
The point is, 9x% of users use X locally now. Focus on that, and when that is good, then you can turn to network transparency. And you know what? That would allow them to focus the attention on network transparency it deserves, rather than shoehorning it in as an all of nothing proposition.
The problem with that is that when you decide to "turn on" network transparency you find it breaks everything because people haven't written Wayland nor the apps running Wayland to work properly with a network. Then you end up with disgusting hacks like RDP and VNC. I've used both and neither can hold a candle to "ssh -CX appname" let alone something a hundred times better like NoMachine NX.
No, network transparency has to be part of the system from the very get-go, or just don't bother mentioning it. You can't build in a feature like that at a later date.
I am in love with NoMachine NX; it's remote X11 done right. The free version is more than sufficient for my needs, and there's FreeNX too if you care about running OSS-only.
Let's see... korganizer reminder daemon generating a gazillion popup errors when the network is disconnected and it can't get to a remote calendar? kopete's history window being terribly slow and clunky? general input window slowness with large blocks of text (kmail generally, but other apps do it too)? general fucktardery with kontact's calendaring (timezone issues, slowness/laggy response, crashing)? akgregator becoming horribly, horribly slow when left idle "too long"? amarok being a steaming pile of shit since 4.x? the kde power daemon profiles not actually doing what they're set to do? plasma-desktop crashing out on a regular basis?
Is that enough to get you started? I've been using KDE since before 3.x and let's be honest: 4.x, even 4.7, is still an obese, steaming pile of bug-ridden feces. I try to be diligent and file the bug reports but very little seems to get done outside of the traditional games of hot potato between the respective Ubuntu and KDE teams.
I've been running KDE since the 3.x days, stuck through the early 4.x, and am currently running the latest and greatest that comes with Kubuntu 11.04.
You've got your head up your arse if you think KDE4 is even close to flawless. It is *still* a HUGE memory hog. It is *still* a HUGE CPU hog. It's leaps and bounds better than Gnome 3 and now Gnome 4, but it's still very big, slow and unstable compared to the 3.5 days. Almost all the old Enlightenment jokes now apply to KDE, except the hardware is a bajillion times faster.
What's funny is that I've never had a problem with how KDE 4's taskbar works. I just have issues with its stability. Plasma crashes far too often for a "point 7" release. Maybe you guys should focus on stability and trimming the fat rather than racing Gnome and seeing how many flashy glossy things you can add.
So what happens when you have to write a long email, and you're not in your car?
I wait, just like I would if the battery was dead or I had no cell signal or any of the other reasons someone would have to wait to send a long email. Tapping out a long email even on a physical but tiny keyboard is still tedious, albeit marginally less so than on an onscreen keyboard.
Anyway, you should understand that the need for physical keyboard depends on one's usage scenarios. For most people, one is not, generally speaking, a requirement, and then they either do what you do, or - the majority - do not bother with BT at all. But there is a reason why e.g. all Blackberries used to have one (and all business ones still do). For people who do use the device for email and other written communication a lot, having an integrated keyboard is indispensable.
Yes, you're absolutely right, it depends on use cases. However I think it would be hard to argue that 70% of those who have blackberries (or smartphones to a larger extent) actually have a real need for the device. They're status symbols to a large extent, much like "I need an SUV." :-)
You know, I've never given a shit how many cores there are on my phone. I want it to be reasonably responsive and if it needs one or 12 cores to do this, I don't care. I don't need or even WANT a zillion choices. I want things that work well, and that's why even today, I will choose an iPhone 4 over anything Android. I've seen them and I've used them and Android is exactly like Linux on the desktop. Full of choices, but nothing's quite finished. I'm fine with that on my desktop, but not on my phone.
I disagree very strongly with your assertion that an on-device keyboard is vastly superior to an external one. The number of times I actually have to use a keyboard with my iOS device is low enough that I am happy to keep the BT keyboard in the car and not have the extra weight/bulk on me all the time. When I have to ssh in somewhere and it's for more than a dozen commands, or when I want to have a real conversation on IM/IRC, or I have to write a long email, I bust out the BT keyboard. For everything else the onscreen one is sufficient, and I don't have any slider/hinge mechanisms to break or wear out.
I've used gogo's internet several times. It's exactly the same speed as a modern CDMA (Verizon) mobile internet connection. No drops, consistently 50ms latency, no apparent port blocking, although forced web proxy. I really don't know what kind of shit hardware you're using, because my 4yo macbook with original battery still gets 4h of battery life doing the kinds of things you'll be using internet for on a flight (i.e. not compiling or gaming). Hell, my 4yo Toshiba U300 with a new battery running Ubuntu gets between 2 and 3h doing the same thing.
I like their service, and $10 really isn't a bad price to pay if you want to chat on IRC or talk to your friends on MSN on an otherwise boring flight. It's about the same price you'd pay for a frou-frou coffee to do the same thing in a coffee shop. Sure, I'd prefer free, but let's be realistic here. All this "need internet badly" and "attention problems" bullshit comments I see here are missing the point entirely. I get by just fine without internet, but sometimes it's just nice to virtually hang out with your friends when you're flying over them.
I call bullshit. I have long preferred laser but my 3 year old $200 Canon Pixma MP530 not only does duplexing and has ADF for scanning but the ink tanks are refillable. I put new tanks in it maybe once every 3-4 months, at a total cost of $45 for all 5 tanks. A colour laser that did duplexing would have cost me a good deal more than $400 today, let alone 3 years ago, and I'd still have to buy a scanner with ADF for what, $50 used off of ebay.
You can tell by my use (new tanks 3-4 times a year) that I don't print a whole lot, yet the heads have never needed cleaning. I just wish it did hole punching and binding. :-)
Nonsense. It's a rule that you are obligated to abide by when working for the company. "Don't piss in the boardroom" is another rule. Are you suggesting that breaking that rule shouldn't be a firing offense, or are things somehow different because it's an IT policy?
I bought a little APC battery about 5 or 6 years ago; it's got a USB and microUSB port on it. You plug it into a USB port on a computer or charger to charge it, and you plug your phone into it to charge/power the phone. Works great, stores way more than any secondary phone battery I've ever used, can be charged without a proprietary connector and works for more than one device. Also works great for watching movies on long-haul flights without power sockets.
Seems a lot better solution to me than carrying around a second phone battery.
Sure, with all the hidden node problems and access problems a busy wifi network gives as well. Standard 802.11abg wifi is not suited for this kind of application. You really want to use TDMA, not CSMA/CD.
You're seriously timing even a hundred file opens and closes when you're talking about ANY operation centered around a human clicking a button or using a keyboard? You're a fool. Registries are hideous single points of failure with no redeeming value. If you're doing some high speed repetitive operation and need to access data you MIGHT have an argument, but then again the VFS takes care of keeping that data in the filesystem cache already and you're not going to be hitting disk more than once.
Honestly... Do you optimize your code as you write it too?
I have no problems with Dropbox in KDE (Kubuntu). I did install the nautilis plugin, but my Dropbox folder works, the systray icon works... *shrug*