Why am I being offered Gnome and KDE? What's the difference?
Or you installed Ubuntu, and that question is resolved, until you decide you want to dig deeper.
Why can I not play DVDs?
Because you didn't buy it preinstalled from a place like Dell, which handles that for you.
Try installing a fresh copy of Windows, and you'll find that it cannot play DVDs either, unless you happened to get a copy of PowerDVD with that DVD drive. You'd have to download VLC anyway.
What's the app called for playing MP3s and ripping CDs? Alright everyone, which one of the dozen is going to work best?
Probably the one that's preinstalled.
How would you answer this on Windows? iTunes? There are dozens of others you could've downloaded. Windows Media Player? Sucks, but it's preinstalled. Those are the choices.
Oh, and if you have to download one: Amarok for playback, k3b for ripping -- that's what came by default on my Kubuntu. I'm not sure what comes by default on Ubuntu, but chances are, it will have a description right next to it in the menu.
Contrast with Windows: Where are my drivers?
Ok, let's see. What kind of video card do I have? Let's look up the manufacturer's website, looks like it's nvidia, ok, head over to nvidia.com. WTF? Where's the download? Nvidia tells me I have to go to my manufacturer's website. OK, back to Dell.com. WTF? Vista-only? This is bullshit, nvidia certainly maintains XP drivers, but due to a contract with Dell, can't provide them.
Ok, contact support. Support gives me a download for a different model of laptop that happens to have the same video card. Support informs me that there's no way I could know or discover this on my own -- I would just have to contact support.
And that's just the video card. The best wireless drivers were actually not the default ones, as Intel ProSet is utter crap compared to the default Windows wireless tools. I base this mostly on the fact that ProSet is completely unnecessary -- maybe there is something it does better, but I haven't found it yet. Regardless, I had to find the "driver only" download.
And on down the list -- AHCI is really needed to make the best use of this hard disk, but XP doesn't support it without a driver. One option was to disable it in the BIOS and re-enable it later. The option I chose was to slipstream the driver into the XP CD -- a process significantly more involved than burning an Ubuntu ISO.
Why is it running so slow? Why does it take 15 minutes to shutdown?
I noticed you didn't list a solution to that. That's because the solution is pretty much what you said for one Linux problem:
Okay, reinstall. Why did I not make notes?
The solution to each, of course, is to back up a working copy, or to take notes, possibly in the form of a script.
The difference is, I actually know how to fix a Linux problem. The fix for a Windows problem is generally to reboot, and if that doesn't work, reinstall. This may be my own lack of knowledge, but others seem to notice the same thing, and certainly Windows professionals seem to have little to suggest beyond running Spybot and reinstalling. So maybe it's related to the "freedom to tweak"?
And on "enterprise" distros, where versions are locked for years while patches are backported, you suffer dependency hell, which forces you to run older versions of software until you can make the business case to upgrade the OS and all the commercial apps.
While having everything be of a constant version (and thus designed to work well together) is useful, upgrading doesn't generally break everything. Especially when your only compatibility problems will be with that old software, it should be easy enough to test the upgrade and make sure it still works.
This is more or less the same as the case with Windows -- notice how businesses are still running XP, which has had things slowly patched and backported. Most ignored Vista, and most are now looking to skip Windows 7.
Ok, but I'm not a driveling whiny developer enthusiast that needs to have the bazillion levels of freedom that you need to hack the bejeezus out of your computer. I'm a burger flipper, a tire guy, a mechanic, a professional, or a housewife and I just want the stuff to work.
In that case, the bazillion programs available for Windows shouldn't matter -- only the few you need to work. Additionally, the lack of a need for antivirus, and the ease of keeping your system up-to-date and secure, should appeal to you.
In fact, even a package manager and a distribution should benefit you, in the long run. Choosing software supported by the distro means it'll be maintained, likely forever and for free. Using a distro like Debian or Ubuntu, which has separate stable and unstable versions, means that as long as you're on the stable version, all of that software is known to work together -- no "dll hell", no other strange cases of one piece of software causing another to not work.
I don't want to have to make a stupid decision about which distribution I should download
That's why we say "Ubuntu" and move on.
and I don't want to have to answer nine billion technical questions just to get it installed.
I'm sure someone can verify it, but I don't think Ubuntu asks more questions than XP. If you're a professional, you solve this problem by getting it preinstalled.
I want to have that feeling that there is a company that I can blame,
That would be Dell, who is providing you service, if you followed the above option.
I need to have the feeling that there is a group of people that may benefit from my purchase,
That, I really don't get. Since it can be free, why would you need that? If you get it as a product, with someone to blame (the Dell option), then Dell and Canonical both benefit, and some portion of your money goes directly to improving Ubuntu.
Ubuntu is coming the closest to being a product and it's goals are commendable but it is not a mainstream desktop PRODUCT and never will be.
Why not? Putting PRODUCT in all caps (and bold) doesn't make it a valid point. Your actual points here, I think I've refuted.
Just because it's free doesn't mean it is going to be good and just because it costs a lot of money doesn't mean that it's evil.
This is true. However, the fact that it is free, in a truly level market economy, would mean that anything that costs money would have to come with a lot of added value.
As it is, the closest competitor, in the sense of something for which most software is compatible, might be Solaris (and other commercial Unices), but Solaris was recently open sourced -- Linux dominates that market. OS X might count, except their GUI is so proprietary that a truly native OS X app can't be much more easily ported to Linux than a Windows app can.
I'l probably get modded a 0 flame bait for this
I really hope mods stop falling for this tactic.
Hey, mods, I'm about to say something that people might not want to hear! Some people might mod me down for it! You'd better mod me up to compensate!
I'd have modded you overrated, but I actually have something to say.
That, or if you're partnered closely with a designer.
I worked for a web company, and we did use a designer. We'd give them a rough idea of what we wanted, they'd send us back HTML and CSS, and we'd just have to replace Lorem Ipsum with template markup, then go play with backend logic.
The company we hired was basically one guy who was 100% artistic, and one guy who turned it into HTML/CSS and appropriate images.
Eventually, we started doing the design in-house, mostly because we decided that we could do it simpler and better than a web designer -- mostly because we wanted something absurdly minimalist. But it did take a lot of time, even with that "minimalist" route.
As a developer, what was probably the most frustrating was either having to design things myself (and having them look ugly), or being given a gigantic photoshop file and having to slice out the parts that should be images, measure pixel distances between everything, and deal with all the warts of CSS in Internet Explorer. But, it was better than having to deal with something DreamWeaver (or worse, MS Word or Adobe GoLive) crapped out.
So yes, there is definitely a job in it, and it's a job I don't want to do.
Each broadband customer must have one IP anyway, so it does not matter if they give you a fixed one or not.
Yes and no.
Yes, they are gouging. All ISPs seem to, with this.
No, it's not zero cost. Assume they pull it from the DHCP pool -- that means reconfiguring the DHCP server, and possibly killing a lease (thus interrupting service for that person). There are likely graceful ways to do it, but they're not exactly that easy.
But then, you'd think they would want to encourage static IP addresses. It might not be much, but DHCP does require a server, some amount of bandwidth, CPU, and disk space, and is yet another point of failure should something go horribly wrong.
Seriously, just use dyndns, dnsmadeeasy, zoneedit, or any of the many, many free services out there for dynamic DNS.
your assertion that it only costs $0.16 is stupid. First, they need to also pay electricity, pay for employees to take your service calls, employees to do line repairs, repair existing lines when they are damaged, the gas to get those trucks out to repair the line, the lease on the truck itself, and then still have enough for upgrades and to make a profit...
Amazon has to do all of this, and manages to charge $0.17 per gig down, $0.10 per gig up. Flat, across all their services.
Or I suppose AWS doesn't run on electricity, nor require any kind of service or customer support?
Alright, yes, an ISP has a bit more ground to cover, physically, than a datacenter. On the other hand, Amazon is either plugged directly into a backbone, or they have an ISP too.
But really, are those physical lines from my ISP to my house really costing enough to explain that it costs ten times as much?
Worth mentioning, that's initially a win for the service provider -- the slides are packed, but each kid is still paying $20. Even if you assume the waterpark was fully utilized at $20 for 10 slides, that means they have at least twice as many kids getting 5 slides, but still paying $20 each.
What I've humbly suggested is, instead of pocketing the extra money, spend it on building more slides. It may take time, but you'll eventually get even more kids paying $20 each, and having a lot more fun.
Instead, ISPs have done the equivalent of advertising even more, and pulling in even more people, until no one can get more than about 2 slides. Some kids figure out exactly which slides to use, and when to use them, coordinate with each other to plan it out, and somehow manage to get 20 slides each -- thus forcing everyone else down to 1 slide. The pool cracks down by randomly kicking people out who "slide too much" without defining it.
Eventually, after enough people complain, the park goes back to its original model, this time stating explicitly that you get 5 slides, and more than that costs extra.
All of this just to make a few bucks in the short term.
What shocks me is that there's no competing waterpark with five times as many slides to pull all the customers away. Surely, at least some ISPs have to have gotten it right, and invested in infrastructure, right?
Oh, and it was a good analogy, sorry to abuse it...
That's why you need tiered cap pricing if you go to that - for some people 5GB is quite enough, others would use it in a day. Tiered pricing allows people to decide for themselves what capacity they need...
And metered means that if those usage patterns change, they can still pay a reasonable amount, rather than suddenly being bumped up to a higher tier. Given the Steam example, what happens if I never really used much bandwidth, but one day I downloaded Steam? Do I get a pile of overage charges and the offer to add a flat $20/mo to my Internet bill? to cover this sort of usage? Or do I get a few little spikes of $1 or $2 here and there, if that?
VOD, for example, would probably get less uptake as people compare that cost to renting or watching a movie on pay per view.
I assume that's "video on demand", right?
Alright, let's compare. Amazon charges 17 cents per gig for downloading from Amazon Web Services. That gets pricey if you're downloading an absolutely pristine, full Blu-Ray disc -- which you're almost certainly not. A very good quality Blu-Ray rip will fit into 10 gigs, or $1.70 -- still cheaper than the rental, and you don't necessarily have to return it. And most services compress it a hell of a lot more.
If it's metered and still too expensive, sure. If it's metered at a reasonable rate, I don't think it becomes a problem.
Many content owners are waking up to the idea they can cut distribution costs significantly by internet based delivery; thereby making more money by cutting out the middle man and essentially passing the cost of the tangible media to the buyer eliminating all the reproduction, storage and distribution costs; as well as be freed from battles over format (Blu-ray vs HD and whatever comes next.)
The vast majority... well... where do you think Blu-Ray (or HD-DVD) came from in the first place?
Granted, it was probably content creation corporations, not individual creators, but the fact is, they're often the first to shoot themselves in the foot over a piracy scare. Let's not forget the whole "VCR is the Boston Strangler" analogy. The smart ones picked up on Internet distribution a decade ago -- the dumb ones think "Internet Distribution" means getting their music on iTunes -- or worse, PlaysForSure/Zune.
one of the biggest problems: 2 access to get that URL.
How is that a problem?
The main reason it's a problem with a service like tinyurl is that one of those is to a completely different server. This means two points of failure for the request (TinyURL could be down), multiplied by all the stages involved (The TinyURL DNS could be down...)
And of course, the larger problem of a single point of failure -- TinyURL going away breaks all those links. TinyURL also gets data on all those links.
Let's face it -- multiple connections will be opened per page anyway, for ads alone, let alone scripts, CSS, images, etc -- unless they're pipelined via HTTP 1.1. Either way, an extra HTTP GET is just not a big deal these days -- especially when it pretty much only has to return this:
HTTP/1.1 301 Moved Permanently Location: http://foo.com/some/longer/url
Even when you consider the three-way handshake (that is, assuming it doesn't hold the connection open as a pipeline), that's just not a lot of data.
I suppose it matters if you're on an absurdly high-latency connection. Other than that, I see cache coherency and resource integrity as much higher priorities.
The main reason not to is identity. A resource should generally be accessible from a single URL. That's what the redirect is for -- to identify where the canonical resource lives.
Given that I've also never once had a kernel upgrade or OS upgrade of any kind completely trash my filesystem, or hard drive, or controller, etc.
Having a backup of my data and configuration is useful. Backing up several gigs of software really isn't -- worst case, I lose a few hours re-downloading it.
we really had no good way to guide students effectively in learning the interface.
Which makes me curious, again -- how much would they figure out on their own?
Consider that children often figure out the "child-proof" medicine bottles more quickly than adults. Or at least, my parents would often give them to me to figure out.
asking how she could make use of it got nowhere because she would not be allowed to make any use of it.
I assert that this really isn't Sugar's fault. If she's actually being dictated which UIs she can teach and which she can't, I doubt xfce or anything else would've been better, unless it was the XP that eventually got grafted onto it.
The educational bureaucracy in most countries is mostly familiar with Windows on personal and institutional bases, so they will be much harder to interest in adopting computing paradigms (such as Sugar) that seem irrelevant to what they perceive as mainstream, i.e. Windows.
Which was exactly why OLPC wasn't targeting existing beaurocracies so much as developing countries -- but rather targeting the kind of community where it's likely even the teachers haven't been exposed to computers much before.
Of course...
And you can be sure M$ was leaning on them to continue to see it that way.
That's the real tragedy here. Because I believe there was an opportunity to introduce something genuinely new to a large portion of the world.
Instead, Microsoft wins again.
But no, I don't think differentness played a large role there. After all, the common challenge for Linux has always been, if it is too different, people used to Windows won't be able to follow that learning curve. If it's too much the same, people won't see a compelling reason to use it over the alternative.
Let me ask you: If it looked and felt just like XP, only wouldn't run Windows programs, and was only a few dollars cheaper, would anyone be interested? Only in places where a few dollars really matters -- and even there, the perception would be that this is the cheaper version, and that you get what you pay for -- not that this is the revolutionary new vision, that also happens to be cheaper.
I'd say what's good about multiplayer games is that you're playing against human opponents. A bot doesn't get angry when it loses, or gloat when it wins. There's a certain amount of creativity lacking in a bot -- and humans tend to have entirely different and unpredictable flaws and strengths.
I've played multiplayer games in which particular humans had a severe handicap, and it was still fun.
No, the main thing I don't like about special treatment for the player is that it's unrealistic, and it's a cop-out -- plenty of games manage to make it work without that. Of course, that's not always a constant -- if the enemies are in a ten-story-tall tank (like the Scarab in Halo), it absolutely should be hard to kill. Even Max Payne had a decent explanation -- the enemies didn't stop with one shot because they were all ridiculously high. But in a GTA game, it really makes no sense.
Mostly because I'm already on 8.10, and I don't like being a full year out of date.
Some of that is actually legitimate -- for instance, various development tools and games would be falling out of date. Some of it's just impulsive -- I occasionally regret updating to 8.10, as KDE4 was really not ready.
My most salient advice is to make a bootable, external-disk full system backup before attempting an Ubuntu upgrade.
I'm far too lazy for that, so I tend to just back everything up to a folder -- even a folder on the same system. It's much easier to downgrade if the entire thing falls apart, but I've never had an Ubuntu upgrade completely kill my filesystem or hard drive.
what is the chance that software compiled for Bolivian KDE will run under Paraguayan KDE? (Hint: WINE)
What is the relevance of that at all? If KDE apps can run under GNOME, and KDE3 apps can run under KDE4 (and vice versa), I seriously doubt Paraguayan KDE will have a problem with Bolivian KDE.
If they really were fully targeting the Win32 API their goals would logically be measured by percent of API implemented rather than based on which programs run under the current Win32 subset.
That would be true, if the Win32 API stuck to any sort of standard.
It doesn't. Like many Microsoft APIs, the spec is spotty and incomplete, and since it's so widely deployed in exactly one version, the only "correct" spec (as far as anyone cares) is "how Windows XP does it."
No, Wine's goal has always been "bug for bug" compatibility with Windows. And you can't measure that in percentage of an API with such buggy and incomplete documentation. You can only measure it in actual programs that will run.
That's ignoring, of course, the really strange things Microsoft has had to do -- certain old software will make really stupid assumptions (like the length of the Windows version number), and modern Windows will detect those programs and change the API accordingly (for example, lying about the Windows version so the app doesn't crash).
I don't know if Wine intends to run such programs, but if so, implementing Win32 isn't even close to enough.
Sure, greater productivity is one benefit, but the language is completely irrelevant for that.
It's about how flexible the system will be when you have to change it. And you will -- that's the whole point of software, that it is soft, and changeable.
Old Cobol apps generally are not flexible. (stolen from this comment). It's worth mentioning that a decent object-oriented system would've gone a long way towards eliminating this problem -- any idiot can stuff a date into a Date class, which then encapsulates all the date-handling code.
Maybe some of it is very well designed. Drupal proves that you can write good, elegant code in any language, even if you are fighting the language and reinventing the wheel every step of the way. But the converse is also true -- you can write bad COBOL in any language.
My point here is that when changing minimum wage is even a techstory at all, that program is really fucking broken*. It's very likely too broken to be patched. Really, we've learned things in the past 50 years, and not all of them are buzzwords or ways to waste five times the RAM.
Not all of them have anything to do with programming languages, either, but if you're building a new system, and you have a choice of languages, why would you choose COBOL?
I agree in spirit. But what people have to remember is, if it ain't broke, don't fix it. So, if it's broke, fix it!
* I apologize for the profanity, but any program that can't change a fucking constant is a broken program. Or did they copy/paste 6.55 all over the place?
Slashdot Mirror
I dual-boot. I boot Windows when I want to game, and Linux when I want to do anything else.
If gaming is the only thing keeping you on Windows, I'd suggest the same. At least for me, it beats buying a console.
Why am I being offered Gnome and KDE? What's the difference?
Or you installed Ubuntu, and that question is resolved, until you decide you want to dig deeper.
Why can I not play DVDs?
Because you didn't buy it preinstalled from a place like Dell, which handles that for you.
Try installing a fresh copy of Windows, and you'll find that it cannot play DVDs either, unless you happened to get a copy of PowerDVD with that DVD drive. You'd have to download VLC anyway.
What's the app called for playing MP3s and ripping CDs? Alright everyone, which one of the dozen is going to work best?
Probably the one that's preinstalled.
How would you answer this on Windows? iTunes? There are dozens of others you could've downloaded. Windows Media Player? Sucks, but it's preinstalled. Those are the choices.
Oh, and if you have to download one: Amarok for playback, k3b for ripping -- that's what came by default on my Kubuntu. I'm not sure what comes by default on Ubuntu, but chances are, it will have a description right next to it in the menu.
Contrast with Windows: Where are my drivers?
Ok, let's see. What kind of video card do I have? Let's look up the manufacturer's website, looks like it's nvidia, ok, head over to nvidia.com. WTF? Where's the download? Nvidia tells me I have to go to my manufacturer's website. OK, back to Dell.com. WTF? Vista-only? This is bullshit, nvidia certainly maintains XP drivers, but due to a contract with Dell, can't provide them.
Ok, contact support. Support gives me a download for a different model of laptop that happens to have the same video card. Support informs me that there's no way I could know or discover this on my own -- I would just have to contact support.
And that's just the video card. The best wireless drivers were actually not the default ones, as Intel ProSet is utter crap compared to the default Windows wireless tools. I base this mostly on the fact that ProSet is completely unnecessary -- maybe there is something it does better, but I haven't found it yet. Regardless, I had to find the "driver only" download.
And on down the list -- AHCI is really needed to make the best use of this hard disk, but XP doesn't support it without a driver. One option was to disable it in the BIOS and re-enable it later. The option I chose was to slipstream the driver into the XP CD -- a process significantly more involved than burning an Ubuntu ISO.
Why is it running so slow? Why does it take 15 minutes to shutdown?
I noticed you didn't list a solution to that. That's because the solution is pretty much what you said for one Linux problem:
Okay, reinstall. Why did I not make notes?
The solution to each, of course, is to back up a working copy, or to take notes, possibly in the form of a script.
The difference is, I actually know how to fix a Linux problem. The fix for a Windows problem is generally to reboot, and if that doesn't work, reinstall. This may be my own lack of knowledge, but others seem to notice the same thing, and certainly Windows professionals seem to have little to suggest beyond running Spybot and reinstalling. So maybe it's related to the "freedom to tweak"?
And on "enterprise" distros, where versions are locked for years while patches are backported, you suffer dependency hell, which forces you to run older versions of software until you can make the business case to upgrade the OS and all the commercial apps.
While having everything be of a constant version (and thus designed to work well together) is useful, upgrading doesn't generally break everything. Especially when your only compatibility problems will be with that old software, it should be easy enough to test the upgrade and make sure it still works.
This is more or less the same as the case with Windows -- notice how businesses are still running XP, which has had things slowly patched and backported. Most ignored Vista, and most are now looking to skip Windows 7.
Ok, but I'm not a driveling whiny developer enthusiast that needs to have the bazillion levels of freedom that you need to hack the bejeezus out of your computer. I'm a burger flipper, a tire guy, a mechanic, a professional, or a housewife and I just want the stuff to work.
In that case, the bazillion programs available for Windows shouldn't matter -- only the few you need to work. Additionally, the lack of a need for antivirus, and the ease of keeping your system up-to-date and secure, should appeal to you.
In fact, even a package manager and a distribution should benefit you, in the long run. Choosing software supported by the distro means it'll be maintained, likely forever and for free. Using a distro like Debian or Ubuntu, which has separate stable and unstable versions, means that as long as you're on the stable version, all of that software is known to work together -- no "dll hell", no other strange cases of one piece of software causing another to not work.
I don't want to have to make a stupid decision about which distribution I should download
That's why we say "Ubuntu" and move on.
and I don't want to have to answer nine billion technical questions just to get it installed.
I'm sure someone can verify it, but I don't think Ubuntu asks more questions than XP. If you're a professional, you solve this problem by getting it preinstalled.
I want to have that feeling that there is a company that I can blame,
That would be Dell, who is providing you service, if you followed the above option.
I need to have the feeling that there is a group of people that may benefit from my purchase,
That, I really don't get. Since it can be free, why would you need that? If you get it as a product, with someone to blame (the Dell option), then Dell and Canonical both benefit, and some portion of your money goes directly to improving Ubuntu.
Ubuntu is coming the closest to being a product and it's goals are commendable but it is not a mainstream desktop PRODUCT and never will be.
Why not? Putting PRODUCT in all caps (and bold) doesn't make it a valid point. Your actual points here, I think I've refuted.
Just because it's free doesn't mean it is going to be good and just because it costs a lot of money doesn't mean that it's evil.
This is true. However, the fact that it is free, in a truly level market economy, would mean that anything that costs money would have to come with a lot of added value.
As it is, the closest competitor, in the sense of something for which most software is compatible, might be Solaris (and other commercial Unices), but Solaris was recently open sourced -- Linux dominates that market. OS X might count, except their GUI is so proprietary that a truly native OS X app can't be much more easily ported to Linux than a Windows app can.
I'l probably get modded a 0 flame bait for this
I really hope mods stop falling for this tactic.
Hey, mods, I'm about to say something that people might not want to hear! Some people might mod me down for it! You'd better mod me up to compensate!
I'd have modded you overrated, but I actually have something to say.
That, or if you're partnered closely with a designer.
I worked for a web company, and we did use a designer. We'd give them a rough idea of what we wanted, they'd send us back HTML and CSS, and we'd just have to replace Lorem Ipsum with template markup, then go play with backend logic.
The company we hired was basically one guy who was 100% artistic, and one guy who turned it into HTML/CSS and appropriate images.
Eventually, we started doing the design in-house, mostly because we decided that we could do it simpler and better than a web designer -- mostly because we wanted something absurdly minimalist. But it did take a lot of time, even with that "minimalist" route.
As a developer, what was probably the most frustrating was either having to design things myself (and having them look ugly), or being given a gigantic photoshop file and having to slice out the parts that should be images, measure pixel distances between everything, and deal with all the warts of CSS in Internet Explorer. But, it was better than having to deal with something DreamWeaver (or worse, MS Word or Adobe GoLive) crapped out.
So yes, there is definitely a job in it, and it's a job I don't want to do.
For what it's worth, Schneier did notice something that makes sense, even if you're neither a security expert nor a geo-strategist:
Read the whole story; there aren't really any facts in it.
Enough said.
Each broadband customer must have one IP anyway, so it does not matter if they give you a fixed one or not.
Yes and no.
Yes, they are gouging. All ISPs seem to, with this.
No, it's not zero cost. Assume they pull it from the DHCP pool -- that means reconfiguring the DHCP server, and possibly killing a lease (thus interrupting service for that person). There are likely graceful ways to do it, but they're not exactly that easy.
But then, you'd think they would want to encourage static IP addresses. It might not be much, but DHCP does require a server, some amount of bandwidth, CPU, and disk space, and is yet another point of failure should something go horribly wrong.
Seriously, just use dyndns, dnsmadeeasy, zoneedit, or any of the many, many free services out there for dynamic DNS.
your assertion that it only costs $0.16 is stupid. First, they need to also pay electricity, pay for employees to take your service calls, employees to do line repairs, repair existing lines when they are damaged, the gas to get those trucks out to repair the line, the lease on the truck itself, and then still have enough for upgrades and to make a profit...
Amazon has to do all of this, and manages to charge $0.17 per gig down, $0.10 per gig up. Flat, across all their services.
Or I suppose AWS doesn't run on electricity, nor require any kind of service or customer support?
Alright, yes, an ISP has a bit more ground to cover, physically, than a datacenter. On the other hand, Amazon is either plugged directly into a backbone, or they have an ISP too.
But really, are those physical lines from my ISP to my house really costing enough to explain that it costs ten times as much?
Worth mentioning, that's initially a win for the service provider -- the slides are packed, but each kid is still paying $20. Even if you assume the waterpark was fully utilized at $20 for 10 slides, that means they have at least twice as many kids getting 5 slides, but still paying $20 each.
What I've humbly suggested is, instead of pocketing the extra money, spend it on building more slides. It may take time, but you'll eventually get even more kids paying $20 each, and having a lot more fun.
Instead, ISPs have done the equivalent of advertising even more, and pulling in even more people, until no one can get more than about 2 slides. Some kids figure out exactly which slides to use, and when to use them, coordinate with each other to plan it out, and somehow manage to get 20 slides each -- thus forcing everyone else down to 1 slide. The pool cracks down by randomly kicking people out who "slide too much" without defining it.
Eventually, after enough people complain, the park goes back to its original model, this time stating explicitly that you get 5 slides, and more than that costs extra.
All of this just to make a few bucks in the short term.
What shocks me is that there's no competing waterpark with five times as many slides to pull all the customers away. Surely, at least some ISPs have to have gotten it right, and invested in infrastructure, right?
Oh, and it was a good analogy, sorry to abuse it...
That's why you need tiered cap pricing if you go to that - for some people 5GB is quite enough, others would use it in a day. Tiered pricing allows people to decide for themselves what capacity they need...
And metered means that if those usage patterns change, they can still pay a reasonable amount, rather than suddenly being bumped up to a higher tier. Given the Steam example, what happens if I never really used much bandwidth, but one day I downloaded Steam? Do I get a pile of overage charges and the offer to add a flat $20/mo to my Internet bill? to cover this sort of usage? Or do I get a few little spikes of $1 or $2 here and there, if that?
VOD, for example, would probably get less uptake as people compare that cost to renting or watching a movie on pay per view.
I assume that's "video on demand", right?
Alright, let's compare. Amazon charges 17 cents per gig for downloading from Amazon Web Services. That gets pricey if you're downloading an absolutely pristine, full Blu-Ray disc -- which you're almost certainly not. A very good quality Blu-Ray rip will fit into 10 gigs, or $1.70 -- still cheaper than the rental, and you don't necessarily have to return it. And most services compress it a hell of a lot more.
If it's metered and still too expensive, sure. If it's metered at a reasonable rate, I don't think it becomes a problem.
Many content owners are waking up to the idea they can cut distribution costs significantly by internet based delivery; thereby making more money by cutting out the middle man and essentially passing the cost of the tangible media to the buyer eliminating all the reproduction, storage and distribution costs; as well as be freed from battles over format (Blu-ray vs HD and whatever comes next.)
The vast majority... well... where do you think Blu-Ray (or HD-DVD) came from in the first place?
Granted, it was probably content creation corporations, not individual creators, but the fact is, they're often the first to shoot themselves in the foot over a piracy scare. Let's not forget the whole "VCR is the Boston Strangler" analogy. The smart ones picked up on Internet distribution a decade ago -- the dumb ones think "Internet Distribution" means getting their music on iTunes -- or worse, PlaysForSure/Zune.
...why?
Really, I have no idea what the point is. Here's a TFA URL:
http://joshua.schachter.org/2009/04/on-url-shorteners.html
Here's what yours might look like:
http://joshua.schachter.org/89dfaf0834055017af95b8cbb8b440819c3db49a
Congratulations, it's longer. What's the gain?
one of the biggest problems: 2 access to get that URL.
How is that a problem?
The main reason it's a problem with a service like tinyurl is that one of those is to a completely different server. This means two points of failure for the request (TinyURL could be down), multiplied by all the stages involved (The TinyURL DNS could be down...)
And of course, the larger problem of a single point of failure -- TinyURL going away breaks all those links. TinyURL also gets data on all those links.
Let's face it -- multiple connections will be opened per page anyway, for ads alone, let alone scripts, CSS, images, etc -- unless they're pipelined via HTTP 1.1. Either way, an extra HTTP GET is just not a big deal these days -- especially when it pretty much only has to return this:
Even when you consider the three-way handshake (that is, assuming it doesn't hold the connection open as a pipeline), that's just not a lot of data.
I suppose it matters if you're on an absurdly high-latency connection. Other than that, I see cache coherency and resource integrity as much higher priorities.
First, is there a real reason to have that?
The main reason not to is identity. A resource should generally be accessible from a single URL. That's what the redirect is for -- to identify where the canonical resource lives.
Yes, it looks hideously long. It also works fine, it's clickable, I really don't get the big deal.
Given that I've also never once had a kernel upgrade or OS upgrade of any kind completely trash my filesystem, or hard drive, or controller, etc.
Having a backup of my data and configuration is useful. Backing up several gigs of software really isn't -- worst case, I lose a few hours re-downloading it.
we really had no good way to guide students effectively in learning the interface.
Which makes me curious, again -- how much would they figure out on their own?
Consider that children often figure out the "child-proof" medicine bottles more quickly than adults. Or at least, my parents would often give them to me to figure out.
asking how she could make use of it got nowhere because she would not be allowed to make any use of it.
I assert that this really isn't Sugar's fault. If she's actually being dictated which UIs she can teach and which she can't, I doubt xfce or anything else would've been better, unless it was the XP that eventually got grafted onto it.
The educational bureaucracy in most countries is mostly familiar with Windows on personal and institutional bases, so they will be much harder to interest in adopting computing paradigms (such as Sugar) that seem irrelevant to what they perceive as mainstream, i.e. Windows.
Which was exactly why OLPC wasn't targeting existing beaurocracies so much as developing countries -- but rather targeting the kind of community where it's likely even the teachers haven't been exposed to computers much before.
Of course...
And you can be sure M$ was leaning on them to continue to see it that way.
That's the real tragedy here. Because I believe there was an opportunity to introduce something genuinely new to a large portion of the world.
Instead, Microsoft wins again.
But no, I don't think differentness played a large role there. After all, the common challenge for Linux has always been, if it is too different, people used to Windows won't be able to follow that learning curve. If it's too much the same, people won't see a compelling reason to use it over the alternative.
Let me ask you: If it looked and felt just like XP, only wouldn't run Windows programs, and was only a few dollars cheaper, would anyone be interested? Only in places where a few dollars really matters -- and even there, the perception would be that this is the cheaper version, and that you get what you pay for -- not that this is the revolutionary new vision, that also happens to be cheaper.
Depends on the game. In Day of Defeat, for instance, the Nazis win as often as the Allies.
I'd say what's good about multiplayer games is that you're playing against human opponents. A bot doesn't get angry when it loses, or gloat when it wins. There's a certain amount of creativity lacking in a bot -- and humans tend to have entirely different and unpredictable flaws and strengths.
I've played multiplayer games in which particular humans had a severe handicap, and it was still fun.
No, the main thing I don't like about special treatment for the player is that it's unrealistic, and it's a cop-out -- plenty of games manage to make it work without that. Of course, that's not always a constant -- if the enemies are in a ten-story-tall tank (like the Scarab in Halo), it absolutely should be hard to kill. Even Max Payne had a decent explanation -- the enemies didn't stop with one shot because they were all ridiculously high. But in a GTA game, it really makes no sense.
Mostly because I'm already on 8.10, and I don't like being a full year out of date.
Some of that is actually legitimate -- for instance, various development tools and games would be falling out of date. Some of it's just impulsive -- I occasionally regret updating to 8.10, as KDE4 was really not ready.
Whoosh.
Mod this +5 Funny so we can see it for the sarcasm it is.
(If it's not sarcasm, of course, you are a sad little man.)
My most salient advice is to make a bootable, external-disk full system backup before attempting an Ubuntu upgrade.
I'm far too lazy for that, so I tend to just back everything up to a folder -- even a folder on the same system. It's much easier to downgrade if the entire thing falls apart, but I've never had an Ubuntu upgrade completely kill my filesystem or hard drive.
what is the chance that software compiled for Bolivian KDE will run under Paraguayan KDE? (Hint: WINE)
What is the relevance of that at all? If KDE apps can run under GNOME, and KDE3 apps can run under KDE4 (and vice versa), I seriously doubt Paraguayan KDE will have a problem with Bolivian KDE.
It means that 10am-bedtime needs to sleep more?
If they really were fully targeting the Win32 API their goals would logically be measured by percent of API implemented rather than based on which programs run under the current Win32 subset.
That would be true, if the Win32 API stuck to any sort of standard.
It doesn't. Like many Microsoft APIs, the spec is spotty and incomplete, and since it's so widely deployed in exactly one version, the only "correct" spec (as far as anyone cares) is "how Windows XP does it."
No, Wine's goal has always been "bug for bug" compatibility with Windows. And you can't measure that in percentage of an API with such buggy and incomplete documentation. You can only measure it in actual programs that will run.
That's ignoring, of course, the really strange things Microsoft has had to do -- certain old software will make really stupid assumptions (like the length of the Windows version number), and modern Windows will detect those programs and change the API accordingly (for example, lying about the Windows version so the app doesn't crash).
I don't know if Wine intends to run such programs, but if so, implementing Win32 isn't even close to enough.
I would probably tune in to watch people play Quake.
In fact, so would you. Look here.
Sure, greater productivity is one benefit, but the language is completely irrelevant for that.
It's about how flexible the system will be when you have to change it. And you will -- that's the whole point of software, that it is soft, and changeable.
Old Cobol apps generally are not flexible. (stolen from this comment). It's worth mentioning that a decent object-oriented system would've gone a long way towards eliminating this problem -- any idiot can stuff a date into a Date class, which then encapsulates all the date-handling code.
Maybe some of it is very well designed. Drupal proves that you can write good, elegant code in any language, even if you are fighting the language and reinventing the wheel every step of the way. But the converse is also true -- you can write bad COBOL in any language.
My point here is that when changing minimum wage is even a tech story at all, that program is really fucking broken*. It's very likely too broken to be patched. Really, we've learned things in the past 50 years, and not all of them are buzzwords or ways to waste five times the RAM.
Not all of them have anything to do with programming languages, either, but if you're building a new system, and you have a choice of languages, why would you choose COBOL?
I agree in spirit. But what people have to remember is, if it ain't broke, don't fix it. So, if it's broke, fix it!
* I apologize for the profanity, but any program that can't change a fucking constant is a broken program. Or did they copy/paste 6.55 all over the place?