The nice thing about Free/Libre Open-Source Software is that you can modify it, if it does not behave the way you want. Assuming Android actually pop an ad at random (I seriously doubt it), you are welcome to remove the ad-poping bit.
Re not being able to run "normal" Linux software, I think it is really a matter of considering the use case and scope of what the XO is trying to achieve. The CPU should be plenty fast enough for a custom web browser, text editing software and IM. "Normal" Linux software, such as Firefox and OpenOffice, do not really scale down to such a tiny screen anyway, so you have to think outside the box a bit when choosing application for these platform. I think power consumption and cost are more important factors that raw CPU power in the use case that the XO is targeting.
BTW, I have seen Compiz running on the ClassMate. Cool, but... what the heck ?
The ClassMate is almost ruggedized enough with the vinyl wallet, but I would trust an XO better in the hand (and school backpack) of my daughter. I would never let her handle a Eee, it is way too flimsy for kids (although the best overall package for adults).
Because you want to do good (whether OLPC is doing good or not is your call), and because you want a piece of tech history.
The XO basically revolutionize the low-end portable computer market. They where the first to talk about ultra low cost, ultra-portable, low-power computing, and as such kick-start the movement which gave us recently the Asus Eee and the Intel ClassMate. Without them, the market would have slowly converge toward cheaper and cheaper hardware, but I think we would still be a couple years away before the laptop industry change its mindset from "let's pack more power for the same amount of money" toward "cheap enough to be ubiquitous". I doubt OLPC and the XO would get credit for that, unfortunately.
The XO is also a marvel of engineering. The Sugar interface is the first completely kid-centric interface to come out of Open-Source, and leading the pack on that front. The hardware have been carefully thought-out, and it show. It may not be the best laptop for an adult (it is completely kid-centric), but this not just a bunch of discrete parts thrown together by an ODM. Finally, as if it was not enough, five words: built-in Wifi mesh network. If that does not scream "cool" to you, you need to hand over your geek card right away.
So go grab a rebranded Asus on Black Monday and enjoy your crap imitation of a real laptop. In the meantime, I will continue to use my five years old Thinkpad, which will probably outlive the pathetic knock-off you can buy in the price range you quote.
I have held both an XO and an Eee in my hands (and a ClassMate, too), and the XO clearly beat the competition in term of built and robustness. It was built expressly for the purpose of surviving usage poor condition (dust, humidity, heat) and are totally centered around the needs of kids, while the Eee and ClassMate where built to be a shrunk-down imitation of full-sized laptop and be as cheap as possible.
As an adult, I prefer the Eee though, mostly because I do not like the XO rubberized keyboard.
Because 8.04 will be a LTS release (Long Term Support), and it is expected that the Ubuntu developers will be conservative with the feature set they allow into 8.04. As such, if you have a feature that is somewhat experimental, you need to push it now (to get it tested and polished before 8.04), or wait until 8.10. At least, that's the theory. In practice, I am fairly certain quite a few experimental features will find their ways into 8.04 anyway.
Managing releases at fixed date and coordinating with upstream project release is probably the toughest challenge Ubuntu is facing. But on the other hand, this is exactly what gave it the edge in the distro war. So far, the execution have been pretty good and Ubuntu reap the benefits.
File system layout have been standardized in the Filesystem Hierarchy Standard and, AFAIK, all Big Three (Red Hat, Novell, Ubuntu) have been adhering to it pretty strictly. Anyway, third-party commercial applications should go in/opt and not mess around with the rest of the system, period.
Different distros ship different version of libraries, yes. But naming convention for libraries differ from distro to distro ? I do not understand what you talk about here.
And, yes, I have installed "enterprise" software, and their installer have been pretty consistently an awful hack slapped together as an afterthought by the vendor. And that include the horrible Oracle Java wizard. So, the blame lie squarely on the vendors as far as the crappiness of their installer is concerned if you ask me.
The difference was how Wozniak went about engineering focusing on usability and openness.
Huh?
I readily acknowledge the focus on usability. But openness ? They did ship schematic with the Apple II, but that does not extend to the Macintosh, which was (and still is to some extent) as closed as you can get.
Outside the billboard (only one, and then only for a few weeks) and the ShipIt program, they have done no advertising whatsoever. I never saw an ad in a printed magazine or on a web site. You are welcome to prove me wrong if I missed something...
Except, in this case, Red Hat *is* the incumbent offering a single one-size-fit-all product, and getting its ass kicked by up-and-coming competitors.
Ubuntu is not the only player in the field, but it will stay the only player with a chance to succeed as long the current incumbents (Red Hat and Novell) continues with their completely inane strategies of charging yearly subscription fees for their desktop products.
And Ubuntu's competition is not Red Hat, it is Microsoft and Apple.
The reasons Ubuntu came to dominate the market is multi-folds.
First, they pionneered three keys Linux distributions improvements: single ISO installer, clean desktop and LiveCD.
People seem to have forgotten that, but back when Ubuntu 4.10 came out, you needed to download *5* ISO to install Fedora (I think you could have gotten away with three if you did the minimal install, but whatever). This was an absolutely horrible experience for Linux first-timer and an important barrier to adoption.
You also have to remember that the Linux desktop in 2004 was quite busy and not very sleek. And quite frankly, Fedora default theme back then was quite ugly (the Gnome/KDE unified theme, cannot remember the name). In contrast, Ubuntu was quite slick, as long as you liked brown. The Gnome-only policy also made it look very well-integrated, and the menus where clear and concise. It was the most Mac-like distro back then.
As for LiveCD, it was really pionneered by Gentoo (credit where due), but I think combined with the two advantages above, significantly contributed to lower the bar for Ubuntu newcomers.
There are other non-obvious reasons why Ubuntu came to dominate the Linux desktop. First and foremost was (and still is) community advocacy. This can be explained by three reasons: a genuinely open development community, a charismatic leader and ShipIt.
The Ubuntu community is, as far as I know, the only Open-Source community that actively try to recruit non-developers and give them the same status as core devs. This made a lot of advocates flock to Ubuntu, as they felt respected and welcomed, where their contribution are considered minor at best (and often seen as distracting or annoying) in other distro's community. These advocate started Local Community Team, got peoples involved in translation and documentation, etc. Democratic and civil community governance mechanisms (the Community Council and the Code of Conduct, among others) kept them under the fold of the community. AFAIK, Ubuntu have been the best and most successful community-building experience in Open-Source.
Leadership is also an important facet of Ubuntu success. Who is the leader of Fedora, or OpenSuse, or Gentoo ? Nobody know outside of their community. A charismatic spokeperson who can make headlines is very useful in spreading the word. And, in Open-Source, a clear leadership is vastly better than a purely democratic and collegial form of governance, which tend to lose focus or spend too much time in unproductive flamewars (witness: Debian). Ubuntu stuck just the right balance of benevolent dictatorship and community governance.
The importance of ShipIt in Ubuntu is fairly obvious. It got Ubuntu in the hands of people who could not have it otherwise. Plus, it's easier to have someone give a spin to Linux if you hand them a nice, professionally made CD with a cool jacket than if you hand them a CD-R with "Fedora CD 1 of 5" scribbled to it. It must have costed Shuttleworth is metric truckload of money to support ShipIt so far, but it bought him the dominant poistion in the market.
There are more reasons why Ubuntu kicked the ass of others Linux distros in the market, but this post is already long enough.
According to TFA, Red Hat is targetting public administration and small business in developing countries. This is a very price-conscious market. If the only competitive advantage they have over the competition (Ubuntu, pirated Windows, etc) is that they offer some warm-and-fuzzy feeling that the product is supported by a corporation, they are doomed to fail. Canonical already offer support à la carte (you buy support only if you need it), which make Red Hat Global Desktop compulsory subscription a fairly though sell.
Red Hat (and Novell) strategy of charging per-seat "subscription" is doomed to fail on the desktop. Really, this is paramount to the proprietary software business model of charging licensing fees per seat. And why would anybody choose to engage a recurring cost for an *operating system* is beyond me (but then, people flocked to "Software Assurance", go figure). To have any chance, they would need to charge very little for this "subscription", which raise the question of profitability. Maybe they would have a chance if they where giving away these desktop "subscriptions" to existing customers of RHEL as a perk.
Red Hat never understood the Linux desktop market, and apparently never will. It is a good thing they dominate a profitable niche in replacing Solaris as a platform to run Oracle and other enterprise software, because they completely suck at market development. I would hate them to go away; they are very goods corporate Open-Source citizens that contributes significantly to key Open-Source project, so I hope this niche will not dry up in the near future.
As a side note, if you think Red Hat can afford to dispatch a field engineer for desktop problem on the premise of a small business customer, your expectations need a little adjustment.
Your choice of curriculum in higher education is not relevant to your career path, at least not for people in IT. The field is full of self-taught and people with degree in unrelated disciplines (math and history being prevalant in my circle). Your degree may help with the first job or two, but that's it. After that, it's all about your network and your soft skills: leadership, people skill, reliability, etc.
But don't despair just yet. University is actually of great help with starting your network, and provide plenty of opportunities to improve the soft skills necessary to succeed in the job market. So, do go get a degree. Any degree will do, just choose a discipline that interest you and will keep you motivated.
If I where 18 again, I would do philosophy. And I would probably end up in the same place, only a bit wiser.
(While here, I would like to say that AMD is becoming less helpful day
by day towards open source operating systems too, perhaps because
their serious errata lists are growing rapidly too).
I bet your client's VPN was a Cisco 3000 (really just bastardized IPSec, apparently). If this is the case, instead of installing the crappy binary Cisco client, you could just have installed the vpnc package, which is an open-source client for this line of VPN. And unlike the Cisco binary crap, it would not disconnect every five minutes and would not lock out all your network interfaces to force all traffic through the tunnel.
No patching kernels and recompiling would have been involved, but you would have had to use Google to learn about it. I guess you are above learning what's required to do your job. Too bad for you, carpenter blaming tools and all that.
I will consider buying a Mac laptop the day their keyboard will stop sucking so bad. I cannot believe that Apple get away with such flimsy and ugly keyboards on their recent laptop line. Even the keyboard on low-end Asus generic laptop are better than those on Powerbook.
In the meantime, I will continue to buy refurbished Thinkpad. These people know how to build reliable laptops with good ergonomy (decent keyboard and no stupid trackpad on my X30).
First and foremost, the big killer for me was that it was a resource hog. Sure, on today's machine, no prob. But back then in 1997, I was running a computer lab made of 133/166 Mhz machines equipped with 32 MB of RAM. Starcraft played no problem (except when everybody in a game where maxing out their units), but TA was dog slow from the start. And thus no fun.
The matter of complexity is also an important one for cusual gamers. TA number of units and technology tree made it way too complex for the average guy.
So, in my case, when people where gathering for a quick and dirty impromptu LAN party, they invariably choosed Starcraft. TA was just too slow, too complex and too long to play.
Do you seriously think the world would be better off if we replaced alcoholics by an equal number of pot addicts?
If such a thing as "pot addicts" existed, I would say yes. Alcoholics tend to be loud, rowdy, violent and irresponsible. Serious potheads tend to be passive and detached. Beside smelling bad and being lazy, I actually have very little against heavy pot smokers.
This discussion also uncover one of the biggest problem about this whole drug frenzy. Unknowing people (including you, apparently) gladly lump pot in the same category as cocaine, heroine, amphetamine and other hard drugs, which it is clearly not, while nicely sheltering themselves from the fact that alcohol is one hell of a dangerous substance by itself. But since alcohol is legal and culturally accepted, it cannot be bad, right ?
Don't get me wrong: I have my fair share of booze regularly, I just find the precious irony and total lack of perspective here delicious.
I do not hear about alcoholics committing robberies to get something to drink remotely as often as I see stories of junkies doing so for crack, heroine and other hard drugs...
What about pot smokers ? I do not hear about pot smokers committing domestic violence, killing people while driving under influence and getting into fist fight remotely as often as I see stories of alcoholics doing so when drunk.
These guys often have a toe or more in organized crime and are far more likely to sell your banking and credit information to get their fix than your average alcoholic.
Back in Prohibition days, getting alcohol to drink involved dealing with the organized crime too. The drug <-> mob link is entirely due to the current state of prohibition; the day you will get your pot from a local farmer or coffee shop, this link will disappear.
And drug use correlates to other behaviors generally considered undesirable in an employee.
So do alcohol. Thus it is sensible and economical to use alcohol screening in the hiring process. Why are'nt bank already doing so ? I do not want to trust my financial information to alcoholics !!!
Slashdot Mirror
The nice thing about Free/Libre Open-Source Software is that you can modify it, if it does not behave the way you want. Assuming Android actually pop an ad at random (I seriously doubt it), you are welcome to remove the ad-poping bit.
Gaaaah, I have been trolled again.
No, but I have two of those!
Re not being able to run "normal" Linux software, I think it is really a matter of considering the use case and scope of what the XO is trying to achieve. The CPU should be plenty fast enough for a custom web browser, text editing software and IM. "Normal" Linux software, such as Firefox and OpenOffice, do not really scale down to such a tiny screen anyway, so you have to think outside the box a bit when choosing application for these platform. I think power consumption and cost are more important factors that raw CPU power in the use case that the XO is targeting.
... what the heck ?
BTW, I have seen Compiz running on the ClassMate. Cool, but
The ClassMate is almost ruggedized enough with the vinyl wallet, but I would trust an XO better in the hand (and school backpack) of my daughter. I would never let her handle a Eee, it is way too flimsy for kids (although the best overall package for adults).
Because you want to do good (whether OLPC is doing good or not is your call), and because you want a piece of tech history.
The XO basically revolutionize the low-end portable computer market. They where the first to talk about ultra low cost, ultra-portable, low-power computing, and as such kick-start the movement which gave us recently the Asus Eee and the Intel ClassMate. Without them, the market would have slowly converge toward cheaper and cheaper hardware, but I think we would still be a couple years away before the laptop industry change its mindset from "let's pack more power for the same amount of money" toward "cheap enough to be ubiquitous". I doubt OLPC and the XO would get credit for that, unfortunately.
The XO is also a marvel of engineering. The Sugar interface is the first completely kid-centric interface to come out of Open-Source, and leading the pack on that front. The hardware have been carefully thought-out, and it show. It may not be the best laptop for an adult (it is completely kid-centric), but this not just a bunch of discrete parts thrown together by an ODM. Finally, as if it was not enough, five words: built-in Wifi mesh network. If that does not scream "cool" to you, you need to hand over your geek card right away.
So go grab a rebranded Asus on Black Monday and enjoy your crap imitation of a real laptop. In the meantime, I will continue to use my five years old Thinkpad, which will probably outlive the pathetic knock-off you can buy in the price range you quote.
... should be "Plus ça change, plus c'est pareil". Same literal meaning, but yours is not the common French usage.
I have held both an XO and an Eee in my hands (and a ClassMate, too), and the XO clearly beat the competition in term of built and robustness. It was built expressly for the purpose of surviving usage poor condition (dust, humidity, heat) and are totally centered around the needs of kids, while the Eee and ClassMate where built to be a shrunk-down imitation of full-sized laptop and be as cheap as possible.
As an adult, I prefer the Eee though, mostly because I do not like the XO rubberized keyboard.
... says the guy who cannot be bothered to use uppercase properly.
Because 8.04 will be a LTS release (Long Term Support), and it is expected that the Ubuntu developers will be conservative with the feature set they allow into 8.04. As such, if you have a feature that is somewhat experimental, you need to push it now (to get it tested and polished before 8.04), or wait until 8.10. At least, that's the theory. In practice, I am fairly certain quite a few experimental features will find their ways into 8.04 anyway.
Managing releases at fixed date and coordinating with upstream project release is probably the toughest challenge Ubuntu is facing. But on the other hand, this is exactly what gave it the edge in the distro war. So far, the execution have been pretty good and Ubuntu reap the benefits.
File system layout have been standardized in the Filesystem Hierarchy Standard and, AFAIK, all Big Three (Red Hat, Novell, Ubuntu) have been adhering to it pretty strictly. Anyway, third-party commercial applications should go in /opt and not mess around with the rest of the system, period.
Different distros ship different version of libraries, yes. But naming convention for libraries differ from distro to distro ? I do not understand what you talk about here.
And, yes, I have installed "enterprise" software, and their installer have been pretty consistently an awful hack slapped together as an afterthought by the vendor. And that include the horrible Oracle Java wizard. So, the blame lie squarely on the vendors as far as the crappiness of their installer is concerned if you ask me.
Huh?
I readily acknowledge the focus on usability. But openness ? They did ship schematic with the Apple II, but that does not extend to the Macintosh, which was (and still is to some extent) as closed as you can get.
Outside the billboard (only one, and then only for a few weeks) and the ShipIt program, they have done no advertising whatsoever. I never saw an ad in a printed magazine or on a web site. You are welcome to prove me wrong if I missed something ...
Except, in this case, Red Hat *is* the incumbent offering a single one-size-fit-all product, and getting its ass kicked by up-and-coming competitors.
Ubuntu is not the only player in the field, but it will stay the only player with a chance to succeed as long the current incumbents (Red Hat and Novell) continues with their completely inane strategies of charging yearly subscription fees for their desktop products.
And Ubuntu's competition is not Red Hat, it is Microsoft and Apple.
The reasons Ubuntu came to dominate the market is multi-folds.
First, they pionneered three keys Linux distributions improvements: single ISO installer, clean desktop and LiveCD.
People seem to have forgotten that, but back when Ubuntu 4.10 came out, you needed to download *5* ISO to install Fedora (I think you could have gotten away with three if you did the minimal install, but whatever). This was an absolutely horrible experience for Linux first-timer and an important barrier to adoption.
You also have to remember that the Linux desktop in 2004 was quite busy and not very sleek. And quite frankly, Fedora default theme back then was quite ugly (the Gnome/KDE unified theme, cannot remember the name). In contrast, Ubuntu was quite slick, as long as you liked brown. The Gnome-only policy also made it look very well-integrated, and the menus where clear and concise. It was the most Mac-like distro back then.
As for LiveCD, it was really pionneered by Gentoo (credit where due), but I think combined with the two advantages above, significantly contributed to lower the bar for Ubuntu newcomers.
There are other non-obvious reasons why Ubuntu came to dominate the Linux desktop. First and foremost was (and still is) community advocacy. This can be explained by three reasons: a genuinely open development community, a charismatic leader and ShipIt.
The Ubuntu community is, as far as I know, the only Open-Source community that actively try to recruit non-developers and give them the same status as core devs. This made a lot of advocates flock to Ubuntu, as they felt respected and welcomed, where their contribution are considered minor at best (and often seen as distracting or annoying) in other distro's community. These advocate started Local Community Team, got peoples involved in translation and documentation, etc. Democratic and civil community governance mechanisms (the Community Council and the Code of Conduct, among others) kept them under the fold of the community. AFAIK, Ubuntu have been the best and most successful community-building experience in Open-Source.
Leadership is also an important facet of Ubuntu success. Who is the leader of Fedora, or OpenSuse, or Gentoo ? Nobody know outside of their community. A charismatic spokeperson who can make headlines is very useful in spreading the word. And, in Open-Source, a clear leadership is vastly better than a purely democratic and collegial form of governance, which tend to lose focus or spend too much time in unproductive flamewars (witness: Debian). Ubuntu stuck just the right balance of benevolent dictatorship and community governance.
The importance of ShipIt in Ubuntu is fairly obvious. It got Ubuntu in the hands of people who could not have it otherwise. Plus, it's easier to have someone give a spin to Linux if you hand them a nice, professionally made CD with a cool jacket than if you hand them a CD-R with "Fedora CD 1 of 5" scribbled to it. It must have costed Shuttleworth is metric truckload of money to support ShipIt so far, but it bought him the dominant poistion in the market.
There are more reasons why Ubuntu kicked the ass of others Linux distros in the market, but this post is already long enough.
According to TFA, Red Hat is targetting public administration and small business in developing countries. This is a very price-conscious market. If the only competitive advantage they have over the competition (Ubuntu, pirated Windows, etc) is that they offer some warm-and-fuzzy feeling that the product is supported by a corporation, they are doomed to fail. Canonical already offer support à la carte (you buy support only if you need it), which make Red Hat Global Desktop compulsory subscription a fairly though sell.
Red Hat (and Novell) strategy of charging per-seat "subscription" is doomed to fail on the desktop. Really, this is paramount to the proprietary software business model of charging licensing fees per seat. And why would anybody choose to engage a recurring cost for an *operating system* is beyond me (but then, people flocked to "Software Assurance", go figure). To have any chance, they would need to charge very little for this "subscription", which raise the question of profitability. Maybe they would have a chance if they where giving away these desktop "subscriptions" to existing customers of RHEL as a perk.
Red Hat never understood the Linux desktop market, and apparently never will. It is a good thing they dominate a profitable niche in replacing Solaris as a platform to run Oracle and other enterprise software, because they completely suck at market development. I would hate them to go away; they are very goods corporate Open-Source citizens that contributes significantly to key Open-Source project, so I hope this niche will not dry up in the near future.
As a side note, if you think Red Hat can afford to dispatch a field engineer for desktop problem on the premise of a small business customer, your expectations need a little adjustment.
Your choice of curriculum in higher education is not relevant to your career path, at least not for people in IT. The field is full of self-taught and people with degree in unrelated disciplines (math and history being prevalant in my circle). Your degree may help with the first job or two, but that's it. After that, it's all about your network and your soft skills: leadership, people skill, reliability, etc.
But don't despair just yet. University is actually of great help with starting your network, and provide plenty of opportunities to improve the soft skills necessary to succeed in the job market. So, do go get a degree. Any degree will do, just choose a discipline that interest you and will keep you motivated.
If I where 18 again, I would do philosophy. And I would probably end up in the same place, only a bit wiser.
From the TFA:
Yay indeed.
I bet your client's VPN was a Cisco 3000 (really just bastardized IPSec, apparently). If this is the case, instead of installing the crappy binary Cisco client, you could just have installed the vpnc package, which is an open-source client for this line of VPN. And unlike the Cisco binary crap, it would not disconnect every five minutes and would not lock out all your network interfaces to force all traffic through the tunnel.
No patching kernels and recompiling would have been involved, but you would have had to use Google to learn about it. I guess you are above learning what's required to do your job. Too bad for you, carpenter blaming tools and all that.
I will consider buying a Mac laptop the day their keyboard will stop sucking so bad. I cannot believe that Apple get away with such flimsy and ugly keyboards on their recent laptop line. Even the keyboard on low-end Asus generic laptop are better than those on Powerbook.
In the meantime, I will continue to buy refurbished Thinkpad. These people know how to build reliable laptops with good ergonomy (decent keyboard and no stupid trackpad on my X30).
TA also had other problem, beside games length.
First and foremost, the big killer for me was that it was a resource hog. Sure, on today's machine, no prob. But back then in 1997, I was running a computer lab made of 133/166 Mhz machines equipped with 32 MB of RAM. Starcraft played no problem (except when everybody in a game where maxing out their units), but TA was dog slow from the start. And thus no fun.
The matter of complexity is also an important one for cusual gamers. TA number of units and technology tree made it way too complex for the average guy.
So, in my case, when people where gathering for a quick and dirty impromptu LAN party, they invariably choosed Starcraft. TA was just too slow, too complex and too long to play.
Clearly, it is in the same league as Apache, Firefox, gcc and the Linux kernel.
Sorry for the disparaging comment about you mixing up pot and hard drug. Rereading your post, I see you do get the nuance.
If such a thing as "pot addicts" existed, I would say yes. Alcoholics tend to be loud, rowdy, violent and irresponsible. Serious potheads tend to be passive and detached. Beside smelling bad and being lazy, I actually have very little against heavy pot smokers.
This discussion also uncover one of the biggest problem about this whole drug frenzy. Unknowing people (including you, apparently) gladly lump pot in the same category as cocaine, heroine, amphetamine and other hard drugs, which it is clearly not, while nicely sheltering themselves from the fact that alcohol is one hell of a dangerous substance by itself. But since alcohol is legal and culturally accepted, it cannot be bad, right ?
Don't get me wrong: I have my fair share of booze regularly, I just find the precious irony and total lack of perspective here delicious.What about pot smokers ? I do not hear about pot smokers committing domestic violence, killing people while driving under influence and getting into fist fight remotely as often as I see stories of alcoholics doing so when drunk.
Back in Prohibition days, getting alcohol to drink involved dealing with the organized crime too. The drug <-> mob link is entirely due to the current state of prohibition; the day you will get your pot from a local farmer or coffee shop, this link will disappear.
So do alcohol. Thus it is sensible and economical to use alcohol screening in the hiring process. Why are'nt bank already doing so ? I do not want to trust my financial information to alcoholics !!!