I've been in this exact situation recently. I regularly go back and forth between macbook pros, microsoft-oriented work laptops, and microsoft-oriented gaming laptops, and hacking-project chromebooks.
I am sure there are exceptions anyone can point to but as a general observation data point from someone who has "been there":
Apple laptops are a known, very high quality, quantity. They don't always have the features you want-- top-of-the-market GPU capability and larger screens is mine, but what Apple does have, they do very well. If apple laptops do deliver everything you really need, then get one, even if you put VMware/Parallels/Other/Bootcamp of it to get solid MS compatibility for everything but high end games. Even if you have to put up with silly dongles.
High-end PC laptops have very good specs but, in general, come with more risks. The more cutting edge features you buy the less likely it is you're going to get good, long term support, and the more likely that you're going to develop some issue a few years out that never showed up in the manufacturer's time-limited testing. The one exception to this may be the MS Surface Pros.
In the midrange prices from reputable manufacturers, support is more predictable, but there's often some performance compromise being made-- you can discover too late that your great price came at the cost of a certain component that's dragging the rest of the system down. Could be wrong CPU for your workload, poor cooling, sub-par screen, KB placement, loud fans, slow HD, inadequate GPU cooling, etc.. It's hard to know what compromise may affect you without doing a lot of review research, and the best, most reliable reviews will be for product that has been out 6+ months, and you may be tempted to get "the latest new model" instead. It's a balancing act.
At the low end you can find some really great values in PCs laptop. If these fit your performance envelope you can get something that's cheap and looks good too.
The above is just on the hardware side. On the software side it goes without saying that there will be a learning curve and a host of annoyances, particularly around driver behavior coming out of sleep mode & hibernation, privacy, and possibly juggling multiple possible audio inputs and outputs.
Great comment, at least in the sense that it resonates with me.
I want to like KDE, because when you get under the hood the Qt framework and many related components are truly well designed, mature software.
But on the visuals side, there's something off. It's not in the tech, it's in the visual polish and consistency, color pallet, icon set, styling. Yes, I know it's customizable but so far, the out of the distro kde-styling seems off to me personally once you open a few different apps and settings panels and see them together; it's hard to pin down.
Despite this unease, I have been generally put off in the design decisions of the GNOME camp in recent years, in particular the intentional removal of optional functionality compared to previous versions, and getting away from a traditional desktop metaphor in favor or something that's more like a smartphone. I don't want a smart phone interface on a high resolution, multi-monitor setup with KB and mouse at the ready.
Cinnamon / XFCE is working for me as a compromise, but I'd love it KDE is able to storm back in with great style, to rescue us from desktop redsign madness.
I've been running crouton on a toshiba chromebook for a couple years now and I use it with surprising frequency.
I spend most of my time on the linux side, running terminals, vpns, and some custom chat and web apps. It's not my main system but I frequently use it as a supplement.
The downside of the config is the annoying/slow "developer mode" bios warning on reboot that you get from unlocking the chromebook bios, and the 3-4 commands you have to run after each reboot to get crouton up and running and happy. Luckily you don't need to reboot often. My average is about once every two months because I've let the battery run down too far. I'm very happy with it especially for the price.
If your passwords are randomly generated and long, it doesn't matter how many attempts to guess them are tried. The likelyhood of a random guess getting through are lower than your chances of winning the lottery. Let people waste their time on futile attempts.
To further decrease your chances, use public keys authentication instead of passwords, or two factor authentication, or limit connections by IP address.
Changing the post does fool most SSH scanners as well.
I don't like fail2ban because it can lead to DoS vectors.
Bottom line is that logged attacks that have no hope of getting through shouldn't cause a panic.
I have an Apple laptop (more like, portable workstation) and I bought it after numerous computer-generations of all kinds of PC laptops, some quite expensive and focused on gaming/performance. I've had it for a year now and I can say that it is the *only* laptop I've ever owned where I've been completely satisfied with the build and service quality. Having a top-flight desktop with an uncompromising unix shell is quite nice too. For gaming I dual boot.
BTW, for a more mainstream data point, the Apple laptops swept Consumer Reports "most recommended buy" in multiple categories recently.
Despite being from a "closed" company, it gives me a platform that lets me natively run Linux, Windows, and MacOSX. It offers more choices. Development tools are much easier to come by as well.
I've done this recently, not just for travelling, but for using networked, 3D accelerated games with rich multimedia that cannot enjoyably run in linux or even played without proprietary kernel drivers outside of a VM sandbox, and which blow normal security practices to bits.
What I did is create a normal linux partition, locked down but still highly usable for multimedia, gaming, and typical virtual machine usage.
Then on one of the internal partitions I created a second, entirely encrypted root partition for a second OS using LUKS. This partition is booted by connecting a USB key and booting from it instead of the normal internal MBR, then entering passwords. The second OS is reasonably secure, locked down, much more limited in functionality, and contains tools to audit the integrity of the multimedia OS and virtual machines, as well as backup and restore them. The USB key is modestly obfuscated so that by default it will boot the multimedia OS unless a sequence of keys are pressed.
At airports I boot the less-secured multimedia OS to show that it's a laptop. Casual inspection shows that it's been used recently and complete.
If someone cares enough to really dig in, notice an encrypted partition, and confiscates my laptop for that reason alone the cost of the laptop is the lower down of my list of concerns. If it happened, I'd probably switch to booting the secure OS fully from removable media like easily hidden flash memory.
If someone wants to threaten me for the secure partition's contents, including lie detectors, drugs, 'enhanced interrogation', etc.. Well that's honestly more than I am concerned with at the electronic level. But if I thought it was a real possibility and worth fighting against, I'd have some tripwire that would self-destruct the data on a particular password (perhaps obfuscated to look like a boot sequence that detects corruption and initiates a disk filesystem check), or not have any data of that importance on a typical laptop drive to begin with.
I was liking both processors up until I found this gem in the article:
If there is egg to be thrown in anyone's face from this article, it is on Intel for its locking down of the Atom platform. Since Computex this year I have been hearing complaints from board vendors on the amount of restrictions Intel is putting on them for Atom products. Vendors are not allowed to build Atom motherboards with PCI Express, digital video outputs or more than one memory slot. VIA on the other hand is openly courting board manufacturers to put as much technology on a mini-ITX design as they can - as long as they DO build one!
Here we see Intel, up to its obnoxious "You'll use our technology only as we prescribe" games. This is the same philosophy that leisurely milked the market for 33Mhz CPU bumps every 6 months, while they sat on years worth of better technology, until AMD lit a fire under their ass.
As good as the software is, there are numerous design decisions that continue to irk many users and 10x as much effort is spent arguing the point rather that just adding in some basic UI code.
Refusal to include a control to adjust the chat font size within the application has to be ridiculous decision #1.
I'm really surprised it hasn't been forked for this reason alone.
I entered in a bug to get a hotly debate option in ( protocol icons ) and it took numerous locked threads, hundreds of messages, and a fantastic amount of vitriol before it was accepted.
That said, aside from this extreme conservatism, Pidgin is good software.
I own an Eee PC and the tab interface really is both simpler and more usable at the same time. Relatives, coworkers, and non-technical friends were able to pick it up for the first time and immediately start using it for their typical internet use. It really is that easy, and the design of the system is much more usable and intuitive than a traditional desktop metaphor for the form-factor of the device.
It seems like your criticism of Linux as 'too technically-oriented' may be true in some cases, but is off-target for the EeePC.
The main complaints were with hardware, not the Linux OS or applications: The keyboard is unusually small and not everyone likes using a trackpad.
I think it's a real tragedy that despite a great demand, utility, and business case for a very low cost laptop like the EeePC 701, there seems to be forces at work to reduce the amount of products like this on the market. Even the newer EeePC models are priced significantly higher.
My current frustration is trying to search for technical documentation online, and discovering that only exists in the form of a video. Text and screenshots would be been a far superior medium. I thought it was bad enough when it happened the first time, but it has happened to me now with two different products. And this article is less useful to me because I do not have the same kind of time to watch videos than I would to scan text.
Gaming on Linux might become a very big deal, if manufacturers can distribute their game as a linux-based liveCD or USB stick, optionally using part of the internal drive for save games/cache/storage.
This has many advantages from a customer support perspective.. Less software environments to test. No antivirus/personal firewalls to kill CPU performance.
This appears to be the biggest obstacle to open source drivers. The current US patent laws have created an environment where any sufficiently complex, or even non-complex piece of software can be construed to violate someone else's bogus software patent.
There's a lot of risk attached to sharing source code. It allows hostile entities a free opportunity to pursue their bogus software patent litigation.
On the positive side, we're starting to see more ways to mitigate this risk, such as creative licensing (GPL3), creative code ownership arrangements, and not but not least pressure to reform the US patent system.
No one should support those particular Linux vendors who assist Microsoft in their efforts to deceptively and in bad faith portray Free Software as illegal. Lenovo - How about some Red Hat or Ubuntu offerings?
On the positive side, one can argue that for a Free Software user it's better to pay for Novell's product than Microsoft's, because at least the hardware is more likely going to be compatible with other, more respectable Linux distributions.
A good step forward, but there is much room for improvement.
In my experience as a custom integrator of dell kit, Dell's premium support leaves much to be desired. Some brief lowlights:
- Dell's kit breaks more often than more vertical integrators. Dell builds systems with a much larger, active set of vendors. This helps them lower initial costs, at a tradeoff to level of confidence in the system as a whole. Especially hard drives.
- Dell (like some other vendors, admittedly) contracts out their field support (and increasingly their internal QA and phone support). That means the technician who shows up doesn't work directly for Dell. Their motive is not to make you, the customer happy. Their motive is to leave your premise as quickly as possible. Their motivation is to maximize profit from Dell, not insure your satisfaction.
- This also means that its difficult to get accountability. You have to go through many layers and multiple companies to fix any problems you have with support work.
- Related to the above points, a startling percentage of the technicians contracted by Dell display a startling lack of basic understanding and expertise. If they have to do more than replace an easy part or run the diagnostic CD, they're over their head. This can be very frustrating when you bump into something like a BIOS issue or a firmware issue with the PERC controllers. It's difficult to get escalation and proper diagnosis.
- There are always ongoing BIOS problems and firmware issues with the PERC storage controller cards.
- It also means that if you're not watching the support technician closely (standing over their shoulder) they will simply replace a part and then leave without even testing if what they did works. If the equipment is at a remote site without 24hr staffing, no amount of signage, verbal communication, or complaining to account managers will insure that the field tech actually informs your company they are "done" and waits for verification before they leave the premise.
It may be that some other vendors are just as bad, but all these things together (and probably others I cannot recall at the moment) make for a poor support experience. The entire feel of the support operation is that of a company attempting to do the bare minimum required by contract, instead of truly standing behind a product with confidence-building reserves of expertise and execution.
The miniseries was amazingly great. It moved the quality bar for TV science fiction more than one notch.
Seasons 1 and 2 were amazingly great. It restored the quality bar for TV drama and deep, immersive storytelling.
Season 3 was fair. It started very strong, but then languished.
The finale last winter was terrible.
It was obvious that deep sacrifices in storytelling quality and production budget were being made to cut costs.
The announcement of the final season is a blessing. It will insure that the series wraps up before a good chunk of the core fanbase becomes alienated. I'm encouraged to know that a strong ending will be the top priority going forward.
ESR's copy_instance defines a class instance's variables/functions based on set of key-value pairs passed in at runtime, where the key is the name of the class function or class variable, and the value is the thing to set it to. You can think of it as a generic, fancy initializer that can define the behavior of class functions/methods as well as typical variables.
This kind of fast-and-loose object manipulation is easy in python. It's built into the language.
But don't feel too bad, the example might not have been written for maximal readability. There's a few things in there that are misleading.
Jabber is a protocol, not an implementation. It's been replaced by the XMPP protocol.
The current releases of one of the most popular implementations of an XMPP server, "ejabberd" supports a dozen different authentication plugins and mechanisms, many of which store passwords only as hashes. Most likely any large scale implementation will use a database authentication backend.
On top of this, modern XMPP clients (including Gaim/Pidgin/Google Talk) encrypt all their communication by default.
--
It's not as easy to find out how passwords are stored for aim/icq/yim/msm, for which the production sources are not available.
I can only speak for free softphones the Linux side.
Ekiga is what I've been using under Fedora Core 5-6 after experimenting with other options. It's an unencumbered SIP client. Make sure to use an up-to-date version. It interoperates well with MS netmeeting. It's works great for personal use.
Most softphones, including the one above, will allow you to choose the audio codec to use for a point to point call. This is a direct tradeoff of bandwidth to quality. You can get a reasonably high quality signal if you have the bandwidth for it. I'd advise experimentation to find the codec that works best with the resources you have available.
There are some serious downsides to VOIP in general:
- The general internet is not 100% reliable. You will experience clipping and dropped calls at some point. You can mitigate this somewhat by configuring your routing equipment at each end to protect and prioritize bandwidth for VOIP.
- There is usually a audio delay by design for buffering. This may be noticable to a third party.
A more professional setup would install a dedicated line between the two premises exclusively for VOIP, making sure that all routers/switches from end-to-end up prioritize and protect VOIP traffic.
There's almost certainly some commercial endpoint hardware just for this situation, with a selection of professional audio-in/out interfaces for hooking up to your gear.
I've experienced something of a culture shock recently, attempting to use and deploy the "DotNetNuke" web framework. Its an IIS+ASP.NET+MSSQL stack, which integrated well with an in-house application. The software has some good press and is open source under a BSD-style license.
There were a number of odd differences in the developer community, as compared to similar types of open source projects I have worked with. Here's some non-comprehensive highlights.
- There wasn't good, free web-based documentation. There was some limited web-based documentation that often repeated the same information readily available on the page you were trying to research. The serious documentation for the project is primarily distributed electronically in finished, non-editable PDF form (which amusingly ship with a license that forbids you from downloading them without written permission) or in video tutorials from other sites with heavy advertisements on their pages. The purpose of putting the documentation in these awkward formats, at least the video tutorials, appeared designed to drive ad revenue or paid subscriptions.
- Every level of add-on was heavily commercialized. Most of the skins available were commercial, even if they were very simple. Most of the modules were commercial, even if they were as basic as modules that allow the site to be backed up. I got the feeling that the hyper-commercialization of add-ons was actually retarding the growth of the core software, because people were protecting the revenue of their simplistic modules, instead of integrating that functionality back into the core.
- To be fair, the software does do what it says it does. The commercial ad-ons do what they say they do. If you shell out for them, you can create a reasonably capable web site quickly, for not too much money.
This kind of "open source" solution exemplifies some of ways that open source software projects can fail to live up to expectations implied by the buzzword. In this case, the core software was under an open source license, but the "complete" solution for most users was not, and there was a culture in place that made it difficult to make the open-source part of the software into a complete solution.
I did not see the difference so clearly before, because most of the BSD-licensed products in the Linux community have more of sharing culture, and perhaps do not feel as strongly the need to recoup the costs of their development tools:D
It is too bad Eric is using his celebrity in such a negative way. He could have focused on what he likes about other distributions, rather than what he doesn't like elsewhere.
For myself, many of the reasons that Eric lists for not liking Fedora, are exactly what I enjoy and want to see more of in a linux distribution.
I wonder if there's a way to twist this scenario again, by Novel setting up a mechanism whereby everyone, everywhere can be legally considered their customer without restriction.
I have used all of the major distributions, including every version of Fedora Core except FC2.
I think Fedora is Great! It's my favorite distribution for playing with bleeding edge developments smoothly and easily (aside from KDE).
There are a few things I keep in mind:
- If you are making a desktop system for personal use, create a home partition. That way you can blow away your OS partition for future updates if needed, without losing any important data.
- The released version without updates will be buggy. I will often wait until a month after release, then install and immediately run an update. For example, Fedora Core 5 out of the box was almost unusable for what I needed, but Fedora Core 5 with updates may be my slickest and most productive linux desktop environment.
- You have to install third-party packages for anything of dubious US legality, such as mp3 codecs and dvd support. However, there are a few sites (freshrpms, livna) that make this simple and painless. Personally I like that Fedora draws the line where they do, and keeps a clean distribution.
- Explore the distribution, extras, and third party repositories. You have more to gain if you attempt to use what they provide instead of try and shoehorn in something on your own. Also, upgrading to future Fedora Core revisions will be easier.
Live Long and Prosper on the gleaming but sharp brim of Fedora!
Slashdot Mirror
I've been in this exact situation recently. I regularly go back and forth between macbook pros, microsoft-oriented work laptops, and microsoft-oriented gaming laptops, and hacking-project chromebooks.
I am sure there are exceptions anyone can point to but as a general observation data point from someone who has "been there":
Apple laptops are a known, very high quality, quantity. They don't always have the features you want-- top-of-the-market GPU capability and larger screens is mine, but what Apple does have, they do very well. If apple laptops do deliver everything you really need, then get one, even if you put VMware/Parallels/Other/Bootcamp of it to get solid MS compatibility for everything but high end games. Even if you have to put up with silly dongles.
High-end PC laptops have very good specs but, in general, come with more risks. The more cutting edge features you buy the less likely it is you're going to get good, long term support, and the more likely that you're going to develop some issue a few years out that never showed up in the manufacturer's time-limited testing. The one exception to this may be the MS Surface Pros.
In the midrange prices from reputable manufacturers, support is more predictable, but there's often some performance compromise being made-- you can discover too late that your great price came at the cost of a certain component that's dragging the rest of the system down. Could be wrong CPU for your workload, poor cooling, sub-par screen, KB placement, loud fans, slow HD, inadequate GPU cooling, etc.. It's hard to know what compromise may affect you without doing a lot of review research, and the best, most reliable reviews will be for product that has been out 6+ months, and you may be tempted to get "the latest new model" instead. It's a balancing act.
At the low end you can find some really great values in PCs laptop. If these fit your performance envelope you can get something that's cheap and looks good too.
The above is just on the hardware side. On the software side it goes without saying that there will be a learning curve and a host of annoyances, particularly around driver behavior coming out of sleep mode & hibernation, privacy, and possibly juggling multiple possible audio inputs and outputs.
Great comment, at least in the sense that it resonates with me.
I want to like KDE, because when you get under the hood the Qt framework and many related components are truly well designed, mature software.
But on the visuals side, there's something off. It's not in the tech, it's in the visual polish and consistency, color pallet, icon set, styling. Yes, I know it's customizable but so far, the out of the distro kde-styling seems off to me personally once you open a few different apps and settings panels and see them together; it's hard to pin down.
Despite this unease, I have been generally put off in the design decisions of the GNOME camp in recent years, in particular the intentional removal of optional functionality compared to previous versions, and getting away from a traditional desktop metaphor in favor or something that's more like a smartphone. I don't want a smart phone interface on a high resolution, multi-monitor setup with KB and mouse at the ready.
Cinnamon / XFCE is working for me as a compromise, but I'd love it KDE is able to storm back in with great style, to rescue us from desktop redsign madness.
I've been running crouton on a toshiba chromebook for a couple years now and I use it with surprising frequency.
I spend most of my time on the linux side, running terminals, vpns, and some custom chat and web apps. It's not my main system but I frequently use it as a supplement.
The downside of the config is the annoying/slow "developer mode" bios warning on reboot that you get from unlocking the chromebook bios, and the 3-4 commands you have to run after each reboot to get crouton up and running and happy. Luckily you don't need to reboot often. My average is about once every two months because I've let the battery run down too far. I'm very happy with it especially for the price.
If your passwords are randomly generated and long, it doesn't matter how many attempts to guess them are tried. The likelyhood of a random guess getting through are lower than your chances of winning the lottery. Let people waste their time on futile attempts.
To further decrease your chances, use public keys authentication instead of passwords, or two factor authentication, or limit connections by IP address.
Changing the post does fool most SSH scanners as well.
I don't like fail2ban because it can lead to DoS vectors.
Bottom line is that logged attacks that have no hope of getting through shouldn't cause a panic.
I have an Apple laptop (more like, portable workstation) and I bought it after numerous computer-generations of all kinds of PC laptops, some quite expensive and focused on gaming/performance. I've had it for a year now and I can say that it is the *only* laptop I've ever owned where I've been completely satisfied with the build and service quality. Having a top-flight desktop with an uncompromising unix shell is quite nice too. For gaming I dual boot.
BTW, for a more mainstream data point, the Apple laptops swept Consumer Reports "most recommended buy" in multiple categories recently.
Despite being from a "closed" company, it gives me a platform that lets me natively run Linux, Windows, and MacOSX. It offers more choices. Development tools are much easier to come by as well.
What is to stop someone from installing proxy or NAT software onto their (perhaps jailbroken) smartphone? Can cell providers really prevent this?
I've done this recently, not just for travelling, but for using networked, 3D accelerated games with rich multimedia that cannot enjoyably run in linux or even played without proprietary kernel drivers outside of a VM sandbox, and which blow normal security practices to bits.
What I did is create a normal linux partition, locked down but still highly usable for multimedia, gaming, and typical virtual machine usage.
Then on one of the internal partitions I created a second, entirely encrypted root partition for a second OS using LUKS. This partition is booted by connecting a USB key and booting from it instead of the normal internal MBR, then entering passwords. The second OS is reasonably secure, locked down, much more limited in functionality, and contains tools to audit the integrity of the multimedia OS and virtual machines, as well as backup and restore them. The USB key is modestly obfuscated so that by default it will boot the multimedia OS unless a sequence of keys are pressed.
At airports I boot the less-secured multimedia OS to show that it's a laptop. Casual inspection shows that it's been used recently and complete.
If someone cares enough to really dig in, notice an encrypted partition, and confiscates my laptop for that reason alone the cost of the laptop is the lower down of my list of concerns. If it happened, I'd probably switch to booting the secure OS fully from removable media like easily hidden flash memory.
If someone wants to threaten me for the secure partition's contents, including lie detectors, drugs, 'enhanced interrogation', etc.. Well that's honestly more than I am concerned with at the electronic level. But if I thought it was a real possibility and worth fighting against, I'd have some tripwire that would self-destruct the data on a particular password (perhaps obfuscated to look like a boot sequence that detects corruption and initiates a disk filesystem check), or not have any data of that importance on a typical laptop drive to begin with.
I was liking both processors up until I found this gem in the article:
Here we see Intel, up to its obnoxious "You'll use our technology only as we prescribe" games. This is the same philosophy that leisurely milked the market for 33Mhz CPU bumps every 6 months, while they sat on years worth of better technology, until AMD lit a fire under their ass.
Don't be fooled again.
As good as the software is, there are numerous design decisions that continue to irk many users and 10x as much effort is spent arguing the point rather that just adding in some basic UI code.
Refusal to include a control to adjust the chat font size within the application has to be ridiculous decision #1.
I'm really surprised it hasn't been forked for this reason alone.
I entered in a bug to get a hotly debate option in ( protocol icons ) and it took numerous locked threads, hundreds of messages, and a fantastic amount of vitriol before it was accepted.
That said, aside from this extreme conservatism, Pidgin is good software.
I own an Eee PC and the tab interface really is both simpler and more usable at the same time. Relatives, coworkers, and non-technical friends were able to pick it up for the first time and immediately start using it for their typical internet use. It really is that easy, and the design of the system is much more usable and intuitive than a traditional desktop metaphor for the form-factor of the device.
It seems like your criticism of Linux as 'too technically-oriented' may be true in some cases, but is off-target for the EeePC.
The main complaints were with hardware, not the Linux OS or applications: The keyboard is unusually small and not everyone likes using a trackpad.
I think it's a real tragedy that despite a great demand, utility, and business case for a very low cost laptop like the EeePC 701, there seems to be forces at work to reduce the amount of products like this on the market. Even the newer EeePC models are priced significantly higher.
I agree 100%
My current frustration is trying to search for technical documentation online, and discovering that only exists in the form of a video. Text and screenshots would be been a far superior medium. I thought it was bad enough when it happened the first time, but it has happened to me now with two different products. And this article is less useful to me because I do not have the same kind of time to watch videos than I would to scan text.
Text please!
Bzip2 consumes far more memory and CPU cycles than gzip. There's a lot of scenarios where this tradeoff isn't desirable for a busy webserver.
Gaming on Linux might become a very big deal, if manufacturers can distribute their game as a linux-based liveCD or USB stick, optionally using part of the internal drive for save games/cache/storage.
This has many advantages from a customer support perspective.. Less software environments to test. No antivirus/personal firewalls to kill CPU performance.
This appears to be the biggest obstacle to open source drivers. The current US patent laws have created an environment where any sufficiently complex, or even non-complex piece of software can be construed to violate someone else's bogus software patent.
There's a lot of risk attached to sharing source code. It allows hostile entities a free opportunity to pursue their bogus software patent litigation.
On the positive side, we're starting to see more ways to mitigate this risk, such as creative licensing (GPL3), creative code ownership arrangements, and not but not least pressure to reform the US patent system.
No one should support those particular Linux vendors who assist Microsoft in their efforts to deceptively and in bad faith portray Free Software as illegal. Lenovo - How about some Red Hat or Ubuntu offerings?
On the positive side, one can argue that for a Free Software user it's better to pay for Novell's product than Microsoft's, because at least the hardware is more likely going to be compatible with other, more respectable Linux distributions.
A good step forward, but there is much room for improvement.
In my experience as a custom integrator of dell kit, Dell's premium support leaves much to be desired. Some brief lowlights:
- Dell's kit breaks more often than more vertical integrators. Dell builds systems with a much larger, active set of vendors. This helps them lower initial costs, at a tradeoff to level of confidence in the system as a whole. Especially hard drives.
- Dell (like some other vendors, admittedly) contracts out their field support (and increasingly their internal QA and phone support). That means the technician who shows up doesn't work directly for Dell. Their motive is not to make you, the customer happy. Their motive is to leave your premise as quickly as possible. Their motivation is to maximize profit from Dell, not insure your satisfaction.
- This also means that its difficult to get accountability. You have to go through many layers and multiple companies to fix any problems you have with support work.
- Related to the above points, a startling percentage of the technicians contracted by Dell display a startling lack of basic understanding and expertise. If they have to do more than replace an easy part or run the diagnostic CD, they're over their head. This can be very frustrating when you bump into something like a BIOS issue or a firmware issue with the PERC controllers. It's difficult to get escalation and proper diagnosis.
- There are always ongoing BIOS problems and firmware issues with the PERC storage controller cards.
- It also means that if you're not watching the support technician closely (standing over their shoulder) they will simply replace a part and then leave without even testing if what they did works. If the equipment is at a remote site without 24hr staffing, no amount of signage, verbal communication, or complaining to account managers will insure that the field tech actually informs your company they are "done" and waits for verification before they leave the premise.
It may be that some other vendors are just as bad, but all these things together (and probably others I cannot recall at the moment) make for a poor support experience. The entire feel of the support operation is that of a company attempting to do the bare minimum required by contract, instead of truly standing behind a product with confidence-building reserves of expertise and execution.
There should be better options.
- The miniseries was amazingly great. It moved the quality bar for TV science fiction more than one notch.
- Seasons 1 and 2 were amazingly great. It restored the quality bar for TV drama and deep, immersive storytelling.
- Season 3 was fair. It started very strong, but then languished.
The finale last winter was terrible.
It was obvious that deep sacrifices in storytelling quality and production budget were being made to cut costs. The announcement of the final season is a blessing. It will insure that the series wraps up before a good chunk of the core fanbase becomes alienated. I'm encouraged to know that a strong ending will be the top priority going forward.ESR's copy_instance defines a class instance's variables/functions based on set of key-value pairs passed in at runtime, where the key is the name of the class function or class variable, and the value is the thing to set it to. You can think of it as a generic, fancy initializer that can define the behavior of class functions/methods as well as typical variables.
This kind of fast-and-loose object manipulation is easy in python. It's built into the language.
But don't feel too bad, the example might not have been written for maximal readability. There's a few things in there that are misleading.
Jabber is a protocol, not an implementation. It's been replaced by the XMPP protocol.
The current releases of one of the most popular implementations of an XMPP server, "ejabberd" supports a dozen different authentication plugins and mechanisms, many of which store passwords only as hashes. Most likely any large scale implementation will use a database authentication backend.
On top of this, modern XMPP clients (including Gaim/Pidgin/Google Talk) encrypt all their communication by default.
--
It's not as easy to find out how passwords are stored for aim/icq/yim/msm, for which the production sources are not available.
I am very glad to see this positive step. When RMS and Linus are aligned, great things happen.
I can only speak for free softphones the Linux side.
Ekiga is what I've been using under Fedora Core 5-6 after experimenting with other options. It's an unencumbered SIP client. Make sure to use an up-to-date version. It interoperates well with MS netmeeting. It's works great for personal use.
Most softphones, including the one above, will allow you to choose the audio codec to use for a point to point call. This is a direct tradeoff of bandwidth to quality. You can get a reasonably high quality signal if you have the bandwidth for it. I'd advise experimentation to find the codec that works best with the resources you have available.
There are some serious downsides to VOIP in general:
- The general internet is not 100% reliable. You will experience clipping and dropped calls at some point. You can mitigate this somewhat by configuring your routing equipment at each end to protect and prioritize bandwidth for VOIP.
- There is usually a audio delay by design for buffering. This may be noticable to a third party.
A more professional setup would install a dedicated line between the two premises exclusively for VOIP, making sure that all routers/switches from end-to-end up prioritize and protect VOIP traffic.
There's almost certainly some commercial endpoint hardware just for this situation, with a selection of professional audio-in/out interfaces for hooking up to your gear.
I've experienced something of a culture shock recently, attempting to use and deploy the "DotNetNuke" web framework. Its an IIS+ASP.NET+MSSQL stack, which integrated well with an in-house application. The software has some good press and is open source under a BSD-style license.
:D
There were a number of odd differences in the developer community, as compared to similar types of open source projects I have worked with. Here's some non-comprehensive highlights.
- There wasn't good, free web-based documentation. There was some limited web-based documentation that often repeated the same information readily available on the page you were trying to research. The serious documentation for the project is primarily distributed electronically in finished, non-editable PDF form (which amusingly ship with a license that forbids you from downloading them without written permission) or in video tutorials from other sites with heavy advertisements on their pages. The purpose of putting the documentation in these awkward formats, at least the video tutorials, appeared designed to drive ad revenue or paid subscriptions.
- Every level of add-on was heavily commercialized. Most of the skins available were commercial, even if they were very simple. Most of the modules were commercial, even if they were as basic as modules that allow the site to be backed up. I got the feeling that the hyper-commercialization of add-ons was actually retarding the growth of the core software, because people were protecting the revenue of their simplistic modules, instead of integrating that functionality back into the core.
- To be fair, the software does do what it says it does. The commercial ad-ons do what they say they do. If you shell out for them, you can create a reasonably capable web site quickly, for not too much money.
This kind of "open source" solution exemplifies some of ways that open source software projects can fail to live up to expectations implied by the buzzword. In this case, the core software was under an open source license, but the "complete" solution for most users was not, and there was a culture in place that made it difficult to make the open-source part of the software into a complete solution.
I did not see the difference so clearly before, because most of the BSD-licensed products in the Linux community have more of sharing culture, and perhaps do not feel as strongly the need to recoup the costs of their development tools
It is too bad Eric is using his celebrity in such a negative way. He could have focused on what he likes about other distributions, rather than what he doesn't like elsewhere.
For myself, many of the reasons that Eric lists for not liking Fedora, are exactly what I enjoy and want to see more of in a linux distribution.
Long live Linux distributions!
I wonder if there's a way to twist this scenario again, by Novel setting up a mechanism whereby everyone, everywhere can be legally considered their customer without restriction.
I have used all of the major distributions, including every version of Fedora Core except FC2.
I think Fedora is Great! It's my favorite distribution for playing with bleeding edge developments smoothly and easily (aside from KDE).
There are a few things I keep in mind:
- If you are making a desktop system for personal use, create a home partition. That way you can blow away your OS partition for future updates if needed, without losing any important data.
- The released version without updates will be buggy. I will often wait until a month after release, then install and immediately run an update. For example, Fedora Core 5 out of the box was almost unusable for what I needed, but Fedora Core 5 with updates may be my slickest and most productive linux desktop environment.
- You have to install third-party packages for anything of dubious US legality, such as mp3 codecs and dvd support. However, there are a few sites (freshrpms, livna) that make this simple and painless. Personally I like that Fedora draws the line where they do, and keeps a clean distribution.
- Explore the distribution, extras, and third party repositories. You have more to gain if you attempt to use what they provide instead of try and shoehorn in something on your own. Also, upgrading to future Fedora Core revisions will be easier.
Live Long and Prosper on the gleaming but sharp brim of Fedora!