http://www.nytimes.com/2011/08/29/business/media/in-times-of-unrest-social-networks-can-be-a-distraction.html?_r=1

http://www.nytimes.com/2012/01/01/opinion/sunday/the-joy-of-quiet.html?pagewanted=1&_r=1&hp

But even in the world of information, this approach has costs.  For example, there is a a hefty bias in the open source community in favor of a very narrow kind of power: technical wizardry, association with powerful techno-business interests, etc.  In some senses, this bias stems directly from the mandate for each individual in the “Linux community” to extract maximum individual value from the community – either in substantive terms, or in terms of garnering increased authority and respect (social capital).  That is, the “open source community” is something of a hyper-competitive meritocracy, with different individuals (or teams) each working to amass the greatest possible shares of technological achievement – the only kind of authority universally recognized by the “community” at large.  Thus, the core value of the open source “community” is distinctly anti-social (ie, vehemently, obsessively technical) in nature, and that fact sets the tone for much of what goes on within it.

And so voices with less technically-oriented priorities (eg, user education; accessibility; social responsibility etc.) are often ignored and discarded, left to either break off to form their own splinter groups (Fedora, Ubuntu, and LibreOffice being the most notable examples) or be quietly suffocated by disinterest.  As a result, unfathomable resources are wasted on parellel but independent efforts, undertaken by various splinter groups who couldn’t quite reconcile their differences, to accomplish almost exactly the same ends.  The guiding principle on which the entire Linux “community” is based is that if something isn’t quite right, it is the responsibility of the individual(s) who find it so to make the necessary corrective adjustments.  And so the entire “community” is maintained primarily by means of powerful individual-centered enticements: people give back, when they do, either as an accidental by-product of pursuing their own agendas, or to increase their own of fame, reputation, and authority within the community.   So if the Linux community at large doesn’t seem to be entirely comfortable embracing any kind of a broader social vision, that’s because it isn’t, at least not intrinsically.

After all, what happens when there is more than one metric of influence and worth in a community?  When people who have vastly different priorities must live and work side-by-side, and do not have the space to fork away from each other because of minor differences in philosophies?  When the resources needed to support such forks are physical rather than informational, and must come from a finite pool, shared by all?

The open-source movement has some intrinsically libertarian characteristics, it would seem, and while some of these – an emphasis on personal responsibility, democracy, egalitarianism, etc. – may be positive, they cannot be separated from their accompanying shadows: deprioritization of social responsibility, the rise of single-factor meritocracy, and the failure to recognize, much less address, pre-existing inequalities (in education, access to resources, and social status within the community) that place potential entrants into a given community on distinctly unequal footing.

Consider the following observation taken from an excellent Ars article on the “app-store” model of software governance, which presents the converse of the argument being made here:

Indeed, from the perspective of the ordinary user, the PC software market may not seem especially free. Lacking the sophistication to distinguish good software from bad, many users don’t feel free to install software at all. In a sense, a curated app store actually increases the freedom of the typical user by enabling him to buy software without the help of his IT-expert sister-in-law.

In other words, because the so-called curated computing philosophy of software governance takes into account and adjusts for differing levels of user competence, it actually gives users on the lower end of the technical spectrum much greater choice and autonomy than does the laissez-faire model of governance, by providing some basic level of user education and support, and safeguards to prevent users from doing something to harm themselves.  Admittedly, all this comes at something of a cost to the most advanced users, who would rather have unfettered control over every aspect of their own “user experience.” But if the goal of the open source community is actually to create a completely egalitarian information society (see Barlow), why shouldn’t this sacrifice be seen as necessary and proper?  The fact is, of course, that it isn’t, because the open source philosophy as a whole is not based on any unifying social ideal – save the dubious deification of individual autonomy.

The overall point, I suppose, is simply that governance is complex and nuanced.  Despotism is not universally evil, and democracy is not universally good.  Hopefully we’ll eventually be able to figure out better ways to incorporate the best elements of both, while discarding the worst.

Several  environment  variables  are  set  up automatically by the cron(8) daemon.  SHELL is set to /bin/sh, and LOGNAME and HOME are set from the /etc/passwd line of the crontab’s owner. PATH is set to “/usr/bin:/bin”

What this means is that even if a certain executable runs just fine from your own command prompt, cron may not know where to find it, because most user shells have PATHs far more extensive than the bare-bones default relied on by cron.  And if cron cannot find an executable, it simply won’t run the task in question.   There are two solutions to this problem.  First, you can tell cron to check all of the paths that are active in your own shell session (issue: echo $PATH at a terminal, then put the results of that command on its own line at the top of your crontab file, like so:

PATH = <what you just got from echo $PATH>

Alternatively, you can just make sure to always use absolute pathnames for every exectuable you invoke in your crontab.  To find the exact location of an executable, issue whereis <simple name of executable>.

Should you run into other problems getting your cron jobs to run (which is likely, given that any flaw in syntax anywhere in a crontab line can result in the entire line silently failing to execute), you can use output and error redirection to force the line in question to dump more information about why it’s failing to a specified location (basically, you’re creating your own mini-logging system).  Set up such a log, and then try running the troublesome task every minute (* * * * *) until you get it to work.

In response to Amazon’s behavior, I have begun looking for alternative shopping hubs that are at once more socially responsible, but also do not ask me to sacrifice too much in the way of selection, convenience, or savings.  Believe it or not, I’ve had a fair amount of success, and have managed to stay almost completely Amazon-free throughout the holiday season.  Here are a few of the sites I’ve uncovered, to which I’ll be returning frequently until Amazon decides to clean up its act:

Better World Books (Books)

This site is simply extraordinary, matching or exceeding Amazon on nearly every practical metric (selection, pricing, speed, etc.) while also expressing a firm committement to making the world a better place in multiple ways.  As the website puts it:

Better World Books uses the power of business to change the world. We collect and sell books online to donate books and fund literacy initiatives worldwide. With more than 8 million new and used titles in stock, we’re a self-sustaining, triple-bottom-line company that creates social, economic and environmental value for all our stakeholders.

Wayfair (Home furnishings):

The socially-beneficial bona fides of this group aren’t quite as obvious, but they at least seem to genuinely care about their workers’ well-being – and their workers, in turn, seem to be a pretty socially-conscious lot.

I’ll post more as I find them!

the shopping center of the Internet

This saga clearly has countless interesting and troubling implications.

Among the most prominent:

• As the Ars article points out, it would seem that the current uproar over PROTECT IP and SOPA may be a distraction from the fact that the U.S. government already permits many of the interventions spelled out in those bills.  If opponents of the bills are right that several of their provisions are blatantly unconstitutional (violating the First Amendment, Due Process clause, etc.), then the bills themselves are in some ways far less troubling than what the U.S. legal system is in fact already doing, generally without any substantial public knowledge.  Indeed, the bills may even prove to be somewhat beneficial, in that they make clear and public the rules under which cases are already being decided, and thereby allow the public to voice its profound discontent with those rules.

• The blanket orders issued by the court in this case further testify to the extent to which the U.S. legal system is disturbingly ill-prepared to handle the complexities of applying existing law to the digital realm.  For even if the intent of the Court’s order in this case were reasonable, the broadness and of its scope and general absolutism would still be absurd.  Any order that addresses “all social media sites” and “all search engines” simply cannot be taken seriously (much less implemented) – particularly, as the Ars article points out, given that many such sites are completely outside of the purview of U.S. law.

• The U.S. puts itself on increasingly unsteady ground in its condemnation of state-sponsored Internet censorship elsewhere when its own federal courts issue edicts to sites requiring that they “de-index and/or remove [domain names] from any search results pages.” This sounds eerily to the sort of blacklisting approach used by most authoritarian governments around the globe to restrict public access to sites that are incompatible with reining ideologies – an approach which the U.S. government ostensibly condemns, and is actively funding projects to circumvent.

In particular, I’m deeply troubled at the “Facebook Privacy” setting that now appears in the “Privacy” section of the Pandora web-app.  I may be mistaken, but I was under the impression that prior to this recent interface overhaul, several steps were required to explicitly link a Pandora account to a Facebook profile.  Those several steps seem to have at some point been condensed into a couple of deceptively simple checkboxes – which apparently someone deemed it appropriate to have automatically checked by default.  Given what I know of web technologies, that means that anyone who was using Pandora before this change was made and happens to stay logged into facebook most of the time (ie, a lot of people) automatically started having their music preferences shared on Facebook without ANY prior warning or consent.  Furthermore, given the somewhat vague descriptions associated with the checkboxes, and the fact that Facebook javascript is active any time I visit the Pandora site, I have very little confidence that Facebook isn’t getting huge amounts of data about me even though I have all of the checkboxes unchecked.  This is completely unacceptable.

To Pandora’s credit, one of the reasons why I am so distressed by the recent changes is that I used to consider Pandora to be one of the very few examples of a site that engaged in “responsible tracking,” in that you folks made it quite clear that Pandora users could choose to pay for the service EITHER by viewing targeted ads and being subject to extensive behavioral tracking, OR by paying a reasonable annual subscription fee.  This new interface move seems to me to be evidence that the distinction between paying and non-paying users may not be quite as clear as it was, in as much you are trying to “nudge” ALL of your users into sharing as much information as possible with either you or your corporate partner.  There is a reason that I chose to pay you $36 per year, and it isn’t just that I thought you had an idea worth supporting (which I did, and still do).  I deeply value my online privacy, and will go to great lengths to protect it.  And any attempt to subvert or second-guess my decision on that front is very, very likely to make me want to stop using the site that makes that move.

In case it isn’t clear what I’m asking for here, my request is simply this: that you make it unequivocally clear to your users when you are sharing their data with facebook, and when you aren’t (a little “Facebook link active” icon in the corner of the screen somewhere might be nice).  And for those users who don’t want their information shared with facebook (ESPECIALLY those users who are Pandora One subscribers), I’d ask that you fully respect that decision, and completely obliterate all of Facebook’s javascript and other script hooks from the version of Pandora that you serve up to these users.

Thank you for taking the time to read this.

As someone who is fairly new to the Mac OS GUI, I was flummoxed and peeved by the fact that <Alt><Tab> cycles only through each of the open applications on any given system – not the individual windows of each of those applications (the window you get when you toggle from application X to Y is always the one you were working on last within application Y).  Apparently, move among windows within application Y, one must instead use <Alt><~> (Yes, that’s a tilda).

Many thanks to Apple for unnecessarily tweaking yet another nearly-universal standard for the purpose of “enhancing user experience”

But suppose that instead of hooking itself into the Startup section of the registry, a particularly clever species of malware instead hooked itself into the executable for another, perfectly valid executable on they system.  Better yet, suppose that this fiend hooked itself in the generic execution routine for every single .exe file on a given system.

I ran into just such a phenomenon during my latest malware removal effort: instead of putting tell-tale (and relatively easy-to-remove) references into Startup folders, this worm modified the file association keys for the .exe file extension, thereby telling the computer to silently invoke (or reinvoke) an evil binary every time any executable file on the system was run.  Of course, this technique is ultimately just as effective as the Run-At-Startup hijack – but has the added bonus of being much, much harder to identify and remedy.

The solution is actually fairly simple, if you know what to do: simply get a clean copy of the pertinent registry keys, and merge the clean copy into the infected registry (the merge will overwrite all existing keys, and remove the evil references).  Happily, because Windows Vista seems to be exceedingly good at corrupting its own .exe associations without any external help, some kind souls have posted clean versions of these keys (along with those for several other file types).  I’ve used this particular .reg file repeatedly, and vouch that it works as expected.

Don’t forget to go find and delete the actual executable files that are invoked by the poisoned .exe keys; there’s a chance that those files are being invoked in other ways as well, and in any case you generally don’t want to make a habit of keeping bits of malicious code in random corners of your file system.

SEO keywords:

.exe files, file associations, programs cannot run

The short version (for experts)

By “expert”, I simply mean that you should have a basic familiarity with Linux commands, and understand the dd command enough to know that it is extremely dangerous when used incorrectly. If you end up destroying the contents of your hard drive as a result of not having done enough background reading before trying these procedures, I am not to blame (though I will most certainly sympathize).

A couple of hours of googling should get you up to speed on both of these points; I’ll also be writing a more novice-oriented version of this guide in the future.

Assuming you’re comfortable with all this, let’s proceed:

Copying entire partitions with one fell swoop from…

A good hard disk to a (presumably larger) disk

1. Boot into the Linux Live-CD of your choosing, with your old and new drives attached to the active system in some way (internally or via USB interfaces).
2. Launch gparted, or another disk partitioning utility
3. On the new/larger drive, create a partition of exactly the same size as the one you’re going to be copying from.
4. Find out the name of the Linux block device associated with the new partition, as well as the old one from which you plan to copy (both will probably take the form /dev/sd**).  (In case you didn’t know, all of those funny-looking files in the /dev directory are known as block devices, and are actually representations – generally – of hardware devices.  Such as partitions.)
5. Use the dd command to copy the contents of the old partition to the new one, using a fairly large bs setting (4096 often works well).  Be absolutely sure that the “if” parameter is set to the source partition, and the “of” parameter to the destination. The command you’ll issue should look something like this:  dd if =/dev/ of=/dev/[destination] bs=4096.
6. Repeat this process for any other partitions you wish to copy.
7. If you’re trying to copy an entire drive, and want the new drive to “just work” the same way the old one did (ie, be bootable), you’ll also need to copy over the bootsectors of the old hard drive’s Master Boot Record (MBR), which are always stored in the first 446 bytes of any given disk.  Instructions for doing this can be found here.  I’m not going to give you the exact syntax here, because if you’re familiar with the dd command you’ll already know what it is, and if you’re not familiar with the dd command you really, REALLY need to read the warnings and disclaimers on the page linked above.

A good hard disk to a raw or compressed image

1. Find the name of the Linux block device associated with the partition you want to make an image of (gparted often helps with this).
2. Choose a destination drive that has enough space to accommodate a full, byte-by-byte image of the partition you plan to image.  If you plan to make a compressed image, assume you’ll need about 1/2 – 3/4 of the space that your source partition is currently taking up (you often won’t, but play it safe).  DO NOT try to save an image to a drive formatted as FAT32 – FAT32 has a theoretical 4GB limit on file sizes.  Any other modern filesystem should do nicely, however (ext3/4; xfs; ntfs (shudder), etc.)
3. To make a simple, byte-by-byte image, simply invoke dd with the “of” parameter set to the name of a (new) file on the drive where you want to save the image, and the bs parameter set to something fairly large (4096 often works well).  Thus, if you’re trying to save the image to the drive mounted at /media/backup, you’d issue something like this: dd if=/dev/

   of=/media/backup/myimage.dd bs=4096</em></strong>.</li>
   <li>To make a compressed image, you’ll need to<strong> <em>pipe</em></strong> the output of the dd command through a compression utility – either bzip2 or gzip.  I’d personally recommend bzip2, since it’s supposed to yield higher compression rations.  The command you’ll issue will look something like this : <em><em><strong>dd if=/dev/

   bs=4096 | bzip2 -cz9 > /media/backup/myimage.dd.bz. The  -cz9 switches tell bzip2 to (1) compress to standard output; (2) to compress whatever it’s being given; and (3) to try achieve the maximum compression ration possible – in that order. 

4. To restore your image, simply invert the command you issued before: dd if=/media/backup/myimage.dd of=/dev/[destination partition] bs=4096 for uncompressed images; and bzip2 -dc /media/backup/myimage.dd.bz | dd of=/dev/[destination partition] for uncompressed ones.
5. If the mood strikes you, you can also view the contents of a given image without restoring it to a partition by mounting it as loop volume, like so: mount -t [filesystem type; you'll have remember this] /media/backup/myimage.dd [any empty directory] -o loop.

A hard disk that has bad sectors

The procedures for making copies and/or images of disks that are going bad are nearly the same as those outline above – the one difference being that wherever the aforementioned instructions tell you to use dd, you’ll need to use a utility called ddrescue instead.  ddrescue comes pre-installed on Knoppix, but you’ll need to apt-get it under Ubuntu (enable universe and multiverse repositories; then issue sudo apt-get install ddrescue).  The syntax of ddrescue is a bit different from dd, and arguably even easier to screw up (unlike with dd, the order of parameters matters), so read the man page for the utility carefully before trying to use it.

UNDER CONSTRUCTION

The longer version (with explanations, warnings, and random asides):

As long as you’re fairly confident that the drive you wish to copy from doesn’t have any physical defects, you can opt to use a data-transfer method which dispenses with certain safeguards – but which in return moves data very,VERY quickly.  The core component of this approach is the dd command – a supremely powerful utility that comes natively with almost every linux distribution.

Before I proceed to the step-by-step how to, I must implore, demand, and beg that you spend a few minutes reading up on the dd command – either online, or in the utility’s man page (type  man dd in a linux terminal window), or right here, before even thinking about using it.  For though the name “dd” was originally intended to be a shorthand for something like “Data Duplicator”, it has come to mean something far more sinister to those who accidentally misuse it: the “Disk Destroyer.”  The reason for this less-than-affectionate reinterpretation is that dd is, to but it bluntly, incredibly stupid: it does precisely what you tell it to do, even if what you’re telling it to do makes absolutely no sense, and is guaranteed to destroy the date on your hard drive.  To show you what I mean, I’ll first have to introduce the basic syntax of the command.  To copy the contents of the first partition on the first hard drive to the first partition on a second drive, you would enter something like this:

dd if=/dev/sda1 of=/dev/sdb1

where a and b represent the order of the drives, and the numbers at the end of the “sd*” represent the partitions you’ll be dealing with on each drive (these are basic linux conventions – if you’re unfamiliar with them, take at least a half-an-hour to figure out how they work before doing anything with dd).  The “of” in the command stands for “output file”; similarly, the “if” stands for “input file” (in linux-land, anything can be a “file,” including an entire hard drive partition).

If you’re thinking that the /dev/blahblah strings look awfully similar, or that “if” and “of” are only one character different, or that it generally looks fairly easy to mess up the order of this command somehow, you’d be right – and that’s precisely the reason dd has earned such an unappealing nickname.  For suppose what would happen if you were trying to copy data from a full hard drive A to an empty hard drive B, and inadvertently entered the following command:

dd if=/dev/sdb1 of=/dev/sda1

Remember when I said dd does exactly what you tell it to do?  I wasn’t exaggerating.  If you were to enter this command, dd would dutifully go and copy complete nothingness, one byte at a time, from the empty drive to the full one.  arise when you don’t match the correct input/output specifier with the right partitions.

the following command built-in to all linux distros). EXTREMELY POWERFUL, BUT EXTREMELY DANGEROUS. Only reliable method I’ve found for transferring disks’ MBRs and boot sectors – which it does flawlessly. Byte-level copying
Making images (which can be compressed, if you’d like): the dd command provides this functionality, and unix piping allows you to compress the “images” dd creates, but only the most intrepid will probably opt to dive into the messy command-line gymnastics that this approach requires. An an excellent pre-packaged alternative can be found in the form of PartImage, which can be run either by itself from a dedicated live-CD (System Rescue CD, I believe), or from within a full-blown linux environment, such as Ubuntu.
ntfsprogs: ntfscopy (I think)
Specialized versions of dd designed for data recovery

Gparted

Time and time again, I’ve been amazed at how much a simple update to a single system driver has improved my overall quality-of-computing.  Updating my iwl3945 wireless networking driver from linuxwireless.org caused my periodic problems with random kernel panics to all but vanish; later, updating it again fixed some odd power problems with my antenna.  Repeatedly applying the latest graphics drivers from the official x-updates and xorg-edgers repositories first fixed my dual-monitor setup, then fixed the destruction of the graphics stack performance issues that accompanied the official 9.04 release.  And manually installing the 2.6.31 kernel seems to have banished my power management (suspend and resume) woes, allowing me to avoid having to reboot my laptop for obscenely long periods of time on end (current uptime: 8 days, 21 hours).

Granted, installing updates is hardly a particularly novel or ground-breaking idea, and doing so certainly doesn’t require much in the way of advanced technical skill.  But for those who don’t know, it’s important to understand that critical system files (including drivers, and obviously the kernel) are undergoing active development all the time – but updates to such files generally aren’t pushed out to users between official release cycles (every six months in the case of Ubuntu).  Consequently, if you’re running into a bizarre pattern of behavior that you can’t explain, but seems to be related to some specific program, piece of hardware, or system (kernel) daemon, it’s definitely worth checking to see if there’s an updated version of that program/driver/kernel available – because there probably is, and applying it may very well fix whatever problems you’ve been having.  And even if a given update doesn’t help with your particular issue, it’ll undoubtedly fix other stability and/or security flaws you haven’t yet encountered, and is very unlikely to make the original problem worse.  I’ve only run into problems of the latter kind once, which were (a) fairly easy to fix, and (b) almost entirely my own fault.

As mentioned above, the components which I’ve found to be the best candidates for periodic, out-of-cycle updates are graphics drivers, networking drivers, and the kernel.  Because I had to do quite a bit of scrounging to figure out how to update each of these, I figured I’d compile my findings here for your future reference:

Networking

You’ll first need to figure out what driver your wireless card is currently using.  In Ubuntu, you should be able to do this by right-clicking on the networking icon on the main taskbar, and selecting <Connection Information>.  Next, go the linuxwireless.org site, find the appropriate entry in the list of available drivers, and click on it.  This should bring you to a page with detailed instructions about how to update your driver.

Graphics

Add the following line(s) to /etc/apt/sources.list (the second bunch may be somewhat unstable; use with discretion):

#Relatively recent X-updates
deb http://ppa.launchpad.net/ubuntu-x-swat/x-updates/ubuntu jaunty main
deb-src http://ppa.launchpad.net/ubuntu-x-swat/x-updates/ubuntu jaunty main

#Bleeding-edge X-updates
deb http://ppa.launchpad.net/xorg-edgers/ppa/ubuntu jaunty main

Then run aptitude update, aptitude upgrade

Kernel

Go to http://kernel.ubuntu.com/~kernel-ppa/mainline/, scroll down to the bottom of the page, and find the highest-numbered folder without an “-rc*” at the end of its name (eg, choose v2.6.31.5 over 2.6.32-rc3).  Enter the folder, and then download the linux-headers-2.6****.deb and linux-image-2.6****.deb files that are appropriate for your system (there are 32 and 64-bit versions available), and linux-image-2.6****all.deb

Then run dpkg -i