Hey there! Thanks for dropping by Theme Preview! Take a look around
and grab the RSS feed to stay updated. See you around!

Archive for March, 2019

AT&T has quietly begun offering DSL service for $10 per month for new customers. Offered as part of the concessions the telecom made to the Federal Communications Commission in order to gain approval for its merger with BellSouth, the speed is nothing to get excited about: 768Kbps down and 128Kbps up. HangZhou Night Net

AT&T is also doing little to publicize the new offering. In fact, I was only able to discover any reference to the low-price service by clicking on the Terms and Conditions link at he bottom of AT&T's residential high-speed Internet product page. A note on AT&T Yahoo! High-Speed Internet buried six paragraphs down says that the "basic speed ($10.00)" tier is available to new customers only, those who have not subscribed to AT&T or BellSouth DSL during the past 12 months, and the service requires a one-year contract.

Customers must also order phone service to get the budget-priced DSL service; those looking for cheap, naked DSL should look elsewhere. Those living in BellSouth's former territory can get naked DSL for the next two-and-a-half years, however.

Along with the budget high-speed Internet and naked DSL, AT&T also promised to maintain a "neutral network and neutral routing in its wireline broadband Internet access service" while also giving up its rights to the 2.5GHz spectrum. (WiMAX provider Clearwire recently completed the purchase of AT&T's unused 2.5GHz holdings.) In addition, AT&T must offer broadband to 100 percent of all residential living units in its territory, with 85 percent of that delivered by wire.

As is the case with the naked DSL offering, AT&T is only required to offer the $10 per month tier for the next two-and-a-half years. After that, the company is free to make whatever changes it wants to the service.

It's only $5 cheaper than AT&T's current lowest-priced service, but at $10 per month, the service could appeal to budget-minded consumers—especially those who are paying about that amount for dial-up service. More importantly for AT&T, it gives the company another platform from which to pitch its U-Verse broadband and IPTV service. After two and a half years of 768Kbps service, U-Verse may look very attractive to lower-tier customers.

A different flavor of BSD

PC-BSD is not a Linux distribution, but rather it could be considered among the first major FreeBSD-based distributions to live outside of the official FreeBSD. Like most distributions, it has implemented certain features in a way that attempts to distinguish it from the competition, and I will focus mostly on these differences. This test drive is intended to give an overview of what PC-BSD is and why one would consider using it. HangZhou Night Net

First and foremost, PC-BSD is an attempt to make a user-friendly Unix. Many Linux distributions have a similar focus and attempt to achieve it in different ways, and PC-BSD should be considered alongside these distributions. Additionally, PC-BSD's developers went to great efforts to make users who are transitioning from Windows more comfortable—more on that later.

The version I tested was PC-BSD 1.3, which is based on FreeBSD 6.1, X.org 6.9, and KDE 3.5.5—none of which are the latest release. The use of older releases fits nicely with PC-BSD's focus on releasing an OS that is stable, secure and friendly. There is a testbed release available for those willing to live on the edge (and bleed a little) that includes more recent software… and the problems associated with it. PC-BSD appears to be available only in the 32-bit x86 flavor.

Hardware test bed:AMD Athlon 64 3200+MSI RS-480-M2 motherboard1GB RAM250GB SATA hard drivePCIe NVIDIA GeForce 7600The installation process

The install program is fast and simple, with limited options for installation. Upon first boot, you are dropped into a ncurses menu that lets you launch the graphical installer, drop into an emergency shell, and so forth. The installer can optionally be run in VESA mode if your video card is not properly detected and initialized (such as the case with my PCIe NIVIDIA GeForce 7600). The fallback mode can be selected from the installation menu.

Once in the graphical installer, you are given a very easy-to-use installation procedure that happens to be a single program running inside Fluxbox. This is only noticeable to the trained eye, as the only clue that you even have a window manager is a one-pixel line running along the bottom of the screen that turns into a taskbar when your mouse gets too close. The installer allows you to choose a "Desktop/Laptop" installation versus a "Server" installation, and it includes things such as automatically setting up the OpenBSD PF (Packet Filter) firewall, which it refers to as the Personal Firewall. Same letters in the acronym… very clever.

There is no package selection, and as a result, installation is very fast, as it's simply a matter of watching the installer extract some tarballs. No configuration is really performed at the time of installation, except for those questions the installer asks. The total time to install was around 20 minutes.

Installation went smoothly until the reboot for me, due once again to my X driver problem. If I was not a *nix professional, I would have panicked at this point. Since I am, I was able to boot into safe mode, log in as root, remount the filesystem as read-write, and try to edit my xorg.conf file. In safe mode, I found that something was wrong with the line terminations when using vi, so I had to use less to view the files and then construct a sed substitution to change the video driver from "nv" to "vesa." Upon reboot, everything worked swimmingly. I should note that the bootloader PC-BSD installs is the FreeBSD default bootloader, which detected my existing SATA drive and always allowed me to boot into my preexisting operating systems if I ran into trouble.

I had selected the option during install to automatically log me into my main user account on boot, and it did just as I requested. I must note that KDE seemed to load much faster on PC-BSD than I'm used to; probably around three times faster than my Kubuntu installation on my other drive (which either says something bad about kubuntu or something great about PC-BSD). In fact, the whole system felt very snappy.

Two law school students filed a lawsuit against the administrator of a web site and 28 of the site's users last week for psychological and economic injury. The two plaintiffs, anonymously listed as Doe I and Doe II, are female students at Yale Law School and claim that the users of a third-party law school message board have consistently and regularly made such disparaging remarks about their characters that it has cost them not only their emotional wellbeing, but internships and jobs. And despite repeated requests to remove the offensive posts, the site's administrators continually refused to do so. HangZhou Night Net

The posts occurred on AutoAdmit, a site that describes itself as the world's "most prestigious" college discussion board and claims to help students with law school information, hiring practices at law firms, and more. The comments against Doe I and II started as far back as 2005 when a poster from Doe I's undergrad university, Stanford, started a thread warning everyone at Yale Law School to "watch out" for her in a thread titled "Stupid Bitch to Attend Yale Law." Thus begun the string of public character assassinations, rumors, and (repeated) rape threats. Various users on the site also posted what she claims to be false information about her LSAT score, accused her of participating in a lesbian relationship with a Yale Law School administrator in order to gain admission, and encouraged others to warn law firms about her alleged illegitimacy.

Similarly in 2007, Doe II became the topic of several threads on AutoAdmit, focusing mostly on certain body parts (complete with pictures of her ripped from sites like Facebook) and also with repeated rape threats. Some posters encouraged others to stalk her and take more photographs, while continuing to encourage various lewd acts.

The complaint

In the complaint as seen by Ars Technica, Doe I and II claim to have lost sleep, fallen behind on schoolwork, suffered strained personal relationships with their families, and were forced to attend therapy as a result of the postings on AutoAdmit. Additionally, Doe I claims to have lost job prospects. She says that at some point, she applied for 16 different on-campus interviews at Yale, which resulted in a mere four callbacks and zero offers. "On information and belief, it is unprecedented for a second-year law student from Yale to participate in so many interviews without obtaining a single summer associate offer," the complaint reads. Her academic qualifications were similar to that of other classmates who had received offers, the complaint says.

The suit names the online pseudonyms of 28 anonymous posters on AutoAdmit in hopes of using subpoenas to identify them in real life. The two women are also suing site administrator Anthony Ciolli, who they say knowingly allowed and profited from these posts staying on the site despite AutoAdmit's "no outing" policy—a policy that states that posts that contain real-life information about other users will be deleted immediately. The women are also concerned that the posts on AutoAdmit are showing up in Google results when users perform queries on their names. The complaint itself mentions that several posters on AutoAdmit have attempted to "googlebomb" the women's names with defamatory comments, and that the first several Google hits for one of the women's names do in fact point to threads from AutoAdmit about them.

Fallout? What fallout?

Targeting Ciolli may prove difficult, however, partly because he did not author the posts himself. Ciolli may also be protected by laws stating that a site's administrators aren't responsible for the posts made by its users, such as the DMCA's Safe Harbor for copyrighted content. In March, Ciolli also told the Washington Post that his co-administrator, Jarret Cohen, was solely responsible for approving or deleting comments and that he had no authority to do so. As an interesting tidbit of side trivia, Ciolli—a law graduate himself—recently had an offer from a Boston law firm rescinded over his involvement with and the content on AutoAdmit, according to the Wall Street Journal.

Discovering the identities of the 28 posters could be difficult as well, since AutoAdmit apparently does not retain IP addresses for its users and does not require them to register with real names, according to the Washington Post—just valid e-mail addresses. However, those e-mail addresses could still eventually give away the identities of the posters involved, as it's probable that the e-mail service providers have more personal information stored about their users than AutoAdmit does and could be forced to give it up through subpoenas.

Ciolli and the AutoAdmit gang may not exactly have precedent on their side either. A student blogger from UC Berkeley recently lost a defamation case brought against him by journalist Lee Kaplan last week. The student, Yaman Salahi, had set up a blog called Lee Kaplan Watch in which Salahi cited articles written by Kaplan and publicly disputed various claims. Kaplan sued Salahi for business interference and libel, which Salahi lost in small claims court not once, but twice. On his blog, Salahi argues that because he was sued in small claims court and not a "real" court, he was unable to take advantage of California's anti-SLAPP—Strategic Lawsuits Against Public Participation—protections. "I have absolutely no doubt that had this lawsuit been filed in a real court, I would have won," Salahi wrote.

Doe I and II are asking for punitive damages in the amount of $245,000 as well as unspecified actual and special damages. The complaint also requests that the threads be permanently removed from AutoAdmit and that the administrators authorize Google to permanently remove cached versions of the threads.

Some experts believe that this case will go a long way towards testing the legal limits of anonymous Internet postings. University of Texas law professor Brian Leiter told Reuters that "the most vile posters on that board are two subpoenas away from being outed," which he says led to "much amusement" by AutoAdmit posters. "But they are about to find out that this is how it works," he added ominously.

A new study says that on average, more than half of the ink from inkjet cartridges is wasted when users toss them in the garbage. Why is that interesting? According to the study, users are tossing the cartridges when their printers are telling them they're out of ink, not when they necessarily are out of ink. HangZhou Night Net

The study by TÜV Rheinland looked at inkjet efficiency across multiple brands, including Epson (who commissioned the study), Lexmark, Canon, HP, Kodak, and Brother. They studied the efficiency of both single and multi-ink cartridges. Espon's printers were among the highest rated, at more than 80 percent efficiency using single-ink cartridges. Kodak's EasyShare 5300 was panned as the worst printer tested, wasting 64 percent of its ink in tests. TÜV Rheinland measured cartridge weights before and after use, stopping use when printers reported that they were out of ink.

That's the first problem. Printers routinely report that they are low on ink even when they aren't, and in some cases there are still hundreds of pages worth of ink left.

The second issue is a familiar one: multi-ink cartridges can be rendered "empty" when only one color runs low. Multi-ink cartridges store three to five colors in a single cartridge. Printing too many photos from the air show will kill your cartridge faster than you can say "blue skies," as dominant colors (say, "blue") are used faster than the others. Therein lies the reason Epson backed the study: the company is singing the praises of its single-ink cartridge approach, an approach which is necessarily more efficient in terms of wasted ink because there's only one color per cartridge, and thus only one cartridge to replace when that color runs out.

Single ink cartridges aren't exactly perfect, however. Such cartridges still were reported as empty with an average of 20 percent of their ink left, which means that an entire cartridge worth of ink is wasted for every five which are used. Given the sky-high prices of ink, this is an alarming find. Epson's own R360 posted the best numbers, with only 9 percent wasted. Yet again, Epson commissioned the tests, so we must ask what's missing.

The study did not measure how much ink is lost due to lack of use, or through cleaning processes. Inkjet cartridges are known to suffer from quality problems if they are not used for long periods of time, sometimes "drying up." This problem has been addressed in recent years, but it has not been eliminated.

The study also did not calculate the total cost per page, which arguably is more important than efficiency. If Epson's multicartridge approach is more efficient, it could nonetheless still be more expensive per page than multi-ink cartridge systems. In its defense, Epson and TÜV Rheinland said that their study focused on the ecological impact of inkjet printing. This is a familiar argument: hybrid cars have also been criticized for their supposed efficiency, with debates raging as to whether or not your average driver will ever see cost savings from better miles-per-gallon given the relative expensive of hybrid engines.

As such, anyone in the market for an inkjet printer still needs to compare specific models to one another to get a feel for efficiency, and Epson's efficiency claims needs to be weighed next to the comparative cost of competing inkjet solutions.

Still, the unintended result of this study is that regardless of the battle between single- and multi-ink cartridges, inkjet printers themselves are significantly off the mark when it comes to reporting the fullness of their cartridges. As the Eagles would say, you're best off when you "take it, to the limit." (Or with a laser printer, one can always do the toner cartridge cha-cha.)

Further reading:

Kodak
inkjets doomed to failure, says EpsonEpson
pushes single-ink cartridges

On Monday, Microsoft launched a new version of its MSN Mobile web site. Besides fitting nicely inside a mobile browser, the site offers most, if not all of the content that comes with MSN. That content includes access to Windows Live products like Windows Live Hotmail, Windows Live Local, Windows Live Messenger, Windows Live Spaces, and Live Search. HangZhou Night Net

One of the major aspects of the new MSN Mobile site is that it has been developed with just about every mobile browser in mind. Specifically, Microsoft claims that the site will render appropriately for all browsers that utilize Wireless Application Protocols 1.2 and higher. Besides optimal rendering, Microsoft has also included quick links to frequently requested content such as e-mail, messaging, and maps.

One of the main user groups that Microsoft is focusing on with this new release is sports fans. Thanks to an exclusive deal with FOXSports.com, fans can use MSN Mobile to access statistics, schedules, scores, and player information. Because MSN Mobile also offers streaming content, there's a good chance that you'll also be able to access plenty of JB, Terry, Howie, and Jimmy's insightful commentary during football season. If that's not a deal-breaker, I don't know what is.

Consumers aside, Microsoft is also positioning the new MSN Mobile to operators. Using Microsoft adCenter and ScreenTonic, the company hopes to generate advertising revenue from both local and national businesses. The acquisition of ScreenTonic will certainly be of use here as its STAMP technology considers factors like screen size, geolocation, and formats when creating targeted advertisements.

Currently, MSN Mobile is only available in the United States, but Microsoft plans to expand the site to a global market throughout the remainder of 2007.