You are quite correct that it could be insecure! However, that's a completely different issue from whether or not it is free software. Yes, free software code auditing would be good for the community running the servers, but would probably be disastrous right now for Bioware.
I would take the following precautions with *any* NWN server running on GNU/Linux: * Run it in a chrooted environment * Run it as an unprivileged user * Run it only on a machine dedicated for the purpose, with no other important services running on it (i.e. a game server running 2 or 3 different games is OK, but running NWN alongside your CVS repository probably isn't.) * Put it into a dedicated DMZ with no "friendly" relations to neighboring machines * Put it on a switched network.
Me, the way I'm running my NWN dedicated server? Leasing a box at Serverbeach for $99 a month. This way, it's completely isolated from any of my important network services. I don't have any user account on the system that's common to any other system of mine. It seems a fairly ideal solution... if the box is cracked, one phone call to Serverbeach and the box is restored to its original pristine condition. This makes life much easier. I hear Rackshack is good for this too... lots of bandwidth, reasonably powerful machines, crappy tech support so you'd better know what you're doing.
Most systems administrators are required to run proprietary code in-network with other code. A machine is as secure as you are willing to make it. That one service may be insecure, sure, but if you take precautions you can avoid a root-level compromise of your system by being sane in your usage, and sufficiently isolating the box you know may contain vulnerabilities. Heck, I remember running a system we *knew* to contain some significant security vulnerabilities for over a year at one point, hanging out on the Internet. Sure, it was unsafe, but the reality was we had no time to keep the antiquated code up-to-date, and some customers demanded the service. Therefore we put the box isolated on its own network, with no special identifying information, arranged a secure log server to receive syslogs from it, and watched our daily reports for any hacking activities. When someone would attempt to exploit vulnerabilities (this was some code unique to this box, so it wasn't like someone would write an automated tool to scan the Internet for this problem), we'd simply ban their IP. After a year, we eventually migrated customers to another system.
IMHO, no Internet-connected system can withstand the attacks of a dedicated system cracker. They *will* find a way in, it's simply a matter of time. Your goal is to keep your life easier by keeping out the script kiddies that comprise 99.999% of the "cracking" community, and make the remaining intelligent system crackers' lives harder by using smart tools to discourage entry (portsentry is a *wonderful* thing! Port-scan my box one time, and you are now perma-banned). And then keep up-to-date on security patches, watch your logs, keep up-to-date on security announcements so that you are aware of potential vulnerabilities, and try really, really hard not to intentionally piss off dedicated system crackers. Even then there's no guarantee. Life is risk. Learn to manage it, and it can work for you. Fear it, and never accomplish anything.
Dang, I cannot seem to post anything to Slashdot shorter than a hundred words lately!
You suggest that the GNU/Linux-using community as a whole is somehow hypocritical when it comes to free software. Most free software advocates (and I count myself as part of this number) acknowledge that there are certain software markets where free software simply may not make sense. These include computer games and applications with massive development cost and tiny markets (dozens of users or fewer).
Supporters of free software run the whole gamut on their opinions, and this is simply my own. There is no hypocricy, as you suggest, because the community is composed of millions of individuals with their own opinions. I personally believe that the approach used by some companies (id Software, Sleepycat Software, etc.) of releasing the source code to older versions of their products, while keeping newer versions proprietary until they have recouped the cost, is an excellent approach that could be profitably mimicked by many more companies.
No, the GNU/Linux client is supposed to have a native installer, if I recall correctly. What you describe is the current state of the GNU/Linux dedicated server, because there is no supported way of extracting the required data from the CD's short of installing the game. Since there's no GNU/Linux installer yet, other methods of getting the dedicated server up would be rather clunky at the moment.
Roughly 6 billion people on the planet (some say as high as 9 billion, it depends on which wild-assed guesses you choose to believe).
If 70% of them have had sex by age 18, then 30% haven't.
0.30 * 6 billion = 1.8 billion people had not had sex by age 18. I'd say that qualifies as at least hundreds of millions, and perhaps billions.
I doubt the veracity of your statistic, however. "Unbiased" is virtually impossible to come by in this question. Most of my peers in high school (I am now 29) had not had sex by age 18, and a great many of them were virgins at marriage.
If you abstain from sex, you will not get any sexually transmitted disease. Despite your skewed perspective, there are hundreds of millions (perhaps billions) of young adults who choose to abstain from sex until marriage due to personal belief, religious creed, or a sound understanding of right and wrong or the risks involved.
Sexual abstinence is a 100% foolproof method of preventing the acquisition of STD's not acquired in vitro. Your doubletalk is a thinly-veiled attempt to justify promiscuity. I was a virgin at the time of my marriage. My wife was a virgin at the time of her marriage. I personally know dozens of people who can say the same.
Can you say the same? If you cannot, you have no leg to stand on saying that condoms are more effective than abstinence at preventing STD's.
Go Bruce! Until you came out here to Utah and talked to the local LUG I thought you were just an overinflated ego with few social skills. I found out that, instead, you have a good reason to have an ego and can make geeks laugh a lot.
Go Bruce!
No, I'm not trying to kiss up, but after seeing his presentation in Logan, I realized Mr. Perens is one guy who has the time, talent, motivation, and POSITION to help destroy software patents and the DMCA. Other people (*cough* Linus *cough*) don't give a damn as long as you can play with code...
Thanks, I'm glad you appreciated it! The simple fact is, people were paying for someone with 5 years of experience what required 15 years of experience before -- and the network administration field has only been around as a viable career choice for around 12 or 14 years. Netware was the king of the hill from 1988 to 1996, with NT bringing up the rear but gaining fast frmo 1996 to 1998. After 98, Microsoft gained the lead in commodity server operating systems. However, for larger corporations, UNIX has been and will continue to be the standard by which all others are measured:)
The basic thrust of my post was to say: don't feel bad. At this point in my career, I value stability over finances so that I can plan for the future, sock away enough money for retirement, and build my side business in my spare time.
Realize too there is a wide disparity in the type and quality of network administrators. Run-of-the-mill Windows NT/2K/XP admins are really hard to find jobs over $60,000 in the Salt Lake City area. Most hover around $40,000. However, if you are a UNIX admin or information systems security specialist, particularly with major, relevant certifications, it really helps a LOT. A CCNA, CCIE, or RHCE is a really big plus. And you can't "fake" your way or "study" your way into passing some of these critical certifications; you *must* have practical problem solving experience, since for some of the higher-end certifications you are thrust into a lab to fix these machines. The Cisco stuff is particularly grueling, but no certification is a "blank check" like some of those MCSE puppy mills had people think.
My opinion: Silicon Valley inflates wages a LOT. Take any salary figure which is representative of your skills and experience which you get from those online polls and chop about 20-30% off the top (at least) to get what would be a reasonable salary for you. A lot of it boils down to negotiating skills since it is MUCH easier to negotiate a higher salary with another company than to negotiate the same raise with your current employer. For the next 12-18 months though, I wouldn't try to job-hop! It stinks.
Well, I hope this was informative at least for somebody.
No, no, Kleenex is useful. You can wipe your nose with it. It has a reason for existence, and what it is supposed to do, it does well.
People who get a certification without having any real experience have no purpose. Get some experience, THEN get a certification. That makes it worth something to indicate that you learned some stuff in your years on the job.
Hmm, I can just speak as a currently out-of-work UNIX admin. Here's the way my salary progression went:
1994-95: $20/hour working contract for a small ISP doing sales & support.
1995-1996: $13.50 an hour, but full-time work, with a small screwdriver shop in Las Vegas doing UNIX (my first experiences), Windows, and Netware support. We were heavy into Netware.
1996-1999: $37,000/year plus bonusses. (Note that this equates to roughly $18 an hour). Was a unix/linux/windows/netware network admin for a now out-of-business computer game company called Singletrac. They were the ones that did the original Twisted Metal series for the Playstation. Unfortunately, they over-expanded, sought a buyer, lost major talent, dried up. The usual "game house that gets too big for their britches" syndrome.
1999-2001: $55,000 a year. Shot the moon in the interview and they gave it to me; I thought I had won the lottery!. Was doing UNIX support exclusively, and got to run the systems administration team (that was fun!). Got regular raises up to $77,000 a year by the time I left for the next big thing. The company was thirty seconds from doom anyway, but many got ticked off that I jumped out of the tub while they were circling the drain:) In any case, this was the height of the dot-com boom, and this was another dot-com.
2001: $85,000 a year. Telecommuted to a small Silicon Valley company. I was all fired up about it, but I discovered that telecommuting is not really for me.
2001-2002: $85,000 a year. Worked for a tech startup here in Salt Lake City, Utah. They were still coasting off the dot-com boom, but just barely shut down. Did UNIX and Cisco support mostly.
At this point, I almost consider my salary history a liability. Realistically, my family would get along just fine on US $45,000 a year. More money than that is really nice, but it's gravy beyond our expenses. We're a typical middle-America family, three kids, no car payments, house payment, student loans, etc. I expect at my next job that I'll get somewhere between $65,000 to $75,000, and that will be just fine. The ride was nice while it lasted, but with my experience, I was overpaid.
Re:Puritan war cry: What about the children!?
on
LoTR Takes 4 Oscars
·
· Score: 2, Flamebait
Wow, could you be any more delusional, offensive, or egotistical? It's obvious to me you consider yourself superior to the majority of the American public. Guess what? You aren't.
"It solves the problem in a way that the stupid deny themselves exposure to interesting things."
What a wonderful way for you to feel self-righteous about your viewing habits. Did you ever stop to think that some of us prefer not to see movies with an "R" rating simply because we have more important things to think about than this image of some guy having his head blown off play through our heads the next day? Movies receive an R rating generally due to language, violence, or sexual content (and, in some rare cases, mature themes absent any of the above). Some people choose to be polite by refusing to use gutter language, and choosing to ask those using it to refrain from doing so in their presence. Some people choose to remain faithful to their spouses in both thought and action by avoiding graphic depictions of others having sexual relations. Some people choose to avoid seeing gratuitous violence because even the thought of killing another human being makes them physically ill, and movies depicting them makes them want to barf.
"the poorer, and land owning people of the midwest have disproportunate power to that of the richer, more liberal people in the cities."
Oy, veh, what another spurious helping of horse manure this is. Farmers are not all poor. Some are, some aren't, same as you get in cities. Land owners, similarly, are not poor. Land is a commodity by which wealth has been measured since time immemorial. As to the issue of disproporionate power, it depends on which side of the fence you sit on. Clinton stole thousands of square miles of revenue-generating land from Utah public school systems (Grand Staircase Escalante Monument) to please green-loving city folk. I choose to live by standards established since the dawn of recorded history as set down in the Ten Commandments. That I vote for legislators who think the same way simply reflects my belief that it is my obligation to support the self-interest of my self and family, the same as you do when you elect foul-mouthed womanizers with a drinking problem.
"The poorer, more religious, people seem to eschew personal responsability in favor of a "smaller government" that's big enough to make sure that everyone isn't thinking impure thoughts."
More pure hogwash. I take this personally simply because you are insulting my people and culture -- that of the mid-western religious family. I don't work to restrict your rights. That some do is within *their* rights, and it is your obligation to see that they do not succeed. I wrestle with civic and moral questions daily, examining each decision in light of what will help my family, community, and nation continue to function. That I choose to eliminate R rated movies from my patronage is simply an expression of my belief that most movies are crap, and I'm glad to have a large portion of them ruled out from the get-go from wasting my valuable time deciding whether to see them or not. I take personal responsibility that I am educated on views I deem important, particularly regarding Constitutional rights and technology issues. If I were to take the time to also view every twisted rated R movie made by a director who simply likes to watch women writhe around naked I would have no time for anything else.
I take personal responsibility for helping ensure that I and my family make wholesome viewing decisions. I couldn't care less about enforcing your viewing habits, but I am grateful for the rating system simply because it means someone else is doing an unpleasant job. I am not the one cleaning latrines at the campground, picking up poop after elephants at the circus, or doing every single job at my place of employment. Others do it so I don't have to, and by simplifying and specializing in my life I am able to contribute a larger amount to my pocketbook and the welfare of the nation.
"Personally, I think God doesn't like hypocrisy so he sends tornados to destroy their trailers."
That's like saying "I think God doesn't like hypocrisy so makes people crazy enough to fly airplanes into large buildings".
I sure hope that in real life you aren't as much of a bigot as your post portrays. There are decent, hard-working people who take their religious, civic, and family obligations seriously in all walks of life, and in every city, town, and rural area of the planet. That you deride them, and the efforts of the ratings boards to help people simplify their choices of entertainment, does both an injustice.
Sure, call me and my neighbors sheeple if you want. But what I care about, I care about passionately, and I'm glad to delegate here and there so I don't go crazy from trying to be everything for everybody at once. All of life is choices. That I and others choose to use our time wisely by avoiding the worst of the filth Hollywood has to offer does not impugn your choices, nor deprive you of any essential liberties.
As an aside, I *do* choose to view rated R movies from time to time. I decide whether or not to view them based upon reviews by my peers and whether I am interested in the subject matter. I've discovered that, in general, the ratings board seems to be right: most rated R movies have foul language, excessive violence, and strong sexual content which I would rather not see. I also do not allow my children to see these until they are of an age (17) where I won't have to wake up repeatedly in the middle of the night to deal with their terrible nightmares from the filth they have viewed.
This gentleman's situation certainly doesn't seem like his ISP is using WCCP though (I love WCCP, particularly version 2.0, while a bit tricky to set up, works a charm). I suspect they're just using a route-map port 80 to some host, with exclusion from the rule for that proxy host.
Simple and effective, but the IP address information from the packet is discarded. Also as far as I know, Squid does not support reading the IP address from the packet it transparently intercepts; it always refers to the DNS list included in its conf file. On the other hand, some other (commercial) caching boxes, which have licensed WCCP 2.0, seem to do better in this kind of strange split-DNS situation.
Then again, it's been about 2 years since I used Squid and several commercial proxies heavily. They may have updated Squid to support packet decoding for IP address destination information in that time. Guess I'll have to go check out that page now...
I disagree. Modem users are not forced to do anything, including using the "upgrade" option of a distribution. They should probably whip out their credit card and pony up the dough to buy a boxed set online. In the U.S., using a modem is saving you $20-$40/month versus a broadband connection anyway, making the $29.95 price of a boxed set of Ximian Gnome, or $59.95 for RedHat 7.2 a trivial price to pay for not having to download hundreds of megabytes of data.
I'm not saying distro makers should abandon the "upgrade" option; I'm saying I have had bad experiences with attempting to "upgrade", versus backing up critical data and installing from scratch. Even Richard Stallman got his start with GNU Emacs selling tapes of Emacs for $150 a pop because people didn't have the pipes to download it!
Back to the topic of this thread, however: Version-specificness (is that a word?) of Ximian is annoying, and I don't envy the modem user unwilling to purchase the boxed set their download of 200-300Mbytes after an upgrade. However, even minor revision numbers of the same distribution often break binary compatability. Is there a better way to do it than Ximian provides?
Hmm, I didn't find it on Ximian's site. If you have Ximian Gnome installed, you can simply click "Help", and select "Frequently Asked Questions". I believe the FAQ is installed with Ximian in file:///usr/share/gnome/html/ximian-faq/index.html . At least, that's where it is on my RedHat 7.2 install with Ximian Gnome.
I've never been a big fan of using the "upgrade" option for any operating system, though... there are usually so many things broken that it's far easier to simply back up the old installation and install from scratch.
Re:Ximian and Distribution upgrades.
on
Evolution 1.0 Released
·
· Score: 3, Informative
This question is a Ximian FAQ:
How do I uninstall Ximian Gnome?
If you're unhappy with Ximian GNOME, please write to support@ximian.com and let us know what the problem is. It's entirely possible we can help you, and you won't need to remove your desktop software.
If you really want to remove Ximian GNOME, it's quite easy: just use Red Carpet to uninstall all the Ximian GNOME packages. Start Red Carpet, then go to the Ximian GNOME channel, click "Remove," then choose "Select All" from the "Edit" menu. Click "Remove Packages." Red Carpet will remove all the GNOME software from your system. Then, remove Red Carpet itself, either by using Red Carpet or your native package management system.
Once that's done, proceed with your upgrade as if you never had Ximian Gnome installed on your system...
Well, not necessarily. The cell and PSTN networks are designed around carrying audio and that is still what they do best. Today, it's a toss-up as to whether it's better to approach text-to-speech from the back-end (where you can have more flexibility) servers, or by embedding pieces into phones which gives you a whole new set of problems and potentially great solutions.
The problem is, the idea of using this tech in phones is fighting against hundreds and hundreds of millions of deployed telephones without any tech newer than perhaps a microchip for caller ID. Over the long-term, text-to-speech embedded in the device is the more efficient and user-controllable format. Over the short haul, though, we're going to see many years still of central-office-controlled voice apps on your phone.
Niche applications, like on a Pocket PC, now there something like this would absolutely rock. Get a toehold, and eventually low-power text-to-speech and speech-to-text devices will be all the rage.
Now if only someone would perfect a speech-to-text engine that didn't require hours of training to recognize my accent...
Heh, I already have this daily on my cell phone using "Voicelink", a service provided by WorldCom (and my employer, Talk2 Technology). I routinely get emails like this:
Just imagine how that sounds read back to you over your cell phone. It really beats having to lug a laptop with me just to check my email, but the kinds of email a sysadmin receives often don't translate well into spoken English. However, it's fun to hear this female voice try to get it right. One of these days, I've gotta get together with the programmers here and make sure these things get read right, like "kill back-tick pee-ess dash ee-eff pipe grep user-name"...
Well, I jumped ship to this little company I work for now called Talk2 Technology (free plug, I guess). We've taken a different tack in voice-enabling applications. I think there are different target markets -- the Talk2 stuff uses servers on the back-end, which go out and fetch your email to read it to you. Putting this on-chip in the cell phone itself is a great step in the right direction.
Fundamentally it's a different approach than today's "voice portal" technology. Voice Portals retrieve data for you, and read it over standard cell or PSTN network. There are many benefits to this approach, principal among them being improved processing power for additional functionality such as voice-processing (speech to text, or compressing speech for reply email voice attachment). By putting the power into the phone, instead of at an expensive central office, this chip could either be a great advancement for text-to-speech technology, or a "killer app" that puts my company out of business:)
Regardless, I'm excited to see this happening. I've long envisioned a PDA with the only interface being spoken, rather than requiring any video component. This would bring the power consumption and delicacy of these devices down within reason for extended usage. The downside is that speech is necessarily a rather slow interface to a machine; it will be interesting to see how we adapt speech for greater speed with speech-based devices, and how English as a whole will fare.
Now that I've used voice-enabled email, it would be really hard to go back to the "old" way. I still do an enormous amount of correspondence every day by typing, but when I'm on the road I don't need to bother with a laptop since I can have my email read to me over the phone *and reply* with a voice message via email. Until you've used it, it's tough to realize how convenient it is.
I want one of these for my Agenda VR3! Or something...
I'm not sure you're entirely correct, that 802.11b is only CSMA/CD. You can run 802.11 in CMSA/CA (Collision Avoidance) mode by using Medium Reservation for hidden nodes. The Hidden Nodes problem happens when you have an access point where A can talk to the AP, and B can talk to the AP, but A and B cannot see each other. It's very normal in 802.11b WAN setups to ensure every card has medium reservation turned on, so they can avoid collisions on the access point. Basically, Medium Reservation causes a node to "listen" for a free "line", if it's free it reserves the channel with the AP. There's a chance that the medium reservation request may collide, but by and large it helps avoid collision difficulties.
The downside? You're limited to 20 or 30 nodes per AP channel, so you have a similar problem. Also, in my experience, medium reservation very slightly downgrades performance per machine under heavy loads for the network.
I've been using an 802.11b uplink to my ISP for two years. It's been an on-and-off proposition. We share airspace with a bunch of emergency alert towers since our town is near a chemical weapons depot (Tooele, UT), and the interference from those siren communications is pretty enormous (right in the same band with 2.4GHz). However, within a mile or so you can get really good performance using directional or YAGI antennas for an ad-hoc peer-to-peer net without an access point, or with an amplified AP (but you have to be careful you're staying under the FCC 1-watt limitation if you amp your AP).
A friend of mine and I had a discussion on just this topic not long ago. We called the concept "ghettonet" -- that people shared connections amongst themselves, with gateways out to DSL, modem, or other wireless links. Routing and channel bonding were the thing we figured would stymie the effort, beyond just a few nodes in the NAN.
Thanks for the info about the reduced-preamble mode; I cut my MTU down to 576 to try to get around the issues with some packet loss due to the interference from the aforementioned towers, but the negotiation overhead with my ISP's AP makes the reduced MTU of somewhat dubious value. It's reduced the number of dropped packets at the expense of bandwidth. Eh well. It's a fun game to play!
I think it's interesting that the eepro100 driver is selected to be staticly compiled into the kernel, rather than a module, by default. With ReiserFS file system support statically compiled, and the eepro100 compiled as a module, network performance was horrid. Compiled the eepro100 into the kernel instead of a module and performance was beautiful over the network.
I'm not quite certain why, but that was my experience.
On that topic, however, I've had the reverse experience -- 3Com NICs have been my fave for years. Exceptional reliability and they always just work under Linux, for me since 5.2. YMMV, but my completely subjective experience is yeah, I've had a few 3com nics blow out on me, but they've been far more reliable than Intel on the whole.
It is important to note that, at the time of the Constitution, the legal fiction of the "corporation" was a completely different animal than we see today. The first known European corporations were founded in the 17th century, and had both the power to form lasting trade arrangements as well as make war on behalf of their respective countries against their competitors. U.S. Corporations were tightly controlled entities, with their principal shareholders held personally responsible for the conduct of the corporation. Corporate charters were regularly revoked by state judiciaries against monopoly or abusive corporations.
In 1886, the U.S. Supreme Court declared that corporations are "real persons". This idea, combined with the 14th amendment's declaration that all "persons" are entitled to equal rights under the constitution, did away with that delicate balance upon which the founding fathers depended. A Corporation, previously an entity subservient to the people who controlled it, now existed in its own right to use its assets in whatever amoral way it would.
This is a case where the judiciary has made an overreaching determination, and by so doing created, by judicial fiat, binding legislation on the rest of us. There is no check nor balance against judicial legislation. We have simply relied on the integrity of the judges in that institution to support sanity and right-thinking. In this case, that long-dead judiciary could never have foreseen the horror they created.
However, there is a straightforward solution to modern-day corporatism. Revoke that expectation that corporations are "real persons". Again make the principals of a corporation criminally responsible for the conduct of their organization. Bring the power of the corporation back to the hands of the *people*! We can prevent this vacuous kowtowing to the siren's song of profit, which preys upon the greed of both our elected officials and corporate shareholders, and we can reign in this horrifying beast we have created.
Revoke the judicial legislation of 1886, and we can win. Otherwise, a corporation remains a "person" of obscene wealth and privilege, against whom no normal person can compete.
But right now, there exists no way to do this within U.S. law. The only entity which can reverse the decision of the Supreme Court *is* the Supreme Court. Figure out how we can change that, and you've figured out an important piece of the puzzle of how to reign in global corporatism...
This is not quite correct. If I recall correctly, Mandrake has some pretty signficant bundling deals with (I think) MacMillan to distribute Mandrake with their books, and that gives them a positive revenue stream. Additionally, it seems like when Linux Mandrake gets store shelf space, it disappears very quickly. My local Wal-Mart sold out of Mandrake 8.0 boxes within a few weeks of their arrival on store shelves, and hasn't been able to keep them in stock. (Yes, that's purely anecdotal evidence which does not prove causation; it's merely an observation.)
Contrary to some "dot-bomb" companies of recent times, it seems like the guys who make Mandrake Linux 8.1 are conservative in their approach to selling their product, and drop non-revenue-generating offerings very quickly. They started a whole portal thing for the release of 7.2, then dropped it fairly quickly since it didn't look like it would make money. As a user of Linux-Mandrake since version 5.2 (the first version), I've seen them grow the package from a small offshoot of a stock RedHat install to a robust and up-to-date distribution with an enormous developer base. Subscribe to the "cooker" mailing list for a few weeks; the volume of mail is just incredible, and package announcements come in almost daily. Yeah, the products are pushed out a little early, without incredibly robust quality assurance. I hate to say it, but the strategy worked for Microsoft, and it works for Mandrake: People (like me!)*want* the latest and greatest, and put up with problems in exchange for benefits, so put it out there early and bugfix it later. Mandrake Linux, IIRC, is not yet a publicly traded company, but it seems to me that they have their act together, and know how to market the product so they stay in the black.
On a side note: I noticed Linux-Mandrake changed the name to "Mandrake Linux" with version 8.1. I wonder if that's a step towards calling it "Mandrake GNU/Linux" with version 8.2?
Just so you know, there is a pretty significant difference between versions of packages marketed by Mandrake. The "download edition" doesn't nearly include all the software, or even quite the same install, as the commercial versions, which bundle Star Office, Real Player, and a host of other useful, but non-free, applications.
I highly recommend buying an off-the-shelf version and comparing it to the version you downloaded. It's rather enlightening, and it's tough to go back to the "download edition" afterward.
"Collisions" are far less of a concern on a switched, full-duplex network such as this. If you have an intelligent switch, it will queue packets for an interface, and the back-end switched fabrics of these switches generally mean zero packet loss or collisions. Your full-duplex test of transfer rates very closely simulates the transfer rate you'd receive if you had those devices plugged into a very busy but high-quality switch.
Collisions can, however, be a concern if for some reason the device at the end doesn't support full-duplex operation -- then it is possible for the switch and the device to collide with each other, but you still don't have nearly the same problems you have with traditional hubs. Additionally, it is possible that you can have line errors which force device negotiation at a lower speed, half-duplex, or simply cause random lost packets and noise on the line. This is far less likely with fiber to the home, but if the ends of the cable are not polished well you'll have lots of lost packets -- but still, generally no collisions because the switch and the end-device are not transmitting & receiving in full-duplex mode.
However, I largely agree with your point. Hard drive transfer rates are often abominable. However, the latest drives can be faster than 100Mbps. The article mentioned copying files from one hard disk to another. Write speed on hard drives is generally a small fraction of the read speed; while you may read at 18 or 19 Mbytes/sec (easily saturating a 100Mbps link), writing often only happens at 4 to 5 MBytes/sec, and on many hard drives even slower than that (one here at my house consistently comes in at 780Kbytes/sec!). Once we get writes up in the 10Mbytes/sec range for run-of-the-mill consumer hard disks, even 100Mbps connections will begin to seem quite slow...
I love switches!
Slashdot Mirror
You are quite correct that it could be insecure! However, that's a completely different issue from whether or not it is free software. Yes, free software code auditing would be good for the community running the servers, but would probably be disastrous right now for Bioware.
I would take the following precautions with *any* NWN server running on GNU/Linux:
* Run it in a chrooted environment
* Run it as an unprivileged user
* Run it only on a machine dedicated for the purpose, with no other important services running on it (i.e. a game server running 2 or 3 different games is OK, but running NWN alongside your CVS repository probably isn't.)
* Put it into a dedicated DMZ with no "friendly" relations to neighboring machines
* Put it on a switched network.
Me, the way I'm running my NWN dedicated server? Leasing a box at Serverbeach for $99 a month. This way, it's completely isolated from any of my important network services. I don't have any user account on the system that's common to any other system of mine. It seems a fairly ideal solution... if the box is cracked, one phone call to Serverbeach and the box is restored to its original pristine condition. This makes life much easier. I hear Rackshack is good for this too... lots of bandwidth, reasonably powerful machines, crappy tech support so you'd better know what you're doing.
Most systems administrators are required to run proprietary code in-network with other code. A machine is as secure as you are willing to make it. That one service may be insecure, sure, but if you take precautions you can avoid a root-level compromise of your system by being sane in your usage, and sufficiently isolating the box you know may contain vulnerabilities. Heck, I remember running a system we *knew* to contain some significant security vulnerabilities for over a year at one point, hanging out on the Internet. Sure, it was unsafe, but the reality was we had no time to keep the antiquated code up-to-date, and some customers demanded the service. Therefore we put the box isolated on its own network, with no special identifying information, arranged a secure log server to receive syslogs from it, and watched our daily reports for any hacking activities. When someone would attempt to exploit vulnerabilities (this was some code unique to this box, so it wasn't like someone would write an automated tool to scan the Internet for this problem), we'd simply ban their IP. After a year, we eventually migrated customers to another system.
IMHO, no Internet-connected system can withstand the attacks of a dedicated system cracker. They *will* find a way in, it's simply a matter of time. Your goal is to keep your life easier by keeping out the script kiddies that comprise 99.999% of the "cracking" community, and make the remaining intelligent system crackers' lives harder by using smart tools to discourage entry (portsentry is a *wonderful* thing! Port-scan my box one time, and you are now perma-banned). And then keep up-to-date on security patches, watch your logs, keep up-to-date on security announcements so that you are aware of potential vulnerabilities, and try really, really hard not to intentionally piss off dedicated system crackers. Even then there's no guarantee. Life is risk. Learn to manage it, and it can work for you. Fear it, and never accomplish anything.
Dang, I cannot seem to post anything to Slashdot shorter than a hundred words lately!
You suggest that the GNU/Linux-using community as a whole is somehow hypocritical when it comes to free software. Most free software advocates (and I count myself as part of this number) acknowledge that there are certain software markets where free software simply may not make sense. These include computer games and applications with massive development cost and tiny markets (dozens of users or fewer).
Supporters of free software run the whole gamut on their opinions, and this is simply my own. There is no hypocricy, as you suggest, because the community is composed of millions of individuals with their own opinions. I personally believe that the approach used by some companies (id Software, Sleepycat Software, etc.) of releasing the source code to older versions of their products, while keeping newer versions proprietary until they have recouped the cost, is an excellent approach that could be profitably mimicked by many more companies.
No, the GNU/Linux client is supposed to have a native installer, if I recall correctly. What you describe is the current state of the GNU/Linux dedicated server, because there is no supported way of extracting the required data from the CD's short of installing the game. Since there's no GNU/Linux installer yet, other methods of getting the dedicated server up would be rather clunky at the moment.
Even using your statistic:
Roughly 6 billion people on the planet (some say as high as 9 billion, it depends on which wild-assed guesses you choose to believe).
If 70% of them have had sex by age 18, then 30% haven't.
0.30 * 6 billion = 1.8 billion people had not had sex by age 18. I'd say that qualifies as at least hundreds of millions, and perhaps billions.
I doubt the veracity of your statistic, however. "Unbiased" is virtually impossible to come by in this question. Most of my peers in high school (I am now 29) had not had sex by age 18, and a great many of them were virgins at marriage.
"Condoms work better than abstinence."
That is a load of horse shit.
If you abstain from sex, you will not get any sexually transmitted disease. Despite your skewed perspective, there are hundreds of millions (perhaps billions) of young adults who choose to abstain from sex until marriage due to personal belief, religious creed, or a sound understanding of right and wrong or the risks involved.
Sexual abstinence is a 100% foolproof method of preventing the acquisition of STD's not acquired in vitro. Your doubletalk is a thinly-veiled attempt to justify promiscuity. I was a virgin at the time of my marriage. My wife was a virgin at the time of her marriage. I personally know dozens of people who can say the same.
Can you say the same? If you cannot, you have no leg to stand on saying that condoms are more effective than abstinence at preventing STD's.
Go Bruce! Until you came out here to Utah and talked to the local LUG I thought you were just an overinflated ego with few social skills. I found out that, instead, you have a good reason to have an ego and can make geeks laugh a lot.
Go Bruce!
No, I'm not trying to kiss up, but after seeing his presentation in Logan, I realized Mr. Perens is one guy who has the time, talent, motivation, and POSITION to help destroy software patents and the DMCA. Other people (*cough* Linus *cough*) don't give a damn as long as you can play with code...
Thanks, I'm glad you appreciated it! The simple fact is, people were paying for someone with 5 years of experience what required 15 years of experience before -- and the network administration field has only been around as a viable career choice for around 12 or 14 years. Netware was the king of the hill from 1988 to 1996, with NT bringing up the rear but gaining fast frmo 1996 to 1998. After 98, Microsoft gained the lead in commodity server operating systems. However, for larger corporations, UNIX has been and will continue to be the standard by which all others are measured :)
The basic thrust of my post was to say: don't feel bad. At this point in my career, I value stability over finances so that I can plan for the future, sock away enough money for retirement, and build my side business in my spare time.
Realize too there is a wide disparity in the type and quality of network administrators. Run-of-the-mill Windows NT/2K/XP admins are really hard to find jobs over $60,000 in the Salt Lake City area. Most hover around $40,000. However, if you are a UNIX admin or information systems security specialist, particularly with major, relevant certifications, it really helps a LOT. A CCNA, CCIE, or RHCE is a really big plus. And you can't "fake" your way or "study" your way into passing some of these critical certifications; you *must* have practical problem solving experience, since for some of the higher-end certifications you are thrust into a lab to fix these machines. The Cisco stuff is particularly grueling, but no certification is a "blank check" like some of those MCSE puppy mills had people think.
My opinion: Silicon Valley inflates wages a LOT. Take any salary figure which is representative of your skills and experience which you get from those online polls and chop about 20-30% off the top (at least) to get what would be a reasonable salary for you. A lot of it boils down to negotiating skills since it is MUCH easier to negotiate a higher salary with another company than to negotiate the same raise with your current employer. For the next 12-18 months though, I wouldn't try to job-hop! It stinks.
Well, I hope this was informative at least for somebody.
No, no, Kleenex is useful. You can wipe your nose with it. It has a reason for existence, and what it is supposed to do, it does well.
People who get a certification without having any real experience have no purpose. Get some experience, THEN get a certification. That makes it worth something to indicate that you learned some stuff in your years on the job.
Hmm, I can just speak as a currently out-of-work UNIX admin. Here's the way my salary progression went:
:) In any case, this was the height of the dot-com boom, and this was another dot-com.
1994-95: $20/hour working contract for a small ISP doing sales & support.
1995-1996: $13.50 an hour, but full-time work, with a small screwdriver shop in Las Vegas doing UNIX (my first experiences), Windows, and Netware support. We were heavy into Netware.
1996-1999: $37,000/year plus bonusses. (Note that this equates to roughly $18 an hour). Was a unix/linux/windows/netware network admin for a now out-of-business computer game company called Singletrac. They were the ones that did the original Twisted Metal series for the Playstation. Unfortunately, they over-expanded, sought a buyer, lost major talent, dried up. The usual "game house that gets too big for their britches" syndrome.
1999-2001: $55,000 a year. Shot the moon in the interview and they gave it to me; I thought I had won the lottery!. Was doing UNIX support exclusively, and got to run the systems administration team (that was fun!). Got regular raises up to $77,000 a year by the time I left for the next big thing. The company was thirty seconds from doom anyway, but many got ticked off that I jumped out of the tub while they were circling the drain
2001: $85,000 a year. Telecommuted to a small Silicon Valley company. I was all fired up about it, but I discovered that telecommuting is not really for me.
2001-2002: $85,000 a year. Worked for a tech startup here in Salt Lake City, Utah. They were still coasting off the dot-com boom, but just barely shut down. Did UNIX and Cisco support mostly.
At this point, I almost consider my salary history a liability. Realistically, my family would get along just fine on US $45,000 a year. More money than that is really nice, but it's gravy beyond our expenses. We're a typical middle-America family, three kids, no car payments, house payment, student loans, etc. I expect at my next job that I'll get somewhere between $65,000 to $75,000, and that will be just fine. The ride was nice while it lasted, but with my experience, I was overpaid.
Wow, could you be any more delusional, offensive, or egotistical? It's obvious to me you consider yourself superior to the majority of the American public. Guess what? You aren't.
"It solves the problem in a way that the stupid deny themselves exposure to interesting things."
What a wonderful way for you to feel self-righteous about your viewing habits. Did you ever stop to think that some of us prefer not to see movies with an "R" rating simply because we have more important things to think about than this image of some guy having his head blown off play through our heads the next day? Movies receive an R rating generally due to language, violence, or sexual content (and, in some rare cases, mature themes absent any of the above). Some people choose to be polite by refusing to use gutter language, and choosing to ask those using it to refrain from doing so in their presence. Some people choose to remain faithful to their spouses in both thought and action by avoiding graphic depictions of others having sexual relations. Some people choose to avoid seeing gratuitous violence because even the thought of killing another human being makes them physically ill, and movies depicting them makes them want to barf.
"the poorer, and land owning people of the midwest have disproportunate power to that of the richer, more liberal people in the cities."
Oy, veh, what another spurious helping of horse manure this is. Farmers are not all poor. Some are, some aren't, same as you get in cities. Land owners, similarly, are not poor. Land is a commodity by which wealth has been measured since time immemorial. As to the issue of disproporionate power, it depends on which side of the fence you sit on. Clinton stole thousands of square miles of revenue-generating land from Utah public school systems (Grand Staircase Escalante Monument) to please green-loving city folk. I choose to live by standards established since the dawn of recorded history as set down in the Ten Commandments. That I vote for legislators who think the same way simply reflects my belief that it is my obligation to support the self-interest of my self and family, the same as you do when you elect foul-mouthed womanizers with a drinking problem.
"The poorer, more religious, people seem to eschew personal responsability in favor of a "smaller government" that's big enough to make sure that everyone isn't thinking impure thoughts."
More pure hogwash. I take this personally simply because you are insulting my people and culture -- that of the mid-western religious family. I don't work to restrict your rights. That some do is within *their* rights, and it is your obligation to see that they do not succeed. I wrestle with civic and moral questions daily, examining each decision in light of what will help my family, community, and nation continue to function. That I choose to eliminate R rated movies from my patronage is simply an expression of my belief that most movies are crap, and I'm glad to have a large portion of them ruled out from the get-go from wasting my valuable time deciding whether to see them or not. I take personal responsibility that I am educated on views I deem important, particularly regarding Constitutional rights and technology issues. If I were to take the time to also view every twisted rated R movie made by a director who simply likes to watch women writhe around naked I would have no time for anything else.
I take personal responsibility for helping ensure that I and my family make wholesome viewing decisions. I couldn't care less about enforcing your viewing habits, but I am grateful for the rating system simply because it means someone else is doing an unpleasant job. I am not the one cleaning latrines at the campground, picking up poop after elephants at the circus, or doing every single job at my place of employment. Others do it so I don't have to, and by simplifying and specializing in my life I am able to contribute a larger amount to my pocketbook and the welfare of the nation.
"Personally, I think God doesn't like hypocrisy so he sends tornados to destroy their trailers."
That's like saying "I think God doesn't like hypocrisy so makes people crazy enough to fly airplanes into large buildings".
I sure hope that in real life you aren't as much of a bigot as your post portrays. There are decent, hard-working people who take their religious, civic, and family obligations seriously in all walks of life, and in every city, town, and rural area of the planet. That you deride them, and the efforts of the ratings boards to help people simplify their choices of entertainment, does both an injustice.
Sure, call me and my neighbors sheeple if you want. But what I care about, I care about passionately, and I'm glad to delegate here and there so I don't go crazy from trying to be everything for everybody at once. All of life is choices. That I and others choose to use our time wisely by avoiding the worst of the filth Hollywood has to offer does not impugn your choices, nor deprive you of any essential liberties.
As an aside, I *do* choose to view rated R movies from time to time. I decide whether or not to view them based upon reviews by my peers and whether I am interested in the subject matter. I've discovered that, in general, the ratings board seems to be right: most rated R movies have foul language, excessive violence, and strong sexual content which I would rather not see. I also do not allow my children to see these until they are of an age (17) where I won't have to wake up repeatedly in the middle of the night to deal with their terrible nightmares from the filth they have viewed.
This gentleman's situation certainly doesn't seem like his ISP is using WCCP though (I love WCCP, particularly version 2.0, while a bit tricky to set up, works a charm). I suspect they're just using a route-map port 80 to some host, with exclusion from the rule for that proxy host.
Simple and effective, but the IP address information from the packet is discarded. Also as far as I know, Squid does not support reading the IP address from the packet it transparently intercepts; it always refers to the DNS list included in its conf file. On the other hand, some other (commercial) caching boxes, which have licensed WCCP 2.0, seem to do better in this kind of strange split-DNS situation.
Then again, it's been about 2 years since I used Squid and several commercial proxies heavily. They may have updated Squid to support packet decoding for IP address destination information in that time. Guess I'll have to go check out that page now...
Guess I'll respond to the flamebait :)
I disagree. Modem users are not forced to do anything, including using the "upgrade" option of a distribution. They should probably whip out their credit card and pony up the dough to buy a boxed set online. In the U.S., using a modem is saving you $20-$40/month versus a broadband connection anyway, making the $29.95 price of a boxed set of Ximian Gnome, or $59.95 for RedHat 7.2 a trivial price to pay for not having to download hundreds of megabytes of data.
I'm not saying distro makers should abandon the "upgrade" option; I'm saying I have had bad experiences with attempting to "upgrade", versus backing up critical data and installing from scratch. Even Richard Stallman got his start with GNU Emacs selling tapes of Emacs for $150 a pop because people didn't have the pipes to download it!
Back to the topic of this thread, however: Version-specificness (is that a word?) of Ximian is annoying, and I don't envy the modem user unwilling to purchase the boxed set their download of 200-300Mbytes after an upgrade. However, even minor revision numbers of the same distribution often break binary compatability. Is there a better way to do it than Ximian provides?
Hmm, I didn't find it on Ximian's site. If you have Ximian Gnome installed, you can simply click "Help", and select "Frequently Asked Questions". I believe the FAQ is installed with Ximian in file:///usr/share/gnome/html/ximian-faq/index.html . At least, that's where it is on my RedHat 7.2 install with Ximian Gnome.
I've never been a big fan of using the "upgrade" option for any operating system, though... there are usually so many things broken that it's far easier to simply back up the old installation and install from scratch.
Once that's done, proceed with your upgrade as if you never had Ximian Gnome installed on your system...
S/Mime is slated for 1.1 or 1.2 release. See http://bugzilla.ximian.com/show_bug.cgi?id=1350 for details.
Ximian was waiting on the Evolution 1.0 release to release the Mandrake 8.1 version of it. Expect an 8.1 release in the next two weeks.
Well, not necessarily. The cell and PSTN networks are designed around carrying audio and that is still what they do best. Today, it's a toss-up as to whether it's better to approach text-to-speech from the back-end (where you can have more flexibility) servers, or by embedding pieces into phones which gives you a whole new set of problems and potentially great solutions.
The problem is, the idea of using this tech in phones is fighting against hundreds and hundreds of millions of deployed telephones without any tech newer than perhaps a microchip for caller ID. Over the long-term, text-to-speech embedded in the device is the more efficient and user-controllable format. Over the short haul, though, we're going to see many years still of central-office-controlled voice apps on your phone.
Niche applications, like on a Pocket PC, now there something like this would absolutely rock. Get a toehold, and eventually low-power text-to-speech and speech-to-text devices will be all the rage.
Now if only someone would perfect a speech-to-text engine that didn't require hours of training to recognize my accent...
Heh, I already have this daily on my cell phone using "Voicelink", a service provided by WorldCom (and my employer, Talk2 Technology). I routinely get emails like this:
From: root
Subject: cron kill `ps -ef | grep username | awk '{print $2}'`
Just imagine how that sounds read back to you over your cell phone. It really beats having to lug a laptop with me just to check my email, but the kinds of email a sysadmin receives often don't translate well into spoken English. However, it's fun to hear this female voice try to get it right. One of these days, I've gotta get together with the programmers here and make sure these things get read right, like "kill back-tick pee-ess dash ee-eff pipe grep user-name"...
Well, I jumped ship to this little company I work for now called Talk2 Technology (free plug, I guess). We've taken a different tack in voice-enabling applications. I think there are different target markets -- the Talk2 stuff uses servers on the back-end, which go out and fetch your email to read it to you. Putting this on-chip in the cell phone itself is a great step in the right direction.
:)
Fundamentally it's a different approach than today's "voice portal" technology. Voice Portals retrieve data for you, and read it over standard cell or PSTN network. There are many benefits to this approach, principal among them being improved processing power for additional functionality such as voice-processing (speech to text, or compressing speech for reply email voice attachment). By putting the power into the phone, instead of at an expensive central office, this chip could either be a great advancement for text-to-speech technology, or a "killer app" that puts my company out of business
Regardless, I'm excited to see this happening. I've long envisioned a PDA with the only interface being spoken, rather than requiring any video component. This would bring the power consumption and delicacy of these devices down within reason for extended usage. The downside is that speech is necessarily a rather slow interface to a machine; it will be interesting to see how we adapt speech for greater speed with speech-based devices, and how English as a whole will fare.
Now that I've used voice-enabled email, it would be really hard to go back to the "old" way. I still do an enormous amount of correspondence every day by typing, but when I'm on the road I don't need to bother with a laptop since I can have my email read to me over the phone *and reply* with a voice message via email. Until you've used it, it's tough to realize how convenient it is.
I want one of these for my Agenda VR3! Or something...
I'm not sure you're entirely correct, that 802.11b is only CSMA/CD. You can run 802.11 in CMSA/CA (Collision Avoidance) mode by using Medium Reservation for hidden nodes. The Hidden Nodes problem happens when you have an access point where A can talk to the AP, and B can talk to the AP, but A and B cannot see each other. It's very normal in 802.11b WAN setups to ensure every card has medium reservation turned on, so they can avoid collisions on the access point. Basically, Medium Reservation causes a node to "listen" for a free "line", if it's free it reserves the channel with the AP. There's a chance that the medium reservation request may collide, but by and large it helps avoid collision difficulties.
The downside? You're limited to 20 or 30 nodes per AP channel, so you have a similar problem. Also, in my experience, medium reservation very slightly downgrades performance per machine under heavy loads for the network.
I've been using an 802.11b uplink to my ISP for two years. It's been an on-and-off proposition. We share airspace with a bunch of emergency alert towers since our town is near a chemical weapons depot (Tooele, UT), and the interference from those siren communications is pretty enormous (right in the same band with 2.4GHz). However, within a mile or so you can get really good performance using directional or YAGI antennas for an ad-hoc peer-to-peer net without an access point, or with an amplified AP (but you have to be careful you're staying under the FCC 1-watt limitation if you amp your AP).
A friend of mine and I had a discussion on just this topic not long ago. We called the concept "ghettonet" -- that people shared connections amongst themselves, with gateways out to DSL, modem, or other wireless links. Routing and channel bonding were the thing we figured would stymie the effort, beyond just a few nodes in the NAN.
Thanks for the info about the reduced-preamble mode; I cut my MTU down to 576 to try to get around the issues with some packet loss due to the interference from the aforementioned towers, but the negotiation overhead with my ISP's AP makes the reduced MTU of somewhat dubious value. It's reduced the number of dropped packets at the expense of bandwidth. Eh well. It's a fun game to play!
I think it's interesting that the eepro100 driver is selected to be staticly compiled into the kernel, rather than a module, by default. With ReiserFS file system support statically compiled, and the eepro100 compiled as a module, network performance was horrid. Compiled the eepro100 into the kernel instead of a module and performance was beautiful over the network.
I'm not quite certain why, but that was my experience.
On that topic, however, I've had the reverse experience -- 3Com NICs have been my fave for years. Exceptional reliability and they always just work under Linux, for me since 5.2. YMMV, but my completely subjective experience is yeah, I've had a few 3com nics blow out on me, but they've been far more reliable than Intel on the whole.
It is important to note that, at the time of the Constitution, the legal fiction of the "corporation" was a completely different animal than we see today. The first known European corporations were founded in the 17th century, and had both the power to form lasting trade arrangements as well as make war on behalf of their respective countries against their competitors. U.S. Corporations were tightly controlled entities, with their principal shareholders held personally responsible for the conduct of the corporation. Corporate charters were regularly revoked by state judiciaries against monopoly or abusive corporations.
In 1886, the U.S. Supreme Court declared that corporations are "real persons". This idea, combined with the 14th amendment's declaration that all "persons" are entitled to equal rights under the constitution, did away with that delicate balance upon which the founding fathers depended. A Corporation, previously an entity subservient to the people who controlled it, now existed in its own right to use its assets in whatever amoral way it would.
This is a case where the judiciary has made an overreaching determination, and by so doing created, by judicial fiat, binding legislation on the rest of us. There is no check nor balance against judicial legislation. We have simply relied on the integrity of the judges in that institution to support sanity and right-thinking. In this case, that long-dead judiciary could never have foreseen the horror they created.
However, there is a straightforward solution to modern-day corporatism. Revoke that expectation that corporations are "real persons". Again make the principals of a corporation criminally responsible for the conduct of their organization. Bring the power of the corporation back to the hands of the *people*! We can prevent this vacuous kowtowing to the siren's song of profit, which preys upon the greed of both our elected officials and corporate shareholders, and we can reign in this horrifying beast we have created.
Revoke the judicial legislation of 1886, and we can win. Otherwise, a corporation remains a "person" of obscene wealth and privilege, against whom no normal person can compete.
But right now, there exists no way to do this within U.S. law. The only entity which can reverse the decision of the Supreme Court *is* the Supreme Court. Figure out how we can change that, and you've figured out an important piece of the puzzle of how to reign in global corporatism...
This is not quite correct. If I recall correctly, Mandrake has some pretty signficant bundling deals with (I think) MacMillan to distribute Mandrake with their books, and that gives them a positive revenue stream. Additionally, it seems like when Linux Mandrake gets store shelf space, it disappears very quickly. My local Wal-Mart sold out of Mandrake 8.0 boxes within a few weeks of their arrival on store shelves, and hasn't been able to keep them in stock. (Yes, that's purely anecdotal evidence which does not prove causation; it's merely an observation.)
Contrary to some "dot-bomb" companies of recent times, it seems like the guys who make Mandrake Linux 8.1 are conservative in their approach to selling their product, and drop non-revenue-generating offerings very quickly. They started a whole portal thing for the release of 7.2, then dropped it fairly quickly since it didn't look like it would make money. As a user of Linux-Mandrake since version 5.2 (the first version), I've seen them grow the package from a small offshoot of a stock RedHat install to a robust and up-to-date distribution with an enormous developer base. Subscribe to the "cooker" mailing list for a few weeks; the volume of mail is just incredible, and package announcements come in almost daily. Yeah, the products are pushed out a little early, without incredibly robust quality assurance. I hate to say it, but the strategy worked for Microsoft, and it works for Mandrake: People (like me!)*want* the latest and greatest, and put up with problems in exchange for benefits, so put it out there early and bugfix it later. Mandrake Linux, IIRC, is not yet a publicly traded company, but it seems to me that they have their act together, and know how to market the product so they stay in the black.
On a side note: I noticed Linux-Mandrake changed the name to "Mandrake Linux" with version 8.1. I wonder if that's a step towards calling it "Mandrake GNU/Linux" with version 8.2?
Just so you know, there is a pretty significant difference between versions of packages marketed by Mandrake. The "download edition" doesn't nearly include all the software, or even quite the same install, as the commercial versions, which bundle Star Office, Real Player, and a host of other useful, but non-free, applications.
I highly recommend buying an off-the-shelf version and comparing it to the version you downloaded. It's rather enlightening, and it's tough to go back to the "download edition" afterward.
A correction:
"Collisions" are far less of a concern on a switched, full-duplex network such as this. If you have an intelligent switch, it will queue packets for an interface, and the back-end switched fabrics of these switches generally mean zero packet loss or collisions. Your full-duplex test of transfer rates very closely simulates the transfer rate you'd receive if you had those devices plugged into a very busy but high-quality switch.
Collisions can, however, be a concern if for some reason the device at the end doesn't support full-duplex operation -- then it is possible for the switch and the device to collide with each other, but you still don't have nearly the same problems you have with traditional hubs. Additionally, it is possible that you can have line errors which force device negotiation at a lower speed, half-duplex, or simply cause random lost packets and noise on the line. This is far less likely with fiber to the home, but if the ends of the cable are not polished well you'll have lots of lost packets -- but still, generally no collisions because the switch and the end-device are not transmitting & receiving in full-duplex mode.
However, I largely agree with your point. Hard drive transfer rates are often abominable. However, the latest drives can be faster than 100Mbps. The article mentioned copying files from one hard disk to another. Write speed on hard drives is generally a small fraction of the read speed; while you may read at 18 or 19 Mbytes/sec (easily saturating a 100Mbps link), writing often only happens at 4 to 5 MBytes/sec, and on many hard drives even slower than that (one here at my house consistently comes in at 780Kbytes/sec!). Once we get writes up in the 10Mbytes/sec range for run-of-the-mill consumer hard disks, even 100Mbps connections will begin to seem quite slow...
I love switches!