One thing i haven't seen considered here is that some geeks may actually go out and buy a mac JUST for the purpose of running OS X... Personally, I've already planned on buying a G4 when OS X is available. Why? Why buy 500-pound SPARC servers? Why build multiple dual Xeon machines? or alpha systems? Nobody in the geek community thought it so odd when I did THAT, after all.
Experience is the only true killer of inspiration. If OS X does suck, or doesn't actually run X/GTK/KDE/whatever stuff to any decent level, I'll be the person who will know it and know it from firsthand experience, while 95% of you reading this won't have a damn clue, because you aren't even willing to learn the system in the first place... hell, many of you won't even consider taking a look at it in your local computer store!
Apple does "wierd" stuff. Some of their "wierd" stuff has flat-out revolutionized home and low-end business computing. Hypercard... ANYONE remember that? Product of the YEAR for 1985 THROUGH 1987. The precursor to PDF, HTML, XML, Visual Basic-style WYSIWYG devlopment, and more. Apple's OS has brought lots of concepts to the end-user for the first time, including multitasking, time-slicing, and memory-swapping. Even Quarterdeck's DESQView for DOS didn't show up for another 7 years or so.
All i'm saying is that perhaps those who don't automatically hold some sort of bias will spend their money in order to educate themselves on the product, regardless of it being a proprietary solution or not. I still have win9x,NT,2k, and whistler beta machines in the house, despite the fact that i run exclusively Linux and Solaris/SPARC on my desktop machines. Please consider that self-education might just muck up your auto-disqualification demographics. If OS X has features as revolutionary as many of Apple's previous products, you'd better be prepared to see a lot of real geeks pick themselves up a 5.5GFLOP titanium box to toy around on.
Just to be a nitpicking annoyance, I wanted to point out that most large distribution sites are running a variant of BSD.:) a handful (such as sunsite.unc.edu) run solaris, but most of them (cdrom.com, freesoftware.com, many of tucows.com's mirrors) are hosting from BSD:) It may be more accurate to view BSD as supporting Linux's distribution:)
If you want to make an uber-distro like this, it would certainly make sense to start with Debian and add in the features from Slackware as you see fit (The.tgz packages are too easy to install [cd/;tar -zxf path/to/file.tgz])... You can use the Debian or slackware RPM package as a core for adding your redhat-supported features without a whole helluva lot of trouble.
As far as retaining a safe package base that interacts properly with most administrative programs, Debian would certainly be the best. RedHat makes many major modifications to administrative tools that makes their tools a bit sketchy. Slackware deviates from the GNU core quite a bit too, to retain the BSD-like feel.
If you're feeling lazy and brave all at once, you could look at taking VectorLinux (sorry, I'm too lazy to find their URL), which is a stripped-down and all-in-one version of Slack and use that for a base. I definitely wouldn't put this choice above going with Debian base plus outside packages, though.
"Difficult" would be more clearly stated as "unfamiliar"... If you are accustomed to more pure SysV systems, such as Solaris or AIX, Slack is quite different and unfamiliar. However, if you come from a BSD environment, Slackware feels like FreeBSD (or OpenBSD if you like to keep Slack clean) with pretty directory colors. The reason why you say "difficult" is because it is unfamiliar due to its architectural difference in this respect.
I felt the same way when I was spontaneously stuck fixing a RedHat machine for the first time, or when I first ran Debian. I didn't end up becoming accustomed to the SysV way of handling things until running Solaris. Now, I have no problem with those design differences when using RedHat or Debian, and although I prefer using Slack (Praise Bob!), it's due to other reasons other than unfamiliarity.
Lack of personal knowledge of a system does not make a system difficult - it makes the user uninformed. Don't be afraid to RTFM:)
I'd like to also say that I've seen very few major-distribution developers that are as personable with their users as David is, even in some less-than-fortunate circumstances. I just wanted to thank David publicly for being such a great example of a community-friendly Linux Distribution developer. It's good to see him get a well-deserved interview!
THIS is one that Katz should have in his Hellmouth series.
This is so dead-on. In this little slashdot community we often get caught up in some gang-mentality of saying we were all right and they were all wrong, or some other similar nonsense. Here is a case where both sides need to change. Neither side is perfect. Neither side is necessarily right, for that matter. I know I couldn't have said it as well as Dido, and I hope that this is recognized.
NSI is chock full of questionable actions. Here's one that I'm 100% sure is going on in regards to their auctioning of expired names:
The WHOIS query tool (which is now limited in function to a few select methods) logs every WHOIS lookup of a domain name, so that when that domain name expires, they know how many people have looked at it. This gives them a very unscientific estimate of how much they can (over)charge for that particular domain name.
We, as an Internet Community, may benefit from a petition to NSI and ICANN, requesting that informational searches are open to all users in a method that does not contain inherent logging and is NOT logged by the registrar. Proof of compliance may be met in many ways, and I'm sure many of you would know something better than I'd propose.
By cutting down on NSI's statistics-gathering, we can make it a little more difficult for them to price-gauge based on the popularity or perceived popularity of a domain name.
I have the feeling your subject didn't get quite enough attention.
Censorship fulfills a purpose: It keeps our "impressionable children" from viewing things they "shouldn't see". It also keeps employees from looking at porn while supposedly working on their latest paperwork.
If our children are _so_ impressionable, why does the trend not exist to impression them with self-disgretion? If they are taught to choose between looking at porn or watching the Barney Lobotomy TV show, and they can think and act on their own, this censorship for the ignorant and incapable will be limited to a small few.
This same idea of placing responsibility at the individual also works well in proper business. If you are hiring the type of employees that prefer to look at porn all day instead of doing their work, you are hiring employees with a very poor work ethic (or your work environment really sucks). Get a better interviewing policy and hire better workers.
American culture has been leading the charge of dissipating responsibility for reality since the end of World War II. We have lawyers suiing lawyers because they didn't play nice in court. We have people suing fast food restaurants for them spilling coffee on themselves. We have grown children dying because they ride a bicycle into traffic, and the manufacturer is somehow held responsible. It is very rare that we have a valid situation (such as the Firestone Tire fiasco) where responsibility needs to be corrected by a lawsuit.
Instead of turning off the TV, we censor the TV shows. Instead of closing the browser window (or going to a different site), we censor the content.
We find a problem, and we try to edit the result, like applying a `|sed -e "...."` to a program's output, rather than just fixing the program (the root of the problem) in the first place!
I certainly hope I am not the only person who sees this cultural syndrome and actively disowns it.
the REASON why 64-bit usage isn't currently practical is because it is not currently available for developers to create optimized software for!
Frankly, by that definition, a car over 50 HP is also impractical, because it will still be able to handle top speed in city limits most (non-interstate) highways.
There's a lot more available that just more data in the datapaths... smart assembly hackers will learn to pack and hack smaller bits of data through single registers to reduce processor ticks, just as they did when 32bit processors became available. Current on-the-fly rendering (like desktop animations, software and hardware DVD playback, etc) will greatly benefit from the bus increases, datapath size, and capability of simply dumping bigger numbers through fewer cycles. It sure as hell beats chomping a 36-bit number into 32-bit segments to process each segment of it and then reassemble it for the user. Tasks like that will be orders of magnitude faster (that example would be 2^4, or 16 times, faster), and you will see a whole lore more of them in the very near future.
I don't think it really matters if we have to do little things like rework our compilers...
Intel's last true architectural change was with the introduction of the 386SX processor. Since then, we have only had patchlevel additions of little things here and there. The 386DX was a huge step up from the 386SX, but it was still only a patchlevel increase in functionality.
The fact of the matter is that we are hitting some logical and structural limitations in Intel's current 32-bit architecture that we simply must overcome. This has been even more apparent with the influx of flaky and poorly-performing motherboard chipsets from Intel, which has been a rarity until recently.
I don't think it's a matter of if we're ready to go forward - are we willing to stay where we are, on a backwards-designed architecture with design bottlenecks?
Much agreed. Honda is a great example of how quality assurance and checking is simply superior in Japan, and how their products do not lag behind US products anywhere near the degree reflected in the article.
I have an '88 Honda Civic that i getting retired this month. It's been running hard (and i mean HARD) for the last 12 years, with only the expected maintenance. I managed to snap the axles, crack the radiator, and flood the car, but that's because i'm an absolutely insane driver.:) I still *love* the car (and it is in better condition than a friend's 1998 Kia Sephia, and she is very gentle with her car), but the mechanics of any 12-year-old vehicle that has been blatantly abused (but i had fun abusing it!) are going to become substandard.
Its replacement is an Acura Integra, which is another Honda vehicle. Obviously, you know why - you brought up the relation in the first place... It's just raw quality. The consumers that choose to spend a little bit (and it's normally a VERY little bit) more money for a higher quality product have a much lower rate of problems.
In a commercial environment, You get what you pay for. Linux and most open-source software "goes against" this cliche, but there are other factors governing its success.
Agreed. I actually have benchmarks proving that a particular four-letter Major PC manufacturer's default installation of NT workstation 4 is outperformed by a 100% to 200% increase in all file I/O, all IPC, and 30-75% decrease in program load time, all by reinstalling NT workstation from scratch, installing the right drivers, and applying the latest service pack.
A "fix" to this is to install a certain Big-Name caching program (nine letters long) which speeds up disk I/O, most of the IPC, and some of the load time. This Major PC manufacturer is negotiating a license with certain Big-Name caching program creators, which will increase the price of their business workstations by at least $100.
I won't even go into what happens when I build a custom PC with next-to-identical parts... the hardware is just as bad as their packaged software:(
Yes, it's a sad condition... What's really scary is that it applies to hardware, too.
Back circa 1985, if you bought yourself a computer (Lets just say you get a Mac SE), the rate of failure out-of-the-box was extremely low. In the case of 40 particular Macintosh systems bought in 1985-1987, I know that only one of them needed hardware service at all by 1991. However, in 1993, the same company bought 13 new Macintosh systems (to replace some of those old Mac SE's that were STILL running), and had hardware problems with 3 of those 13 within a year.
Apple has statistically held a much better defect/failure rate than most IBM-clone PC manufacturers (the big ones, like Gateway, Dell, Compaq, etc), let alone all the independent stores building machines themselves.
I can say from personal experience (formerly being a system production manager for such a small company) that the return rates on computer hardware are absurdly high. Our store was among the best (sadly) with an average of a 10-12% return rate for components. If you statistically sum up all the 10-12% possible breakdowns of components in a computer, that should give you a clue as to how many of our full systems we saw back - more than 50% within the year. Talk about depressing...
Personally, I think that the software situation is worse in that respect, but it's just a lot easier to upgrade a program than it is to solder a new north bridge on your motherboard.:)
This is a horrible example, verging on a flame, but eliciting ideas is what i'm all about:
Do you think that, for example, OpenBSD could successfully authorize other developers' software (Lets use ISC BIND server as an easy example) without code changes? My opinion would be that they could not do this. The reason is that the OS team would have to FULLY audit every third-party program. They would have to report back possible bugfixes (to CVS, hopefully with commit priveleges), wait for another release version, and re-audit again! I believe it would tie down the OS developers too much in babysitting others' code to work on their own.
It would be a very cool idea if they could find a better-than-lukewarm way of doing this, but I for one do not see a feasible way of doing it.
I would agree strongly with Chocodile on, well, all of his points!
OpenBSD is by no means any more difficult to learn or use than, say, RedHat. OpenBSD has very thoroughly and consisely documented information on their system, their methodologies, etc. Their tools are respectably inter-operable and close-knit. RedHat has a number of undocumented and/or poorly/obscurely documented system files, system utilities, and other core components that should be handled with better care.
I started using Debian, then went with Slackware (still my #1 Linux), tried FreeBSD (and HATED IT@##@$@), and then used OpenBSD. Currently there is ONE reason why OpenBSD is not on all of my servers (it's on half of them): support for only 8 character usernames. Its stability also exceeds that of Linux by orders of magnitude, with my hardware.
As far as workstations go, I have a SparcStation IPX (40mhz) that served as an X workstation until 3 weeks ago, when I got a Sparc 5/110 for next-to-free. I entirely didn't need to upgrade from the IPX - it was a fully useable workstation that ran surprisingly well. Not many operating systems can do that with a 40mhz machine.:)
I'll make one complaint (and it's probably not really OpenBSD's fault): On the SPARC architecture, I can't compile ORBit, thus I can't compile GTK, and thus I can't use xmms. That's my only complaint.
And it takes a weaker mind to be influenced by entertainment.
You describe responses you had to films - These are great examples. If these films were to influence you or effect you for any longer than an emotional moment, perhaps you would find yourself having totalitarian, Nazi-like tendencies after watching Schindler's list. Perhaps you would feel very threatened and live with a more paranoid deameanor. Such a "response" would then merit the word "influence". Films rarely influence people on any grand scale. Almost everyone can mention a film that really did influence them, but these are individual events having to do with that individual's experiences.
Jumping around all out of order, I'd like to add a perspective to the quote "This child is rehearsing for life." It is true that they are rehearsing for life while they are playing games. Is it not true that they are experiencing and learning first-hand the unreality of their games? Are you saying that they are growing to understand these games as real? Rather, the contrary! They increasingly come to understand the fact that games are an escape from reality. They are entirely and wholly separate from reality. People feared that as games became more realistic these logical lines would fade. What was ignored there was that the human mind is adaptive, and adapts to the higher level of understanding that even these new, fancy, T&L-rendered games are still as unreal as Pong.
Back to the middle somewhere...
The pavlovian response you mention appears to be common. You can go to any teenage LAN party where Doom/Quake/etc is being played and see how much fun they have. It is easy to see this as a Pavlovian response to the blood. Have you considered that the reason such violence and such extreme action is so appealing is because of how UNREAL it is? If you play a game of Quake II, and the Q2 server sets the numgibs variable to 50 (instead of the default of like 8), You will literally see exploding meat when someone is killed. There is absolutely no doubt in my mind or the mind of ANY person I have met that the game is LESS real with all that additional gore and blood than it was before. It is the unreality, the need for escape, that drives this response you perceive as a "Pavlovian response."
Enough scattered ranting. You should be able to see the message I'm trying to get across: MANY people use games for an escape from reality. Actions in games are unreal and KNOWN to be unreal, and do not influence us or our children in the despicably elementary manners that video game opponents encourage us all to believe.
I am very very very close (probably too close) with a number of people at Sunbelt Software... Here's some useful info:
They have an Exchange server that they use (Quad Xeon 500, 2gb RAM, 200something gig RAID-5)... firstly, they don't even use "Real View". They hate it. They also hate their mail server, but since they are an NT-software distributor, they can't just fly off and switch to *nix and a real mail server.
Just to make a decent comparison, my single PII-266 with 64mb RAM can push 80% of the load that their Exchange server pushes... It runs Qmail on Linux. Hell, it's not even configured for speed (20 maximum instances).
I just wanted to let ya know that Sunbelt won't recommend that software to their friends... good to have an honest, inside view sometimes.
Here we have a hardware device which does almost exactly the same things that DeCSS does, with the same "feared" effect by the MPAA: being able to record mpeg-2 directly to your hard drive.
This seems like it would be in direct conflict with the MPAA's position on DeCSS... Is it possible that this product's release may cause a stir in the DeCSS case [ MPAA | 2600/Burns ], particularly in the view of copying digial media of which one already has a copy? Seems evident to me - hopefully the same sentiment will creep into the heads of everyone involved in the case.
The high school I attended was the cliche jock/prep/normal school in the local area. Like any school, it had a small population of non-conformists (10 to 20 out of 2100 students)... it just happened that this school had a very small group.
I met another member of this classification as a junior, as he was a freshman. A tall, lanky, evil "goth" (oh how ambiguous) looking 'freak' with plenty of piercings (about 20), he was the picturesque extreme of the school's 'freak' culture.
He wasn't liked by anyone, and that was his biggest advantage of all. He would get picked on, harassed, attacked, you name it. Instead of viewing these events as tragic causes for his victimization, he simply used them as motivation to shove their closed-minded views back in their faces.
He was never a 'politically correct' individual. In fact, he offered to piss on the principal's desk when the principal wouldn't let him use the restroom because there was cigarette smoke in there. Political correctness would have only made him an inert object, getting passed through the bowels of the school system.
He showed up for a whopping 60 of 180 days that first year, and by the end of it there wasn't a single person in the school who didn't know his name. He pulled some standard pranks, like stealing a golf cart and driving it through the halls, climbing out of fire escape windows during lunch... but those pranks only made his face known, not his name or attitude.
He was probably the only person who wouldn't think twice about screaming out his disagreements with the administration (normally something like "these fucking dimwits think it's reasonable to suspend me for telling a teacher she's a stupid bitch! If she wants to teach math, she needs to learn fucking multiplication!"), especially in front of the administration, in order to publicly provoke them and let their responses be known to the thousand students standing and watching.
The next year (my senior and his sophomore, er, um, well, freshman again), the situation escalated to the point of urban legend. He showed up for a record-low of 10 days throughout the 180-day school year, occasionally making additional appearances at lunches. He managed to receive more than 30 suspensions, despite his 10 day attendance. One particular day he had seen the principal thrice (throwing chairs on the cafeteria roof, walking in late and obnoxiously as hell into class, and (in another class) being painfully polite to a teacher). When he came in for the last visit (which baffled the principal), he received a 10 day suspension (maximum in this district, also guarantees failure for the semester) with the comment of "i'm just tired of seeing you."
His stupid antagonist pranks were the biggest problem for the administration... They were somehow foolhearty enough to ignore the message he was making clear to the other students: There are absurd restrictions and violations of freedom going on here, and the only way that you can make any change is to become active. Prove by example (which is how all U.S. Legal precedent is set) that their restrictions and censorships are not going to be tolerated.
About halfway through the year, a number of interesting changes started to happen. The school newspaper, previously known for typical school-censoring measures, published a handful of opinion columns that directly attacked and questioned the administration and their actions. One of the columns was regarding a friend of mine who was expelled for 'posession of a lethal weapon while under the influence of alcohol'. He lived far from school, actually outside of the school zone, and worked at the local grocery store. He had a 1-inch pocketknife (you can't even unscrew a case-screw with this flimsy POS) which he used for opening boxes. This was his 'lethal weapon'. Being "under the influence of alcohol" was based on a particular teacher of ours (i was sitting next to him) overhearing him talk about having a really bad hangover... She missed the part of him talking about a party we went to months before. He was never given a blood test, breathalizer test, walk-the-line-and-touch-your-nose test, or anything else. Just expelled.
There was a big uproar in the student population, with testimony coming from 4 students (one being myself), and the superintendent reversed the expulsion decision and (in more politically-correct terms) called the principal an idiot. He was. (He was also stoned off his ass every time I saw him, but that's a whole separate story!)
Back somewhere near topic... the newspaper actually published and distributed a column in regards to this expulsion and, at that time, pending reversal. It was crossing the line of polite, and deeply attacked the administration. Of course, the administration didn't like this and tried to retrieve all the released copies of the newspaper. The principal himself came into my classroom ("my" classroom was literal; i co-taught a 2D/3D modeling/rendering class which the most awesome teacher and I started that year) to collect the newspapers. This was also the first class of the day for my expelled friend. When he requested all the newspapers be returned, and some of the kids started reluctantly shuffling up the newspapers, I shook my head clear and started arguing with him. I knew that my testimony was _the_ primary evidence against my friend's expulsion, and since I was also co-teaching the class I was in, the principal couldn't afford to shoot himself in the foot and strongarm my argument. I asked him what he was trying to prove by covering up his actions, which were public record. I also offered to release the public information myself to all the students. I asked him if covering up this information would help his argument to the superintendent, and I (rather unwisely) threatened to flood the school with this article if he were to remove it. He didn't challenge me, and left.
Had it not been for my acquaintance (and now best friend and co-owner of a company) who had managed to become a 'King of Geeks and Freaks', my quiet, conservative, previously sheepish self would NEVER have made that stance against the principal and policy.
If I were to recommend something to help the situation of students being hassled like this, it would be to find a smart "freak/punk/geek/goth/etc" at your local school and tell them to _make_ the school system prove their policies, as my friend did. This does a LOT more than merely shake up the administration - it makes the other students aware that in the constitutionally-breaching school systems, they can still force their civil liberties to be respected, and gain the respect (if not fame) of their school class.
Start with improving the school environment, and the other problems (self-esteem of the outcasts, inter-student harassment, administrative violations of liberty) will become MUCH easier to improve.
No system can be fixed unless it is first fixed at its core. Avoid Fixing the surface issues - it does nothing but help the core problems rot in neglect.
I don't see how I could continue bantering on with someone when the purpose of argument is so drastically misconstrued between participants.
A quality argument cannot ensue when a broad point is recanted with overly specific, if not nit-picking, examples. I could turn around and nitpick equally well... it will serve no purpose other than to satisfy some pending need to vent about something (read: anything).
My arguments of updated/outdated/etc were made to emphasise the structural and organizational differences in the distributions. It was construed as a much broader statement.
Enough of this - i can't expect to rationalize with someone who can change understandings of the overall point of the discussion arbitrarily.
Ideology: An operating system is an operating system. This is the identity property. Basic logic. An Operating System is NOT an office suite. An Operating System is NOT an mp3 player.
My point is that the concept of making an operating system is that you build the core on top of which applications run. Slackware does not follow this ideology the way I like it (OpenBSD-style), but they do MUCH better than RedHat.
Quantity of software is also a null point. Firstly, the greater sludge of packages added by RedHat (same situation with debian) are rival programs striving to serve the same purpose. In fact, you may even end up with 3 packages that are just different versions of the same program! The ones that aren't duplicate are almost always outdated, oftentimes due to buffer overflows and other bugs. That is EXACTLY why RedHat distributions have so many security problems.
Follow the simple idea that you install your operating system and then get your application software from the proper sources. You end up with the latest releases (that work) and much fewer security problems.
It's a simple ideology issue. RedHat takes the AOL hodgepodge mentality. Slackware takes the erector-set mentality (a frame, get your own gear). I'll give you three guesses which the technically apt prefer.;>
IMO, actual power users don't use packages. They are the ones who MAKE the packages.
As far as mechanical quality on BMW bikes, you're correct, but that was not the point of that statement. My point was to illustrate the riders of such bikes. You have your BMW riders, who can't quite hack it on their own, and then you have your harley riders. If you're familiar with the biking scene this should be crystal clear.
BMW releases (yet another) new experimental motorcycle. Everyone who has been in the motorcycle industry for more than 10-15 years tells BMW to shove it, whilst the newer crew rants and raves at how great it is.
Harley-Davidson releases (this is rare) a new experimental motorcycle. Everyone who has been in the motorcycle industry for more than 10-15 years thanks them, whilst the newer crew bitches and moans.
You've got two distributions (Slackware, RedHat) who serve the same core purpose: make a distribution. They go about it in different ways and cater to different people. Slackware, in relativity to redhat, is geared more toward a power-user or server admin. RedHat is geared toward a new user or less experienced admin.
Redhat has had more than 142 security problems published on securityfocus.com in the last year. Slackware has had 14. Of course, these include older versions of the OS's, as well as some included third-party programs, but nonetheless it shows that RedHat's integrity in releasing solid PRODUCTION code is outright horrible, let alone their development. Slackware's record is far superior (and not without falter!), and the community in general respects Slackware for this qualitative approach of releasing a true alpha to the public for open development. They are not only supporting the community with a product, but also with the opportunity to participate in some good ole'fashioned development.
(just to nitpick and advocate Slackware (Praise Bob!), i'd like to make clear that ZERO of the vulnerabilities listed to effect slackware had to do with the implementation or compilation of the distribution. They were either GNU.org code bugs, kernel bugs, or third party software bugs.)
Development code is only as good as your developers... Says a lot more for Slackware's developers than it does RedHat's.
Slackware/SPARC is in SUPER pre-release. It is downloadable only by rsync at this point. This release was intended to be a semi-private release to a small group of developers for the sake of stabilizing the codebase. After that was completed, an FTP-accessible release would be available.
Some important notes from the developers:
Slackware/SPARC is not ready for production.
SILO is a bit broken and needs some work.
many other packages are very unstable, if installable.
changes for these problems will be occuring DAILY and WILL BE UPDATED AS SUCH on the rsync server. A cron job is in place on their rsync server to create a fresh ISO every night at midnight.
I'd offer an FTP site with the daily updates of Slack4Sparc, but my bandwidth can't handle it. If any of you have bandwidth to do this job, I can get you going with the updates.
Props to my crew man Toby Freaksoft, though he tends to get a bit antsy about these sorts of things.:)
Whoever said slackware was dead should gingerly pry their cranium out of their colon. Thanks!:)
I cannot speak for your area, but I'd like to take a shot at outlining a local ISP which I use (and LOVE), their basic policies, and how to pinpoint a similar ISP in your area.
I live in Clearwater, FL and there are a SLEW of internet providers. I literally have my pick of 100 or more ISP's. I work as a Network Engineer now, and previously built a local ISP from the ground up.
The company I choose to do business with now is called Intelligence Network. They have a strict policy of using only Cisco routers and switches, and only Sun Microsystems workstations and servers. Personally, I like playing with old Sun hardware, but I greatly disagree with many of Sun's policies regarding licensing. Nevertheless, no informed individual in the business can say that your average Intel/Bay Networks enterprise-level network can even compare with your average Sun/Cisco network. That's Point 1: Quality.
Secondly, they are a small-to mid-sized company. They have four static T-1's and a burstable OC-12. I'd prefer they had a burstable T-3 instead of the four T-1's, but nevertheless, that is again one of the signs of a smaller, more personalized company.That's Point 2: A small, non-conglomerate company.
Next, they have little to no advertising. They rely on the quality of their service to keep their business at a solid level. They are not in the business seeking profits (or else they wouldn't be spending well over $1,000,000 per server they have), but rather they are in the business to provide a purely quality-oriented service. Point 3: No major advertising.
Lastly, their tech support is good. This is the hardest to gauge and monitor. They have perhaps a dozen technicians. Some are assigned to dialup service, some to DSL and T1/E1 service, and some to colocation and other services. Atop this they have 3 head technicians who actually comminicate directly with the customer. In fact, they're so good with communication that they gave me a free 10mbit SBUS ethernet card for an old Sparc IPX i've got for a firewall. They are the kind of company that wants to get personal with the customer and actually develop a relationship with them. Point 4: Personal Customer relationships.
Okay, one more point. This one is very arguable and lies entirely in my preference. I like this company because they do NOT try to match the bottom-price market. Instead, they charge a fair price for a very reliable service. They are not motivated by high-volume income, and it is very clear by the quality and (relatively small) size of their company. I look for this with every business or service I deal with.
For you performance geeks out there, my DSL connection has had a total of approximately 2 hours of downtime in the last year. I monitor my uptime by 30-second intervals and keep it logged in MRTG-style reports. Those two hours were an accumulation of many very short service failures, many of which I proved were due to AT&T doing maintenance at 4am on Sunday morning every other month. Nearly all of the rest of it, short of approximately 20 minutes, was due to GTE/Verizon (many bad things to say about them). Their end result from these points is an extremely reliable service, and my performance is ALWAYS 99% or better of my bandwidth provisioning.
If you want a _quality_ provider, open up your local phone book and pick a handful of ISP's you have never heard of in your local area. Call each of them and ask them two simple questions:
What hardware do you use to run your network and servers?
How much bandwidth do you have, and from what providers?
If they have the bandwidth to realistically support the performance they claim, and they are using quality hardware, you've got a company that meets three of the above points: Quality, Small Company, and NO Major Advertising. At that point, there's only one way to really test their customer service, and that's to use their service.
Don't expect your "cheapest and fastest" advertising ISP to get you anywhere. Throw in a few more dollars and ignore the claimed performance gains, and deal with a company that actually wants to provide service.
Slashdot Mirror
One thing i haven't seen considered here is that some geeks may actually go out and buy a mac JUST for the purpose of running OS X... Personally, I've already planned on buying a G4 when OS X is available. Why? Why buy 500-pound SPARC servers? Why build multiple dual Xeon machines? or alpha systems? Nobody in the geek community thought it so odd when I did THAT, after all.
Experience is the only true killer of inspiration. If OS X does suck, or doesn't actually run X/GTK/KDE/whatever stuff to any decent level, I'll be the person who will know it and know it from firsthand experience, while 95% of you reading this won't have a damn clue, because you aren't even willing to learn the system in the first place... hell, many of you won't even consider taking a look at it in your local computer store!
Apple does "wierd" stuff. Some of their "wierd" stuff has flat-out revolutionized home and low-end business computing. Hypercard... ANYONE remember that? Product of the YEAR for 1985 THROUGH 1987. The precursor to PDF, HTML, XML, Visual Basic-style WYSIWYG devlopment, and more. Apple's OS has brought lots of concepts to the end-user for the first time, including multitasking, time-slicing, and memory-swapping. Even Quarterdeck's DESQView for DOS didn't show up for another 7 years or so.
All i'm saying is that perhaps those who don't automatically hold some sort of bias will spend their money in order to educate themselves on the product, regardless of it being a proprietary solution or not. I still have win9x,NT,2k, and whistler beta machines in the house, despite the fact that i run exclusively Linux and Solaris/SPARC on my desktop machines. Please consider that self-education might just muck up your auto-disqualification demographics. If OS X has features as revolutionary as many of Apple's previous products, you'd better be prepared to see a lot of real geeks pick themselves up a 5.5GFLOP titanium box to toy around on.
Just to be a nitpicking annoyance, I wanted to point out that most large distribution sites are running a variant of BSD. :) a handful (such as sunsite.unc.edu) run solaris, but most of them (cdrom.com, freesoftware.com, many of tucows.com's mirrors) are hosting from BSD :) It may be more accurate to view BSD as supporting Linux's distribution :)
If you want to make an uber-distro like this, it would certainly make sense to start with Debian and add in the features from Slackware as you see fit (The .tgz packages are too easy to install [cd /;tar -zxf path/to/file.tgz])... You can use the Debian or slackware RPM package as a core for adding your redhat-supported features without a whole helluva lot of trouble.
As far as retaining a safe package base that interacts properly with most administrative programs, Debian would certainly be the best. RedHat makes many major modifications to administrative tools that makes their tools a bit sketchy. Slackware deviates from the GNU core quite a bit too, to retain the BSD-like feel.
If you're feeling lazy and brave all at once, you could look at taking VectorLinux (sorry, I'm too lazy to find their URL), which is a stripped-down and all-in-one version of Slack and use that for a base. I definitely wouldn't put this choice above going with Debian base plus outside packages, though.
"Difficult" would be more clearly stated as "unfamiliar"... If you are accustomed to more pure SysV systems, such as Solaris or AIX, Slack is quite different and unfamiliar. However, if you come from a BSD environment, Slackware feels like FreeBSD (or OpenBSD if you like to keep Slack clean) with pretty directory colors. The reason why you say "difficult" is because it is unfamiliar due to its architectural difference in this respect.
:)
I felt the same way when I was spontaneously stuck fixing a RedHat machine for the first time, or when I first ran Debian. I didn't end up becoming accustomed to the SysV way of handling things until running Solaris. Now, I have no problem with those design differences when using RedHat or Debian, and although I prefer using Slack (Praise Bob!), it's due to other reasons other than unfamiliarity.
Lack of personal knowledge of a system does not make a system difficult - it makes the user uninformed. Don't be afraid to RTFM
I'd like to also say that I've seen very few major-distribution developers that are as personable with their users as David is, even in some less-than-fortunate circumstances. I just wanted to thank David publicly for being such a great example of a community-friendly Linux Distribution developer. It's good to see him get a well-deserved interview!
THIS is one that Katz should have in his Hellmouth series.
This is so dead-on. In this little slashdot community we often get caught up in some gang-mentality of saying we were all right and they were all wrong, or some other similar nonsense. Here is a case where both sides need to change. Neither side is perfect. Neither side is necessarily right, for that matter. I know I couldn't have said it as well as Dido, and I hope that this is recognized.
NSI is chock full of questionable actions. Here's one that I'm 100% sure is going on in regards to their auctioning of expired names:
The WHOIS query tool (which is now limited in function to a few select methods) logs every WHOIS lookup of a domain name, so that when that domain name expires, they know how many people have looked at it. This gives them a very unscientific estimate of how much they can (over)charge for that particular domain name.
We, as an Internet Community, may benefit from a petition to NSI and ICANN, requesting that informational searches are open to all users in a method that does not contain inherent logging and is NOT logged by the registrar. Proof of compliance may be met in many ways, and I'm sure many of you would know something better than I'd propose.
By cutting down on NSI's statistics-gathering, we can make it a little more difficult for them to price-gauge based on the popularity or perceived popularity of a domain name.
I have the feeling your subject didn't get quite enough attention.
Censorship fulfills a purpose: It keeps our "impressionable children" from viewing things they "shouldn't see". It also keeps employees from looking at porn while supposedly working on their latest paperwork.
If our children are _so_ impressionable, why does the trend not exist to impression them with self-disgretion? If they are taught to choose between looking at porn or watching the Barney Lobotomy TV show, and they can think and act on their own, this censorship for the ignorant and incapable will be limited to a small few.
This same idea of placing responsibility at the individual also works well in proper business. If you are hiring the type of employees that prefer to look at porn all day instead of doing their work, you are hiring employees with a very poor work ethic (or your work environment really sucks). Get a better interviewing policy and hire better workers.
American culture has been leading the charge of dissipating responsibility for reality since the end of World War II. We have lawyers suiing lawyers because they didn't play nice in court. We have people suing fast food restaurants for them spilling coffee on themselves. We have grown children dying because they ride a bicycle into traffic, and the manufacturer is somehow held responsible. It is very rare that we have a valid situation (such as the Firestone Tire fiasco) where responsibility needs to be corrected by a lawsuit.
Instead of turning off the TV, we censor the TV shows. Instead of closing the browser window (or going to a different site), we censor the content. We find a problem, and we try to edit the result, like applying a `|sed -e "...."` to a program's output, rather than just fixing the program (the root of the problem) in the first place!
I certainly hope I am not the only person who sees this cultural syndrome and actively disowns it.
the REASON why 64-bit usage isn't currently practical is because it is not currently available for developers to create optimized software for!
Frankly, by that definition, a car over 50 HP is also impractical, because it will still be able to handle top speed in city limits most (non-interstate) highways.
There's a lot more available that just more data in the datapaths... smart assembly hackers will learn to pack and hack smaller bits of data through single registers to reduce processor ticks, just as they did when 32bit processors became available. Current on-the-fly rendering (like desktop animations, software and hardware DVD playback, etc) will greatly benefit from the bus increases, datapath size, and capability of simply dumping bigger numbers through fewer cycles. It sure as hell beats chomping a 36-bit number into 32-bit segments to process each segment of it and then reassemble it for the user. Tasks like that will be orders of magnitude faster (that example would be 2^4, or 16 times, faster), and you will see a whole lore more of them in the very near future.
I don't think it really matters if we have to do little things like rework our compilers...
Intel's last true architectural change was with the introduction of the 386SX processor. Since then, we have only had patchlevel additions of little things here and there. The 386DX was a huge step up from the 386SX, but it was still only a patchlevel increase in functionality.
The fact of the matter is that we are hitting some logical and structural limitations in Intel's current 32-bit architecture that we simply must overcome. This has been even more apparent with the influx of flaky and poorly-performing motherboard chipsets from Intel, which has been a rarity until recently.
I don't think it's a matter of if we're ready to go forward - are we willing to stay where we are, on a backwards-designed architecture with design bottlenecks?
Much agreed. Honda is a great example of how quality assurance and checking is simply superior in Japan, and how their products do not lag behind US products anywhere near the degree reflected in the article.
:) I still *love* the car (and it is in better condition than a friend's 1998 Kia Sephia, and she is very gentle with her car), but the mechanics of any 12-year-old vehicle that has been blatantly abused (but i had fun abusing it!) are going to become substandard.
I have an '88 Honda Civic that i getting retired this month. It's been running hard (and i mean HARD) for the last 12 years, with only the expected maintenance. I managed to snap the axles, crack the radiator, and flood the car, but that's because i'm an absolutely insane driver.
Its replacement is an Acura Integra, which is another Honda vehicle. Obviously, you know why - you brought up the relation in the first place... It's just raw quality. The consumers that choose to spend a little bit (and it's normally a VERY little bit) more money for a higher quality product have a much lower rate of problems.
In a commercial environment, You get what you pay for. Linux and most open-source software "goes against" this cliche, but there are other factors governing its success.
Agreed. I actually have benchmarks proving that a particular four-letter Major PC manufacturer's default installation of NT workstation 4 is outperformed by a 100% to 200% increase in all file I/O, all IPC, and 30-75% decrease in program load time, all by reinstalling NT workstation from scratch, installing the right drivers, and applying the latest service pack.
:(
A "fix" to this is to install a certain Big-Name caching program (nine letters long) which speeds up disk I/O, most of the IPC, and some of the load time. This Major PC manufacturer is negotiating a license with certain Big-Name caching program creators, which will increase the price of their business workstations by at least $100.
I won't even go into what happens when I build a custom PC with next-to-identical parts... the hardware is just as bad as their packaged software
Yes, it's a sad condition... What's really scary is that it applies to hardware, too.
:)
Back circa 1985, if you bought yourself a computer (Lets just say you get a Mac SE), the rate of failure out-of-the-box was extremely low. In the case of 40 particular Macintosh systems bought in 1985-1987, I know that only one of them needed hardware service at all by 1991. However, in 1993, the same company bought 13 new Macintosh systems (to replace some of those old Mac SE's that were STILL running), and had hardware problems with 3 of those 13 within a year.
Apple has statistically held a much better defect/failure rate than most IBM-clone PC manufacturers (the big ones, like Gateway, Dell, Compaq, etc), let alone all the independent stores building machines themselves.
I can say from personal experience (formerly being a system production manager for such a small company) that the return rates on computer hardware are absurdly high. Our store was among the best (sadly) with an average of a 10-12% return rate for components. If you statistically sum up all the 10-12% possible breakdowns of components in a computer, that should give you a clue as to how many of our full systems we saw back - more than 50% within the year. Talk about depressing...
Personally, I think that the software situation is worse in that respect, but it's just a lot easier to upgrade a program than it is to solder a new north bridge on your motherboard.
This is a horrible example, verging on a flame, but eliciting ideas is what i'm all about:
Do you think that, for example, OpenBSD could successfully authorize other developers' software (Lets use ISC BIND server as an easy example) without code changes? My opinion would be that they could not do this. The reason is that the OS team would have to FULLY audit every third-party program. They would have to report back possible bugfixes (to CVS, hopefully with commit priveleges), wait for another release version, and re-audit again! I believe it would tie down the OS developers too much in babysitting others' code to work on their own.
It would be a very cool idea if they could find a better-than-lukewarm way of doing this, but I for one do not see a feasible way of doing it.
I would agree strongly with Chocodile on, well, all of his points!
:)
OpenBSD is by no means any more difficult to learn or use than, say, RedHat. OpenBSD has very thoroughly and consisely documented information on their system, their methodologies, etc. Their tools are respectably inter-operable and close-knit. RedHat has a number of undocumented and/or poorly/obscurely documented system files, system utilities, and other core components that should be handled with better care.
I started using Debian, then went with Slackware (still my #1 Linux), tried FreeBSD (and HATED IT@##@$@), and then used OpenBSD. Currently there is ONE reason why OpenBSD is not on all of my servers (it's on half of them): support for only 8 character usernames. Its stability also exceeds that of Linux by orders of magnitude, with my hardware.
As far as workstations go, I have a SparcStation IPX (40mhz) that served as an X workstation until 3 weeks ago, when I got a Sparc 5/110 for next-to-free. I entirely didn't need to upgrade from the IPX - it was a fully useable workstation that ran surprisingly well. Not many operating systems can do that with a 40mhz machine.
I'll make one complaint (and it's probably not really OpenBSD's fault): On the SPARC architecture, I can't compile ORBit, thus I can't compile GTK, and thus I can't use xmms. That's my only complaint.
P.S.: Nice quote from The Refreshments. :)
And it takes a weaker mind to be influenced by entertainment.
You describe responses you had to films - These are great examples. If these films were to influence you or effect you for any longer than an emotional moment, perhaps you would find yourself having totalitarian, Nazi-like tendencies after watching Schindler's list. Perhaps you would feel very threatened and live with a more paranoid deameanor. Such a "response" would then merit the word "influence". Films rarely influence people on any grand scale. Almost everyone can mention a film that really did influence them, but these are individual events having to do with that individual's experiences.
Jumping around all out of order, I'd like to add a perspective to the quote "This child is rehearsing for life." It is true that they are rehearsing for life while they are playing games. Is it not true that they are experiencing and learning first-hand the unreality of their games? Are you saying that they are growing to understand these games as real? Rather, the contrary! They increasingly come to understand the fact that games are an escape from reality. They are entirely and wholly separate from reality. People feared that as games became more realistic these logical lines would fade. What was ignored there was that the human mind is adaptive, and adapts to the higher level of understanding that even these new, fancy, T&L-rendered games are still as unreal as Pong.
Back to the middle somewhere...
The pavlovian response you mention appears to be common. You can go to any teenage LAN party where Doom/Quake/etc is being played and see how much fun they have. It is easy to see this as a Pavlovian response to the blood. Have you considered that the reason such violence and such extreme action is so appealing is because of how UNREAL it is? If you play a game of Quake II, and the Q2 server sets the numgibs variable to 50 (instead of the default of like 8), You will literally see exploding meat when someone is killed. There is absolutely no doubt in my mind or the mind of ANY person I have met that the game is LESS real with all that additional gore and blood than it was before. It is the unreality, the need for escape, that drives this response you perceive as a "Pavlovian response."
Enough scattered ranting. You should be able to see the message I'm trying to get across: MANY people use games for an escape from reality. Actions in games are unreal and KNOWN to be unreal, and do not influence us or our children in the despicably elementary manners that video game opponents encourage us all to believe.
I am very very very close (probably too close) with a number of people at Sunbelt Software... Here's some useful info:
They have an Exchange server that they use (Quad Xeon 500, 2gb RAM, 200something gig RAID-5)... firstly, they don't even use "Real View". They hate it. They also hate their mail server, but since they are an NT-software distributor, they can't just fly off and switch to *nix and a real mail server.
Just to make a decent comparison, my single PII-266 with 64mb RAM can push 80% of the load that their Exchange server pushes... It runs Qmail on Linux. Hell, it's not even configured for speed (20 maximum instances).
I just wanted to let ya know that Sunbelt won't recommend that software to their friends... good to have an honest, inside view sometimes.
Here we have a hardware device which does almost exactly the same things that DeCSS does, with the same "feared" effect by the MPAA: being able to record mpeg-2 directly to your hard drive.
This seems like it would be in direct conflict with the MPAA's position on DeCSS... Is it possible that this product's release may cause a stir in the DeCSS case [ MPAA | 2600/Burns ], particularly in the view of copying digial media of which one already has a copy? Seems evident to me - hopefully the same sentiment will creep into the heads of everyone involved in the case.
The high school I attended was the cliche jock/prep/normal school in the local area. Like any school, it had a small population of non-conformists (10 to 20 out of 2100 students)... it just happened that this school had a very small group.
I met another member of this classification as a junior, as he was a freshman. A tall, lanky, evil "goth" (oh how ambiguous) looking 'freak' with plenty of piercings (about 20), he was the picturesque extreme of the school's 'freak' culture.
He wasn't liked by anyone, and that was his biggest advantage of all. He would get picked on, harassed, attacked, you name it. Instead of viewing these events as tragic causes for his victimization, he simply used them as motivation to shove their closed-minded views back in their faces.
He was never a 'politically correct' individual. In fact, he offered to piss on the principal's desk when the principal wouldn't let him use the restroom because there was cigarette smoke in there. Political correctness would have only made him an inert object, getting passed through the bowels of the school system.
He showed up for a whopping 60 of 180 days that first year, and by the end of it there wasn't a single person in the school who didn't know his name. He pulled some standard pranks, like stealing a golf cart and driving it through the halls, climbing out of fire escape windows during lunch... but those pranks only made his face known, not his name or attitude.
He was probably the only person who wouldn't think twice about screaming out his disagreements with the administration (normally something like "these fucking dimwits think it's reasonable to suspend me for telling a teacher she's a stupid bitch! If she wants to teach math, she needs to learn fucking multiplication!"), especially in front of the administration, in order to publicly provoke them and let their responses be known to the thousand students standing and watching.
The next year (my senior and his sophomore, er, um, well, freshman again), the situation escalated to the point of urban legend. He showed up for a record-low of 10 days throughout the 180-day school year, occasionally making additional appearances at lunches. He managed to receive more than 30 suspensions, despite his 10 day attendance. One particular day he had seen the principal thrice (throwing chairs on the cafeteria roof, walking in late and obnoxiously as hell into class, and (in another class) being painfully polite to a teacher). When he came in for the last visit (which baffled the principal), he received a 10 day suspension (maximum in this district, also guarantees failure for the semester) with the comment of "i'm just tired of seeing you."
His stupid antagonist pranks were the biggest problem for the administration... They were somehow foolhearty enough to ignore the message he was making clear to the other students: There are absurd restrictions and violations of freedom going on here, and the only way that you can make any change is to become active. Prove by example (which is how all U.S. Legal precedent is set) that their restrictions and censorships are not going to be tolerated.
About halfway through the year, a number of interesting changes started to happen. The school newspaper, previously known for typical school-censoring measures, published a handful of opinion columns that directly attacked and questioned the administration and their actions. One of the columns was regarding a friend of mine who was expelled for 'posession of a lethal weapon while under the influence of alcohol'. He lived far from school, actually outside of the school zone, and worked at the local grocery store. He had a 1-inch pocketknife (you can't even unscrew a case-screw with this flimsy POS) which he used for opening boxes. This was his 'lethal weapon'. Being "under the influence of alcohol" was based on a particular teacher of ours (i was sitting next to him) overhearing him talk about having a really bad hangover... She missed the part of him talking about a party we went to months before. He was never given a blood test, breathalizer test, walk-the-line-and-touch-your-nose test, or anything else. Just expelled.
There was a big uproar in the student population, with testimony coming from 4 students (one being myself), and the superintendent reversed the expulsion decision and (in more politically-correct terms) called the principal an idiot. He was. (He was also stoned off his ass every time I saw him, but that's a whole separate story!)
Back somewhere near topic... the newspaper actually published and distributed a column in regards to this expulsion and, at that time, pending reversal. It was crossing the line of polite, and deeply attacked the administration. Of course, the administration didn't like this and tried to retrieve all the released copies of the newspaper. The principal himself came into my classroom ("my" classroom was literal; i co-taught a 2D/3D modeling/rendering class which the most awesome teacher and I started that year) to collect the newspapers. This was also the first class of the day for my expelled friend. When he requested all the newspapers be returned, and some of the kids started reluctantly shuffling up the newspapers, I shook my head clear and started arguing with him. I knew that my testimony was _the_ primary evidence against my friend's expulsion, and since I was also co-teaching the class I was in, the principal couldn't afford to shoot himself in the foot and strongarm my argument. I asked him what he was trying to prove by covering up his actions, which were public record. I also offered to release the public information myself to all the students. I asked him if covering up this information would help his argument to the superintendent, and I (rather unwisely) threatened to flood the school with this article if he were to remove it. He didn't challenge me, and left.
Had it not been for my acquaintance (and now best friend and co-owner of a company) who had managed to become a 'King of Geeks and Freaks', my quiet, conservative, previously sheepish self would NEVER have made that stance against the principal and policy.
If I were to recommend something to help the situation of students being hassled like this, it would be to find a smart "freak/punk/geek/goth/etc" at your local school and tell them to _make_ the school system prove their policies, as my friend did. This does a LOT more than merely shake up the administration - it makes the other students aware that in the constitutionally-breaching school systems, they can still force their civil liberties to be respected, and gain the respect (if not fame) of their school class.
Start with improving the school environment, and the other problems (self-esteem of the outcasts, inter-student harassment, administrative violations of liberty) will become MUCH easier to improve.
No system can be fixed unless it is first fixed at its core. Avoid Fixing the surface issues - it does nothing but help the core problems rot in neglect.
I don't see how I could continue bantering on with someone when the purpose of argument is so drastically misconstrued between participants.
... it will serve no purpose other than to satisfy some pending need to vent about something (read: anything).
A quality argument cannot ensue when a broad point is recanted with overly specific, if not nit-picking, examples. I could turn around and nitpick equally well
My arguments of updated/outdated/etc were made to emphasise the structural and organizational differences in the distributions. It was construed as a much broader statement.
Enough of this - i can't expect to rationalize with someone who can change understandings of the overall point of the discussion arbitrarily.
Ideology: An operating system is an operating system. This is the identity property. Basic logic. An Operating System is NOT an office suite. An Operating System is NOT an mp3 player.
;>
My point is that the concept of making an operating system is that you build the core on top of which applications run. Slackware does not follow this ideology the way I like it (OpenBSD-style), but they do MUCH better than RedHat.
Quantity of software is also a null point. Firstly, the greater sludge of packages added by RedHat (same situation with debian) are rival programs striving to serve the same purpose. In fact, you may even end up with 3 packages that are just different versions of the same program! The ones that aren't duplicate are almost always outdated, oftentimes due to buffer overflows and other bugs. That is EXACTLY why RedHat distributions have so many security problems.
Follow the simple idea that you install your operating system and then get your application software from the proper sources. You end up with the latest releases (that work) and much fewer security problems.
It's a simple ideology issue. RedHat takes the AOL hodgepodge mentality. Slackware takes the erector-set mentality (a frame, get your own gear). I'll give you three guesses which the technically apt prefer.
IMO, actual power users don't use packages. They are the ones who MAKE the packages.
As far as mechanical quality on BMW bikes, you're correct, but that was not the point of that statement. My point was to illustrate the riders of such bikes. You have your BMW riders, who can't quite hack it on their own, and then you have your harley riders. If you're familiar with the biking scene this should be crystal clear.
Okay. Think of it like this...
BMW releases (yet another) new experimental motorcycle. Everyone who has been in the motorcycle industry for more than 10-15 years tells BMW to shove it, whilst the newer crew rants and raves at how great it is.
Harley-Davidson releases (this is rare) a new experimental motorcycle. Everyone who has been in the motorcycle industry for more than 10-15 years thanks them, whilst the newer crew bitches and moans.
You've got two distributions (Slackware, RedHat) who serve the same core purpose: make a distribution. They go about it in different ways and cater to different people. Slackware, in relativity to redhat, is geared more toward a power-user or server admin. RedHat is geared toward a new user or less experienced admin.
Redhat has had more than 142 security problems published on securityfocus.com in the last year. Slackware has had 14. Of course, these include older versions of the OS's, as well as some included third-party programs, but nonetheless it shows that RedHat's integrity in releasing solid PRODUCTION code is outright horrible, let alone their development. Slackware's record is far superior (and not without falter!), and the community in general respects Slackware for this qualitative approach of releasing a true alpha to the public for open development. They are not only supporting the community with a product, but also with the opportunity to participate in some good ole'fashioned development.
(just to nitpick and advocate Slackware (Praise Bob!), i'd like to make clear that ZERO of the vulnerabilities listed to effect slackware had to do with the implementation or compilation of the distribution. They were either GNU.org code bugs, kernel bugs, or third party software bugs.)
Development code is only as good as your developers... Says a lot more for Slackware's developers than it does RedHat's.
Some important notes from the developers:
- Slackware/SPARC is not ready for production.
- SILO is a bit broken and needs some work.
- many other packages are very unstable, if installable.
- changes for these problems will be occuring DAILY and WILL BE UPDATED AS SUCH on the rsync server. A cron job is in place on their rsync server to create a fresh ISO every night at midnight.
I'd offer an FTP site with the daily updates of Slack4Sparc, but my bandwidth can't handle it. If any of you have bandwidth to do this job, I can get you going with the updates.Props to my crew man Toby Freaksoft, though he tends to get a bit antsy about these sorts of things.
Whoever said slackware was dead should gingerly pry their cranium out of their colon. Thanks!
I live in Clearwater, FL and there are a SLEW of internet providers. I literally have my pick of 100 or more ISP's. I work as a Network Engineer now, and previously built a local ISP from the ground up.
The company I choose to do business with now is called Intelligence Network. They have a strict policy of using only Cisco routers and switches, and only Sun Microsystems workstations and servers. Personally, I like playing with old Sun hardware, but I greatly disagree with many of Sun's policies regarding licensing. Nevertheless, no informed individual in the business can say that your average Intel/Bay Networks enterprise-level network can even compare with your average Sun/Cisco network. That's Point 1: Quality.
Secondly, they are a small-to mid-sized company. They have four static T-1's and a burstable OC-12. I'd prefer they had a burstable T-3 instead of the four T-1's, but nevertheless, that is again one of the signs of a smaller, more personalized company.That's Point 2: A small, non-conglomerate company.
Next, they have little to no advertising. They rely on the quality of their service to keep their business at a solid level. They are not in the business seeking profits (or else they wouldn't be spending well over $1,000,000 per server they have), but rather they are in the business to provide a purely quality-oriented service. Point 3: No major advertising.
Lastly, their tech support is good. This is the hardest to gauge and monitor. They have perhaps a dozen technicians. Some are assigned to dialup service, some to DSL and T1/E1 service, and some to colocation and other services. Atop this they have 3 head technicians who actually comminicate directly with the customer. In fact, they're so good with communication that they gave me a free 10mbit SBUS ethernet card for an old Sparc IPX i've got for a firewall. They are the kind of company that wants to get personal with the customer and actually develop a relationship with them. Point 4: Personal Customer relationships.
Okay, one more point. This one is very arguable and lies entirely in my preference. I like this company because they do NOT try to match the bottom-price market. Instead, they charge a fair price for a very reliable service. They are not motivated by high-volume income, and it is very clear by the quality and (relatively small) size of their company. I look for this with every business or service I deal with.
For you performance geeks out there, my DSL connection has had a total of approximately 2 hours of downtime in the last year. I monitor my uptime by 30-second intervals and keep it logged in MRTG-style reports. Those two hours were an accumulation of many very short service failures, many of which I proved were due to AT&T doing maintenance at 4am on Sunday morning every other month. Nearly all of the rest of it, short of approximately 20 minutes, was due to GTE/Verizon (many bad things to say about them). Their end result from these points is an extremely reliable service, and my performance is ALWAYS 99% or better of my bandwidth provisioning.
If you want a _quality_ provider, open up your local phone book and pick a handful of ISP's you have never heard of in your local area. Call each of them and ask them two simple questions:
- What hardware do you use to run your network and servers?
- How much bandwidth do you have, and from what providers?
If they have the bandwidth to realistically support the performance they claim, and they are using quality hardware, you've got a company that meets three of the above points: Quality, Small Company, and NO Major Advertising. At that point, there's only one way to really test their customer service, and that's to use their service.Don't expect your "cheapest and fastest" advertising ISP to get you anywhere. Throw in a few more dollars and ignore the claimed performance gains, and deal with a company that actually wants to provide service.