This is why there are kernel modules. As much as linux ricers like to argue otherwise, there's virtually no reason a normal end user should ever build their own kernel. Nor should their be. The idea that compiling a kernel should ever be optimized for average joe end users is stupid.
Oh? What OS does Red Hat make? They just sell services for a free one
I seem to remember Red Hat making Red Hat Enterprise Linux and Red Hat Desktop. Could be that I'm getting old though, ya know.
Mandriva? Who the fuck is that?
The company formally known as Mandrake. Despite a debt reogranization in their not distant past, they're (somewhat suprisingly) turning a profit now.
Sun does not sell a consumer-level OS. They sell servers.
You said "selling operating system" not "selling [presumably retail?] consumer operating systems". And yes, they sell servers, but they have also become increasingly aggressive about pushing Solaris as a standalone product on Intel hardware.
Novell? Hahaha! Yeah, they are doing GREAT! Woah, nelly! I bet Steve Balmer trembles in his sleep over the runaway popularity of the Novell OS.
Over a billion dollars revenue last year, and profitable. Or were you using some esoteric meaning of the phrase "making money" that I'm not aware of?
(And for the record, there is no "Novell OS". There's Netware, Linux Enterprise Server, and Linux Desktop.)
And on a broader front, are you really prepared to trust a company like Kelkea, Inc. (owners of MAPS) to decide what emails gets to you without really knowing how they operate and deal with resolution processes?
There's a reason I stick to Spamhaus as the sole RBL at work (and at home) - professionalism. They spell out criteria and rationale clearly on their website. They list only IPs, rather than blindly blocking entire netblocks or domains. The delisting policy is incredibly liberal by default, but temper that by tracking repeat offenders. And (this is where a _lot_ of lists fall down) they assign a TTL to every entry and automatically expire the entries even if the owner doesn't report a resolution.
We block millions of messages a day based on the SBL/XBL lists and have, to date, recieved only one query from a client about why a particular message was blocked, and it turned out the recipient had a worm outbreak that got them places on the XBL. The block had been lifted before it even made it to our support team.
That only works if the competition doesn't suck as bad or worse. There are a small number of top-tier hardware vendors, and IME they all suck to more or less the same extent. The only difference is they suck in different ways at different times.
The "everyone sucks" mantra works for the build your own approach too, it just shifts to hardware and RMA instead of hardware and service. And frankly, most in-house IT departments I've dealt with are as grossly incompetent as first tier support at a repudable vendor.
Switching vendors probably won't get you any better service -- it will just complicate your support problems, as now you have 2 different sets of machines to support instead of just one.
Another advantage to leasing. Drop the vendor at the end of the lease period. Mind you, if the company is continuously adding new equipment that complicates things, but then you can switch alliance at the next model churn, when you'd have to move to a new machine type anyways.
Wow, who the hell have you been leasing through? Spanky's Uber Rad Discount Computer Warehouse?
Leased hardware isn't yours (by definition), which places certain constraints. For example, it'll usually require an off-site technician from the company leasing out the hardware to do any maintenance.
None of our vendors have ever complained about us doing field swaps from our spare pool. They just ask for the part number of the failed part, give us an RMA number, and advance ship out the part and return shipping label for the broken hardware.
The second constraint is that those doing the maintenance have no ties to you, which mean that they don't have to do anything effective.
Aside from "Wow, your service sucks. We'll be buying from competitors in the future."
The third problem is that you're likely to get refurbished equiptment with an unknown history and minimal to no quality control.
Again, who the hell have you been leasing through? We've gotten exactly one defective field repair unit in the five years I've been working at this company.
Even if there were checks, though, reliability is an unknown.
You know what? That applies to brand new hardware as well. Reputable companies usually test refurb units more thoroughly the new equipment.
How many enterprises have only one server? We're pushing something like eighty Linux servers here.
(Mind you, I'm still pushing for RHEL4 on our next platform revision, but it's easy to imagine that if the company were more cash-strapped that a $27,600 savings would be quite attractive.)
I have at least another fifty CDs I've yet to encode because I'm out of space.
I'll likely be looking to build a bare minimum of a terabyte over the next year, in anticipation of having some place to store digital video of my children once they're born.
Re:Price is important to many people.
on
Return of the Mac
·
· Score: 1
So far I've only heard of people upgrading third-party to 5400rpm, and a couple of side remarks by people saying they thought they were 5400rpm. Friend of mine who does service at CompUSA has only come across the slower 4200s.
In any case, even 5400 is on the low end for desktop drives these days. I'm pretty sure Apple only uses them in iMac and eMacs, and Dell has bumped up to 7200 on even their entry level models.
Re:Keyboard included at your house
on
Return of the Mac
·
· Score: 1
Oh, yeah, because searching for adapters and taking a knife to your keyboard really fits in with the "It Just Works" mantra.
Not sure why you're mentioning monitor prices, since I didn't bring them up at all. Though I'd guess a lot of them are pricing LCD, not CRT, since Dell and other PC vendors are pretty regularly running promotions with a free upgrade to LCD.
Personally, I think Apple's making a mistake in not offering a more modest monitor themselves. A fair number of people are going to be turned off by the prospect of not being able to single-source. You and I might know that pretty much any monitor will work, but it'll make a non-technical buyer nervous.
I'm somewhat ambivilent, myself, on the value of the Minimac, overall. If you really do have spare components lying around, it's not bad in terms of cost, but under those circumstances the "but think of the software!" argument is a lot less compelling, since someone with a pre-existing PC system is likely to have a pre-existing investment in PC software. And that has the highly compelling benefit of being what the person already knows. (Never, ever discount that. I know plenty of people who are still running long since abandoned Win 3.1 apps because they see no good reason why they should learn something new.)
It's not a lack of money, it's realizing there are other things I can spend my money on for my family rather than frittering it away on a particular brand of computer.
(And not that it matters, but I do have a work provided machine, along with wired and wireless network access. That doesn't account for all the other machines in the house.)
Re:Price is important to many people.
on
Return of the Mac
·
· Score: 1
The Mac Mini is no different from any other G4 Mac.
Really? I wasn't aware Apple was shipping 4200RPM hard drives in any other machine. Oh wait, they're not. Swapping is a hell of a lot painful on a notebook drive.
My home PC is 5 years old, and cost me around $700. And the problem isn't coming up with $15 a month for most people, it's coming up with $3000 in one shot.
Some of us have more important things to spend money on - wives, children, homes, et cetera.
Re:Keyboard included at your house
on
Return of the Mac
·
· Score: 1
Odds are you already own a keyboard, mouse, and monitor. The keyboard and mouse are USB, and the monitor output requires a $20 dongle.
Most PCs sold in the last five years come with PS/2 - not USB - peripherals. And even if one does have a USB keyboard around, the fact of the matter is that not having the Mac keys on there is going to be confusing for most consumers.
And of course the "switcher" assumption seems to be that people just throw out their old PCs or something. Virtually everyone I know who retires a machine gives the old one away, including all the peripherals.
My high school average was over 98% and I'll echo the sentiment. To be fair, a degree from a credible school is damn handy in most professions, particularly starting out, but I know ivy leaguers who've been out of work for over a year, as well as high school dropouts making way above the industry average.
With a smart employer job performance trumps academic performance any day.
Not by a long shot. There have been a large number of discovered violations, most of which have been settled without a legal battle.
Most of the recent ones have involved the iptables/netfilter code. Among the companies who have settled after being cited for violations are Gigabyte, Siemens, Fujitsu-Siemens, Asus, and Belkin.
Another fairly recent one was the folks behind MPlayer discover KiSS Technology were using both their code and several other GPL products in their video players.
That benchmark suffers from the common tendancy of FireWire advocates to measure only large sustained throughput. A number of benchmarks, like this one show USB 2.0 at a signifigant advantage when doing smaller reads.
Both have their advantages, and their place. The differences are likely inconsequential for the common use case - i.e. load up a bunch of crap, add and change incrementally from there on in - and packaging the cable for the nigh ubiquitous technology in the box makes a lot of sense from a cost reduction standpoint.
They introduce regular new releases every eighteen months for the people who want the latest and greatest. And under their enterprise license you have the right to run any currently supported version, including newly introduced ones.
I find most people who've never done large scale administration highly overestimate the value of new versions. Especially when there's the option of a stable platform with all the relevent security updates.
Most businesses I've dealt with are prone to be lax with regards to security, even, if it means they can avoid the expense of disruptive upgrades. In many cases we're talking about platforms that haven't seen updates in over a decade.
Sure, it's true that new features have value. But then again, if what you have works, is it really worth it? And how much do the important things really change? Web servers are still webservers and even fairly ancient platforms can run the latest version of apache. File servers still serve files. Mail servers still serve mail. Blah blah blah.
Except they're really fucking useless to those of us are colour-blind, especially since they're horizontally instead of vertically oriented, and are all always "lit".
Not to mention that traffic lights are indicators and not controls, so it's hardly a ripe analogy to begin with.
Yeah, it seems the project is at least somewhat alive. Really doesn't help give the appearance of viability, though, when both the.com and.net addresses have that as their only message on the front page. Kind of illustrates how important control of the name can be, actually.:)
Slashdot Mirror
This is why there are kernel modules. As much as linux ricers like to argue otherwise, there's virtually no reason a normal end user should ever build their own kernel. Nor should their be. The idea that compiling a kernel should ever be optimized for average joe end users is stupid.
Oh? What OS does Red Hat make? They just sell services for a free one
I seem to remember Red Hat making Red Hat Enterprise Linux and Red Hat Desktop. Could be that I'm getting old though, ya know.
Mandriva? Who the fuck is that?
The company formally known as Mandrake. Despite a debt reogranization in their not distant past, they're (somewhat suprisingly) turning a profit now.
Sun does not sell a consumer-level OS. They sell servers.
You said "selling operating system" not "selling [presumably retail?] consumer operating systems". And yes, they sell servers, but they have also become increasingly aggressive about pushing Solaris as a standalone product on Intel hardware.
Novell? Hahaha! Yeah, they are doing GREAT! Woah, nelly! I bet Steve Balmer trembles in his sleep over the runaway popularity of the Novell OS.
Over a billion dollars revenue last year, and profitable. Or were you using some esoteric meaning of the phrase "making money" that I'm not aware of?
(And for the record, there is no "Novell OS". There's Netware, Linux Enterprise Server, and Linux Desktop.)
Sun Microsystems, Red Hat, Novell, Mandriva... Or did all those companies cease to exist last night while I was sleeping?
Um, he didn't agree with the spirit of the agreement, and by not using the product, wasn't party to it.
Yeah, just like that evil Linus stole the design for Unix. Thank God SCO is litigating that.
Obvious answer is anyone doing web design.
And on a broader front, are you really prepared to trust a company like Kelkea, Inc. (owners of MAPS) to decide what emails gets to you without really knowing how they operate and deal with resolution processes?
There's a reason I stick to Spamhaus as the sole RBL at work (and at home) - professionalism. They spell out criteria and rationale clearly on their website. They list only IPs, rather than blindly blocking entire netblocks or domains. The delisting policy is incredibly liberal by default, but temper that by tracking repeat offenders. And (this is where a _lot_ of lists fall down) they assign a TTL to every entry and automatically expire the entries even if the owner doesn't report a resolution.
We block millions of messages a day based on the SBL/XBL lists and have, to date, recieved only one query from a client about why a particular message was blocked, and it turned out the recipient had a worm outbreak that got them places on the XBL. The block had been lifted before it even made it to our support team.
That only works if the competition doesn't suck as bad or worse. There are a small number of top-tier hardware vendors, and IME they all suck to more or less the same extent. The only difference is they suck in different ways at different times.
The "everyone sucks" mantra works for the build your own approach too, it just shifts to hardware and RMA instead of hardware and service. And frankly, most in-house IT departments I've dealt with are as grossly incompetent as first tier support at a repudable vendor.
Switching vendors probably won't get you any better service -- it will just complicate your support problems, as now you have 2 different sets of machines to support instead of just one.
Another advantage to leasing. Drop the vendor at the end of the lease period. Mind you, if the company is continuously adding new equipment that complicates things, but then you can switch alliance at the next model churn, when you'd have to move to a new machine type anyways.
Wow, who the hell have you been leasing through? Spanky's Uber Rad Discount Computer Warehouse?
Leased hardware isn't yours (by definition), which places certain constraints. For example, it'll usually require an off-site technician from the company leasing out the hardware to do any maintenance.
None of our vendors have ever complained about us doing field swaps from our spare pool. They just ask for the part number of the failed part, give us an RMA number, and advance ship out the part and return shipping label for the broken hardware.
The second constraint is that those doing the maintenance have no ties to you, which mean that they don't have to do anything effective.
Aside from "Wow, your service sucks. We'll be buying from competitors in the future."
The third problem is that you're likely to get refurbished equiptment with an unknown history and minimal to no quality control.
Again, who the hell have you been leasing through? We've gotten exactly one defective field repair unit in the five years I've been working at this company.
Even if there were checks, though, reliability is an unknown.
You know what? That applies to brand new hardware as well. Reputable companies usually test refurb units more thoroughly the new equipment.
How many enterprises have only one server? We're pushing something like eighty Linux servers here.
(Mind you, I'm still pushing for RHEL4 on our next platform revision, but it's easy to imagine that if the company were more cash-strapped that a $27,600 savings would be quite attractive.)
[galt@damballah music]$ du -sh . .
140G
[galt@damballah music]$ find . -name \*.flac | wc -l
4794
[galt@damballah music]$ find . -name \*.mp3 | wc -l
4228
I have at least another fifty CDs I've yet to encode because I'm out of space.
I'll likely be looking to build a bare minimum of a terabyte over the next year, in anticipation of having some place to store digital video of my children once they're born.
So far I've only heard of people upgrading third-party to 5400rpm, and a couple of side remarks by people saying they thought they were 5400rpm. Friend of mine who does service at CompUSA has only come across the slower 4200s.
In any case, even 5400 is on the low end for desktop drives these days. I'm pretty sure Apple only uses them in iMac and eMacs, and Dell has bumped up to 7200 on even their entry level models.
Oh, yeah, because searching for adapters and taking a knife to your keyboard really fits in with the "It Just Works" mantra.
Not sure why you're mentioning monitor prices, since I didn't bring them up at all. Though I'd guess a lot of them are pricing LCD, not CRT, since Dell and other PC vendors are pretty regularly running promotions with a free upgrade to LCD.
Personally, I think Apple's making a mistake in not offering a more modest monitor themselves. A fair number of people are going to be turned off by the prospect of not being able to single-source. You and I might know that pretty much any monitor will work, but it'll make a non-technical buyer nervous.
I'm somewhat ambivilent, myself, on the value of the Minimac, overall. If you really do have spare components lying around, it's not bad in terms of cost, but under those circumstances the "but think of the software!" argument is a lot less compelling, since someone with a pre-existing PC system is likely to have a pre-existing investment in PC software. And that has the highly compelling benefit of being what the person already knows. (Never, ever discount that. I know plenty of people who are still running long since abandoned Win 3.1 apps because they see no good reason why they should learn something new.)
It's not a lack of money, it's realizing there are other things I can spend my money on for my family rather than frittering it away on a particular brand of computer.
(And not that it matters, but I do have a work provided machine, along with wired and wireless network access. That doesn't account for all the other machines in the house.)
The Mac Mini is no different from any other G4 Mac.
Really? I wasn't aware Apple was shipping 4200RPM hard drives in any other machine. Oh wait, they're not. Swapping is a hell of a lot painful on a notebook drive.
My home PC is 5 years old, and cost me around $700. And the problem isn't coming up with $15 a month for most people, it's coming up with $3000 in one shot.
Some of us have more important things to spend money on - wives, children, homes, et cetera.
Odds are you already own a keyboard, mouse, and monitor. The keyboard and mouse are USB, and the monitor output requires a $20 dongle.
Most PCs sold in the last five years come with PS/2 - not USB - peripherals. And even if one does have a USB keyboard around, the fact of the matter is that not having the Mac keys on there is going to be confusing for most consumers.
And of course the "switcher" assumption seems to be that people just throw out their old PCs or something. Virtually everyone I know who retires a machine gives the old one away, including all the peripherals.
Visual Basic and Java and Python and Oberon and Object Pascal and Boo and Nemerle and Component Pascal and Forth and Lisp and Smalltalk and Logo and Tachy.
Some of the packages are still in the larval stages, but they're out there, and more are on the horizon.
My high school average was over 98% and I'll echo the sentiment. To be fair, a degree from a credible school is damn handy in most professions, particularly starting out, but I know ivy leaguers who've been out of work for over a year, as well as high school dropouts making way above the industry average.
With a smart employer job performance trumps academic performance any day.
Not by a long shot. There have been a large number of discovered violations, most of which have been settled without a legal battle.
Most of the recent ones have involved the iptables/netfilter code. Among the companies who have settled after being cited for violations are Gigabyte, Siemens, Fujitsu-Siemens, Asus, and Belkin.
Another fairly recent one was the folks behind MPlayer discover KiSS Technology were using both their code and several other GPL products in their video players.
That benchmark suffers from the common tendancy of FireWire advocates to measure only large sustained throughput. A number of benchmarks, like this one show USB 2.0 at a signifigant advantage when doing smaller reads.
Both have their advantages, and their place. The differences are likely inconsequential for the common use case - i.e. load up a bunch of crap, add and change incrementally from there on in - and packaging the cable for the nigh ubiquitous technology in the box makes a lot of sense from a cost reduction standpoint.
They introduce regular new releases every eighteen months for the people who want the latest and greatest. And under their enterprise license you have the right to run any currently supported version, including newly introduced ones.
I find most people who've never done large scale administration highly overestimate the value of new versions. Especially when there's the option of a stable platform with all the relevent security updates.
Most businesses I've dealt with are prone to be lax with regards to security, even, if it means they can avoid the expense of disruptive upgrades. In many cases we're talking about platforms that haven't seen updates in over a decade.
Sure, it's true that new features have value. But then again, if what you have works, is it really worth it? And how much do the important things really change? Web servers are still webservers and even fairly ancient platforms can run the latest version of apache. File servers still serve files. Mail servers still serve mail. Blah blah blah.
Except they're really fucking useless to those of us are colour-blind, especially since they're horizontally instead of vertically oriented, and are all always "lit".
Not to mention that traffic lights are indicators and not controls, so it's hardly a ripe analogy to begin with.
Yeah, it seems the project is at least somewhat alive. Really doesn't help give the appearance of viability, though, when both the .com and .net addresses have that as their only message on the front page. Kind of illustrates how important control of the name can be, actually. :)