All this port 1024 business is is just a default setting, and defaulting to locking down a sensitive port range makes a lot more sense than defaulting to having it open.
Wouldn't a decent access control implementation like GRsec or SELinux be able to open or close port ranges independently of the process's euid? I'm not experienced with either one, so I don't know if they can, but it seems reasonable to me.
If they're the same message, they should have the same Message-id, and even if it's split up by qmail, it might conceivably be the province of the MUA or the local MDA to collapse them into a single copy. Arguably qmail isn't helping, but it is pretty ambiguous who has what responsibility here.
Your third allegation was true until the publication of this PDF which you obviously didn't read since it included a dedication of qmail to the public domain.
Did you forget a link around "This PDF"? Because the licensing page still has the old terms.
I believe the multiple copies business is thanks to VERP. Which was a good idea, executed poorly, to the point of being abusive to the resources of receivers. But don't tell that to djb unless you want to get yourself banned from his litle fiefdom. Dan Bernstein makes Theo de Raadt look like the Queen of Nice and a master of reasoned diplomacy.
I think the only major qmail installation left these days is Yahoo.
Now if I could find a reasonable alternative to djbdns... I'm sure there's something out there, but the alternatives just don't seem to get the same exposure.
There isn't just "talk of real-time gc", it's already out there.
Pardon moi if I would rather not deal with someone forgetting to free mallocs because they decided to shave a few microseconds off something already running within 50 milliseconds of deadline. Especially when the controller is buried under 10 tons of concrete.
Anyway, the research that's all the rage these days is static garbage collection. It works just like manually freeing memory, but you never have to actually do it yourself. It just inserts the proper free statements at the point of execution where it can prove that your data is unreachable. Systems like this tend to be a bit more conservative (not to be confused with conservative collectors) and might hang on to your data a bit longer, but how many times have I seen freed memory dereferenced later?
Yeah, it's exactly what happens with MacOS, if you do something completely different than what was described. Just redefine reality until Apple is right, and the rainbows will still sing happy tunes.
I've actually had jobs doing honest to goodness filing, as an intake secretary while doing social work (a weird time for me for sure). Every time I got a stack of case folders, I was generally expected to merge them with any existing folders. Sure as hell not destroy the old ones.
> This would be a great demonstration of the value of "Time Machine" backups.
Yeah, but it would also be a great demonstration of how the competition doesn't have this behavior at all. If my main chute tears like a nylon stocking, I'm not exactly one to trust the backup chute made by the same people, dig?
> Isn't it standard UNIX to delete the target before moving a folder to an already used name?
Unix most certainly merges directories, and doesn't delete them.
The Mac behavior basically goes back to the original HFS, which technically didn't really have directories at all, just a UI that faked them. A folder was just a file with a list of files in it, so naturally the folder would be overwritten by such operations. Nowadays, the fact that MacOS still doesn't even give you the option to move the freaking conflict out of the way with an automatic rename, let alone the merge many people expect, strikes me as stubborn hanging on to an anachronistic misfeature, all in the name of removing the confusion of choice in the UI.
True, but neither does Jack Thompson. Then again, the people who have to be taught to ignore him are the media wonks. But in JT, they have a "source" that will make newsy-opiniony noises on cue, and they never like giving those up.
> Thompson said, "no one in their right mind would say that a videogame by itself would turn an angel into a demon"
True, but that first clause about being on one's right mind is a real sticky wicket, ain't it, Jack?
In seriousness, he's been acting a lot more rational these days. Maybe the disciplinary hearing before the Florida Bar that he's facing this month has sobered him up?
Seriously though, I remember this sort of bullshit "research" 10 years ago on MOOs. Sure there's room for legitimate research into social dynamics of virtual environments, but I always wondered what kind of state academia was in when this sort of methodology was considered acceptable. It reminds me of the sort of high school science fair exhibits like "comparing the effects of classical vs rock music on insects".
Power supplies are one of the *least* likely components to fail: memory is the most vulnerable, CPU's are second especially with cooling issues.
Holy freakin shit. CPUs? Have you ever actually worked with hardware? The last time I had a CPU actually burn itself up was when I was testing experimental Sparc 15MP CPUs. Those things were on riser cards, stacked above each other.
Memory gets zapped when you install it. ESD is a bitch, even if you wear a strap. On most utility boxes, memory doesn't get installed, you just get more memory when you get a new server. Bigger servers with >4gig are a different matter, but are perhaps a tenth of total servers in most places, which usually fill up racks with Dell 1950's or the like.
PSUs go all the time. I've learned to get redundant PSUs whenever possible. The only thing that fails more often than CPUs are HDD's... which you didn't even mention.
"Failure to use thermal compound"... geez louise, your overclocked gamer rig is not what they fill datacenters and labs with.
> You don't protect trademarks, you lose them, the end. And you know it.
You don't quite lose them. What you do lose is the ability to claim damages due to dilution. You still get to pursue blatant infringement -- Kimberly-Clark could still sue your cojones off if you sold paper products called "Kleenex".
Slashvertisement, in its most distilled form. I guess the "editorship" here wrenched their shoulders after patting themselves on the back during their tenth anniversary. So much for integrity.
Seriously, even though I know all too well how running something like slashdot is a lot harder than it looks, and how not everyone can be satisfied, and how quality sometimes has to come after candor, even after all that, I know deep down I actually could start something better than this dreck. But frankly, "social links" and blog aggregators are already out there, and I won't pour my money down the hole of recreating reddit, digg, or technorati.
This article shows precisely how slashdot is not only not journalism, it's not even a respectable blog. Slashdot occupies the medium precisely inbetween, known colloquially as "The Worst of Both Worlds." You should be ashamed . But I know you aren't.
Actually, RMS has insisted on full sources being available at the point of distribution, not just diffs. He was really quite inflexible on this point with regards to the sole SuperH developer for gcc at the time (I think it was SuperH), and actually chased him off the project because the bandwidth and storage costs for it at the time were infeasable for the developer to accede to the demands. Obviously things have evolved since then (and RMS no longer personally does GPL enforcement), but it does show that the interpretation could go either way.
Now there's certainly nothing stopping redhat from making updated SRPM availability contingent on having a copy of RHEL, though they can't stop them from being redistributed. Redhat figures that if all it takes is a "leak" to redistribute, they may as well not try to restrict it anyway -- they just don't loudly advertise it.
Redhat's largely a services company. RHN is semi-decent for asset tracking to see what servers need what updates at a glance. Nothing like what Sun offers, but certainly a fraction of the price.
> A 1U server is also commonly called a "pizza box" because of its shape (although they're generally rectangular rather than square).
I always heard "Pizza Box" in describing Sun Sparc 10's and other desktop boxes with similar form factors. The 1U rack form factor was just "rackable" -- they're really way too big to call 'em pizza boxes. Maybe Little Ceasar's "pizza pizza" boxes.
Slashdot Mirror
Should have been "port < 1024". Good old slashdot and its web-0.9 post editor.
All this port 1024 business is is just a default setting, and defaulting to locking down a sensitive port range makes a lot more sense than defaulting to having it open.
Wouldn't a decent access control implementation like GRsec or SELinux be able to open or close port ranges independently of the process's euid? I'm not experienced with either one, so I don't know if they can, but it seems reasonable to me.
If they're the same message, they should have the same Message-id, and even if it's split up by qmail, it might conceivably be the province of the MUA or the local MDA to collapse them into a single copy. Arguably qmail isn't helping, but it is pretty ambiguous who has what responsibility here.
Your third allegation was true until the publication of this PDF which you obviously didn't read since it included a dedication of qmail to the public domain.
Did you forget a link around "This PDF"? Because the licensing page still has the old terms.
I believe the multiple copies business is thanks to VERP. Which was a good idea, executed poorly, to the point of being abusive to the resources of receivers. But don't tell that to djb unless you want to get yourself banned from his litle fiefdom. Dan Bernstein makes Theo de Raadt look like the Queen of Nice and a master of reasoned diplomacy.
I think the only major qmail installation left these days is Yahoo.
Now if I could find a reasonable alternative to djbdns... I'm sure there's something out there, but the alternatives just don't seem to get the same exposure.
There isn't just "talk of real-time gc", it's already out there.
Pardon moi if I would rather not deal with someone forgetting to free mallocs because they decided to shave a few microseconds off something already running within 50 milliseconds of deadline. Especially when the controller is buried under 10 tons of concrete.
Anyway, the research that's all the rage these days is static garbage collection. It works just like manually freeing memory, but you never have to actually do it yourself. It just inserts the proper free statements at the point of execution where it can prove that your data is unreachable. Systems like this tend to be a bit more conservative (not to be confused with conservative collectors) and might hang on to your data a bit longer, but how many times have I seen freed memory dereferenced later?
> Um. What exactly do you think a directory is on every other posix filesystem in the world?
A special file with dirents in it that gets treated specially by the filesystem layer in all the relevant syscalls. Rather unlike HFS.
vi / gives me a dired interface, but yeah, just doing a cat on a directory dumps the raw bits. Well, these days you have to force it.
Yeah, it's exactly what happens with MacOS, if you do something completely different than what was described. Just redefine reality until Apple is right, and the rainbows will still sing happy tunes.
I've actually had jobs doing honest to goodness filing, as an intake secretary while doing social work (a weird time for me for sure). Every time I got a stack of case folders, I was generally expected to merge them with any existing folders. Sure as hell not destroy the old ones.
> This would be a great demonstration of the value of "Time Machine" backups.
Yeah, but it would also be a great demonstration of how the competition doesn't have this behavior at all. If my main chute tears like a nylon stocking, I'm not exactly one to trust the backup chute made by the same people, dig?
> Isn't it standard UNIX to delete the target before moving a folder to an already used name?
Unix most certainly merges directories, and doesn't delete them.
The Mac behavior basically goes back to the original HFS, which technically didn't really have directories at all, just a UI that faked them. A folder was just a file with a list of files in it, so naturally the folder would be overwritten by such operations. Nowadays, the fact that MacOS still doesn't even give you the option to move the freaking conflict out of the way with an automatic rename, let alone the merge many people expect, strikes me as stubborn hanging on to an anachronistic misfeature, all in the name of removing the confusion of choice in the UI.
> Ignoring people doesn't promote healthy debate
True, but neither does Jack Thompson. Then again, the people who have to be taught to ignore him are the media wonks. But in JT, they have a "source" that will make newsy-opiniony noises on cue, and they never like giving those up.
> Thompson said, "no one in their right mind would say that a videogame by itself would turn an angel into a demon"
True, but that first clause about being on one's right mind is a real sticky wicket, ain't it, Jack?
In seriousness, he's been acting a lot more rational these days. Maybe the disciplinary hearing before the Florida Bar that he's facing this month has sobered him up?
> Why was a bot used for this anyway?
Perhaps because its behavior is consistent.
Seriously though, I remember this sort of bullshit "research" 10 years ago on MOOs. Sure there's room for legitimate research into social dynamics of virtual environments, but I always wondered what kind of state academia was in when this sort of methodology was considered acceptable. It reminds me of the sort of high school science fair exhibits like "comparing the effects of classical vs rock music on insects".
> this raises the question
I thought it begged the question. I swear, the world is loosing its ability to speak good.
> So it looks like there's only 28 people left in Second Life?
No, there were only 28 connected who weren't engaging in furry/bdsm cybering.
Power supplies are one of the *least* likely components to fail: memory is the most vulnerable, CPU's are second especially with cooling issues.
... geez louise, your overclocked gamer rig is not what they fill datacenters and labs with.
Holy freakin shit. CPUs? Have you ever actually worked with hardware? The last time I had a CPU actually burn itself up was when I was testing experimental Sparc 15MP CPUs. Those things were on riser cards, stacked above each other.
Memory gets zapped when you install it. ESD is a bitch, even if you wear a strap. On most utility boxes, memory doesn't get installed, you just get more memory when you get a new server. Bigger servers with >4gig are a different matter, but are perhaps a tenth of total servers in most places, which usually fill up racks with Dell 1950's or the like.
PSUs go all the time. I've learned to get redundant PSUs whenever possible. The only thing that fails more often than CPUs are HDD's... which you didn't even mention.
"Failure to use thermal compound"
> CentOS is the Fedora that Fedora should have been.
Erm, Redhat is the Fedora that Fedora was 6 months ago. Fedora exists for that reason.
> You don't protect trademarks, you lose them, the end. And you know it.
You don't quite lose them. What you do lose is the ability to claim damages due to dilution. You still get to pursue blatant infringement -- Kimberly-Clark could still sue your cojones off if you sold paper products called "Kleenex".
Yunno, I'm not one to complain about moderation, but how the fuck do you justify defending slashdot here?
Slashvertisement, in its most distilled form. I guess the "editorship" here wrenched their shoulders after patting themselves on the back during their tenth anniversary. So much for integrity.
Seriously, even though I know all too well how running something like slashdot is a lot harder than it looks, and how not everyone can be satisfied, and how quality sometimes has to come after candor, even after all that, I know deep down I actually could start something better than this dreck. But frankly, "social links" and blog aggregators are already out there, and I won't pour my money down the hole of recreating reddit, digg, or technorati.
This article shows precisely how slashdot is not only not journalism, it's not even a respectable blog. Slashdot occupies the medium precisely inbetween, known colloquially as "The Worst of Both Worlds." You should be ashamed . But I know you aren't.
> isn't the GPL kinda a contract?
Only "kinda". It's a license, and is covered by copyright law, and not contract law. They overlap a bit, but not as much as you might think.
Whoops, it was Coldfire, not SuperH.
Actually, RMS has insisted on full sources being available at the point of distribution, not just diffs. He was really quite inflexible on this point with regards to the sole SuperH developer for gcc at the time (I think it was SuperH), and actually chased him off the project because the bandwidth and storage costs for it at the time were infeasable for the developer to accede to the demands. Obviously things have evolved since then (and RMS no longer personally does GPL enforcement), but it does show that the interpretation could go either way.
Now there's certainly nothing stopping redhat from making updated SRPM availability contingent on having a copy of RHEL, though they can't stop them from being redistributed. Redhat figures that if all it takes is a "leak" to redistribute, they may as well not try to restrict it anyway -- they just don't loudly advertise it.
Redhat's largely a services company. RHN is semi-decent for asset tracking to see what servers need what updates at a glance. Nothing like what Sun offers, but certainly a fraction of the price.
> A 1U server is also commonly called a "pizza box" because of its shape (although they're generally rectangular rather than square).
I always heard "Pizza Box" in describing Sun Sparc 10's and other desktop boxes with similar form factors. The 1U rack form factor was just "rackable" -- they're really way too big to call 'em pizza boxes. Maybe Little Ceasar's "pizza pizza" boxes.