++One more reason I have observed is that people get used to a particular platform. More often than not, a commercial vendor enters a market first, or even creates the market. So people start using that vendor's products and then it becomes difficult for them to switch and learn something new. Many are satisfied if something just simply works, and they don't want change. In this SIP case, they probably purchase the hardware and software as a bundle.
Yup! I can see that. But why can't OSS businesses offer the same solution.
++This same thing can be said of peoples' reluctance to stop using Windows. Sure, some games don't run on Linux and there are some other drawbacks, but otherwise Linux can serve pretty well in the personal desktop area
You know what? That aspect hadn't even occurred to me, but I bet you just nailed it. It's GAMES. Now I, as a UNIX administrator, don't have many games available to me, but I do remember how an MS administrator told me when DOOM hit the gaming market that his network was suffering badly because everyone was playing DOOM network games. So a large portion of the user community *don't* want to switch because they'd lose access to their favorite games. Someone *PLEASE* tell me I'm being too cynical. On the other hand... I have a single XP system here - why? Well, my *official* excuse is that we need it for a couple of critical applications which will only run on Windows. Unofficially? Err.. There are a couple of games on that system and.. we really need the shit-hot graphics card and 2G RAM and biiiiig processor and monitor because... LOOK - I OWN THE COMPANY - OK? SHADDAP!!
++"Nobody ever got fired when the PIR yielded the reason for the outage was vendor error. The vendor takes the blame, the employee looks clean. Employee uses unsupported platform with nobody else to take the blame: employee has to suck it up and admit a bad decision."
Or, in the real world: "The reason for the outage was vendor error: employee has to suck it up and admit a bad decision."
++The bells and whistles are often critical functionality. I've seen one place pay up for bounty after bounty to get all their features in open source, however, so that can work.
Err, no. I defined "bells and whistles" as fringe functionality earlier in my post - no cigar. I also pointed out that if a FOSS package won't do what I want, I'll go the proprietary way, so I would certainly not throw money at a dysfunctional FOSS product.
++If my Linux box panics, I get to google for it and eventually give up. If my Sun/Solaris box panics, I get to send a core dump to Sun and annoy them until I get an answer and a patch.
Linux box panics? SOLARIS BOX PANICS? Very rare; however I get your point - I assume you mean O/S or application problems. A few points here: for Linux, I'd google for the correct *forum* and ask there. Funnily enough there are also Solaris forums which can be more helpful than Sun support. In my experience, if the problem proves intractable, an email to the original author of the Linux code usually provides a helpful response within 48 hours. For Solaris I send a full report from the Solaris diagnostic tool (can't remember what it's called, but much more than a core dump). I try not to *annoy* them, because that's usually counter productive.
++"4. It's free, so it can't be worth anything. Nah, this one has pretty much gone away. Used to be a HUGE amount of this mindset though."
I agree that this is much less of a problem than it was but it's still there more than I would like.
++"FOSS" is a really, really stupid sounding abbreviation. It sucks. Most people in the industry will respond to "Open Source", but they'll either glaze over or burst out laughing if someone says "FOSS". Besides, in a number of industries "OSS" is not free software.
I have to agree. FOSS is not a well recognised acronym and, as you say, it sucks. Also, the little internicene wars between the various licences, definitions (GNU, OSS, etc) are doing very little good for the "Open Source" cause. I can (sort of) understand the "religious" differences between open source licences and the GNU licences (no, they're *not* the same - look it up). I can't help feeling, however, that these disagreements within the community are slowing down the adoption of FOSS in the business community. My view is that as soon as FOSS/OS/Freeware/GNU/etc present a united front, business take-up will improve enormously. Remember that these entities are effectively run by bean-counters and lawyers (spit).
++Windows has no free software culture. It has even less than zero open source culture. A lot of Solaris shops run a lot of open source, but a lot of open source is... LINUX ONLY. Proprietary. Want X feature? Only compiles on Linux. User-mode app requires kernel headers. Only supports PC hardware. Requires glibc. Requires gcc. Won't compile on a 64bit platform.
You're partially (mostly?) correct. I run a couple of companies which both use Linux and Solaris. I don't really have a problem with using OSS with Solaris, but that's probably because I only want OSS software to supply services on Solaris which Sun *hardware* is especially good at providing. Maybe that's why I can usually find the right applications in the OSS field. I don't use Linux as the O/S on Sun platforms - I use Solaris. Likewise, I don't use KDE/Gnome or anything else on the Sun systems. From my point of view, Suns are for heavy throughput systems - not desktops. I use Linux for desktop systems. (No, I don't use Windows because almost all our system
Or, more puzzlingly (that a word?), how do some companies get away with competing against FOSS products with highly expensive proprietary offerings? I'm assuming that the proprietary solution has the same functionality as the other; maybe some bells and whistles on the fringes, but essentially the same.
They must make their money from licencing fees (and maintenance, but FOSS can do that, too). So why don't customers choose the cheaper option. Don't get me wrong; while I approve of FOSS and use it whenever I can, I won't hestitate to buy a proprietary product if it does what I need and there isn't a viable FOSS alternative.
I'm no expert in this - which is why I'm puzzled. Can anyone tell me (us) why? Is it any combination of the following?
1. "Noone was ever fired for buying IBM" (MS/Cisco/etc). 2. The bells and whistles are what the buyer craves. 3. Proprietary products have better support. 4. It's free, so it can't be worth anything. 5. What's FOSS? 6. We only run Windows (Solaris, whatever). 7. Proprietary products are better "rounded" or "easier to use".
I know that all these have flaws and, sometimes the reason is valid. But overall, I think my question still stands.
BTW. If anyone can think of anything to add to the list - I'd love to hear it.
"something thats so far beyond our current capabilities its like the Wright Brothers talking about building a 747"
Actually, the Wright brothers had a much more difficult challenge. They were doing something for the first time; something that was supposedly impossible. After they had cracked the problem a 747 was simply a matter of time.
If we had decided to simply achieve within "our current capabilities" we wouldn't be that far from banging rocks together. How do you think we push past "current capabilities"?
I've never really had a problem - but then, I'm *not* politically correct.
I run two companies; one a computer consultancy and the other is a Biotech research lab (don't ask). Both of them have a mix of male and female (ratio ~ 1:1 and 6:4). My recruitment policy is very simple - "Can you do the job?". I don't care what colour you are, what race you are, what your religion is, what your sexual orientation is, what sex you are, etc. I tell my staff that when they're on the job they're Engineers/Scientists/Whatever - NOT male/female or jewish/christian/muslim or any other thing. And I expect them to behave like professionals. This means I *do* care about friction in the workplace and I enforce the professional approach.
A trivial (but good) example was the issue of who made the coffee. In the early days of the computer consultancy, some males of a "certain ethnic background":) thought it was a "woman's job" to do menial tasks such as make the coffee for the rest of us, despite the fact that their job description was just the same as theirs. I pointed out that *I* made the coffee when it was my turn, I also treated the females exactly the same as I did the males (I *really* don't care), so what was their excuse? They decided to do their share of coffee duty. (OK, saying that the boss should make their coffee, but not the other way round is not a good career move).
Others have mentioned that going to the bar together is good for team spirit - and so it is! There isn't any problem when the boss says "Time to close up shop. Who wants to go to the pub? I'm buying the first round." And off we went. BTW; in the pub (bar) I'm not the boss anymore.
We have a good, friendly working environment. We don't need nosey, fanatical agencies with incomprehensible (and unworkable) agendas to help us work together. We do that ourselves quite nicely, thank you. All it takes is strong leadership (not just me) and peer pressure to say what is out of line. We're not perfect - the usual rivalries arise, but they don't destroy the team spirit.
It's not "sterile" either. A few years ago two of my employees got married and we were all invited to the wedding. I wasn't asked to be the Best Man though:-(
Not a great loss, I'm sorry to say.
on
The Future of NetBSD
·
· Score: 3, Informative
The great advantage that NetBSD had was its fast and secure network facilities. Unfortunately, for many potential users the problems far outweigh the benefits. And the situation has been getting worse for some time now.
DISCLAIMER: I am one of those bearded UNIX gurus (28 years experience), but I don't have a beard (Note to self - SHAVE) and I don't live on a mountain (I live in Italy in Porchiano del Monte... damn!)
The point about "rules of thumb" is that they are just that. They are *not* a replacement for a well engineered solution.
I've seen the role of swap space wax and wane over the years and, although its function has changed quite a bit, it still has its uses today.
In the following text I concentrate mostly on storage capacity, not timing (then it became *really* interesting). I have also been a bit free with the technical details over time. (More detail? Finance me to write a book)
In olden days (when a good sysadmin needed to be able to read core dumps in octal) The disparity between RAM (core) and disk space (Fastrand drum, anyone?) was *huge*; eg 2M disk, 16K RAM (about 120:1 - note this figure, it appears again, later). The other important variable was process size (text+data). In those days, programmers went to extreme lengths to reduce the load on RAM. Nonetheless, processes still managed to occupy all available RAM so mechanisms were necessary to manage the overflow. Allocating a portion of disk space was a logical solution when one could sacrifice a small percentage of disk for the ability to run larger programs. Originally these were called "overlays" which just dealt with the text part of the process (the actual program - not the data) but eventually evolved into primitive swapping which dealt with both text and data albeit on a "all or nothing" basis. The important bit is that processes regularly occupied more space than RAM could provide. In these ancient times the "swap" size was carefully calculated, often down to the last word.
Later, both RAM and disk technology started to improve rapidly but not in parallel. RAM was catching up and the new ratio of disk:RAM was more like 50:1 or less. At the same time process size increased, but not to the same extent because programmers were still very aware of memory constraints. Usually a few processes would fit into RAM quite easily. But now real "multitasking" was available and everyone wanted to use it. Multiple processes *wouldn't* fit into RAM but it was possible to cheat by using swap to create the illusion of more RAM than there actually was. This also drove the development of "all or nothing" swapping to something called "demand paging" (BSD something or other; look it up) which tightened the granularity of disk to memory transactions and blurred the distinction between permanent disk store, swap and RAM. So, swap survived because, although RAM capacity had expanded enormously, demands on it had expanded even more. This is the time when the first "rules of thumb" appeared; and they appeared because noone really knew how the process/memory model worked. A good rule of thumb was swap should be between two and three times RAM capacity - it was rough, but it generally worked. There *were* exceptions when the ROT obviously *didn't* work so it was down to trial and error to find out what did. This was always a result of programmers designing programs that weren't really designed for the technology of the day. Some databases and design tools spring to mind here.
Nowadays, we have an entirely new situation. Or do we?
From high-end desktop machines to small/medium sized servers we are looking at a disk:RAM ratio of 120G:1G or 500G:4G and so on. A ratio of about 120:1 - remember that? The big difference between now and the olden days is that the *role* of these computers can differ wildly.
One way of looking at this is that a desktop machine is simply a vastly overpowered server - the same machine could easily deal with hundreds, if not thousands, of server clients but it doesn't need to... just one user. In this case it's very unlikely that it will need to use swap because the number and scale of processes it's required to run are largely known and it's got more RAM than it will likely use. S
"What is generally enough to keep thieves out is a) basic human morality, and b) the law."
You forgot c) which is my preferred solution. Namely:
Three very large dogs who are fanatically devoted to us, well known in the neighbourhood and act appropriately to a threat. This usually involves a "WTF do you think your doing" glare and growl. It's worked with humans so far.*
* Sadly, not with another dog who didn't take the hint.
"If the only thing men wanted was to live their short lives peacefully and raise a family, then nobody would care if he was being watched."
Oh, yeah?
Scenario:
You are having a meal in a restaurant. You aren't doing anything wrong, you have no secrets to hide.
I am standing next to you - not doing anything except watching you. I don't move. I don't speak. I just watch... intently... quietly.
Guaranteed result after 5 mins:
You shout something like, "WTF IS YOUR PROBLEM??!"
But I don't reply. I just stand there. I don't move. I don't speak. I just watch... intently... quietly.
And you think this would be OK?
"Of course the only reason this world would appear unattractive to someone is if he had something to hide. Imagine no lying and no cheating. You do something and you stand by it. Proudly."...and looking nervously over your shoulder ALL the time. But you have nothing to hide... do you?
Really, you'd be amazed at how even the simplest graphical interpretation of complex data can really show up points of interest. And it's not difficult to see why: Humans' primary sense is visual and we have evolved some seriously complex neural algorithms to interpret visual data.
A simple graph is a case in point. Now take a large amount of complex data and apply just about any process you care to name to present a graphical representation and you can easily see the overall picture.
A very simple example which illustrates statistical clustering. Even with totally random numbers, you *will* find islands of apparently significant populations. This is a common counter-claim to action groups who claim, say, a correlation between mobile 'phone masts and incidents of child leukaemia*. Anyway:
Generate a stream of random numbers and assign a symbol for n = 0.5, display the symbols in a grid and, hey presto! Look at those clusters!
On a more positive note:
We often use graphical representation in our work. This ranges from CTK representations of molecules we're looking at (xlation - pretty pictures with balls and lines) to grid based colour indexed representation of multi-dimentional data sets. In each case the point is to present data in a way that we humans can quickly spot potential areas of interest and get a "feel" for the data we're looking at.
It's all good stuff. (Sometimes very pretty, too)
* Actually, this is a good example of why I'm always wary of purely statistical "proofs". In this case the *science* (ie. proposed mechanisms for this) don't hold up to current understanding.
It's interesting but not (yet) as significant as TFA makes out.
These are studies on transgenic mice, so it's more a proof-of-concept rather than clinical trials which will be some way off - mostly due to bureaucracy.
For those who want a quick *scientific-ish* summary:
It is widely believed that a protein called Beta-Amyloid is reponsible synaptic dysfunction in Altzheimer's disease. Another variant (Alpha-Amyloid) also does horrible things to the body such as renal failure and constrictive pericarditis. This often happens as a result of certain auto-immune diseases (which is my speciality).
These tests are based on the accidental discovery that a dysentry drug (PBT-1) has some effects on restoring some cognitive function in patients. The company pursuing this has created a drug which is more specifically targeted towards reducing levels of A-A. And... so far, so good. The mice show greatly reduced A-A levels and they perform better in mazes. I wish them all the best - Altzheimers is a horrible and frightening disease.
1. What is it with these long drawn-out reviews? 2. And why the huge graphics? 3. And why do they seem to assume that we're not interested in price?
OK, (1) is probably so they can put adverts on each page and get revenue. On the other hand they're not nearly as bad as a certain RP whose first page about a robotic snake contained no fewer than 17 ads. On the gripping hand, this article had about 4 ads per page which works out at 76 ads in total. Right! Definitely about showing ads, then. Good thing I have Adblock on my browser - eat my dirt RP, Mwahahahah!
As for (2), I still don't get it. The images are about 50k each which adds up to.. Ahhh, who cares? A lot more than it should be. Do you *really* want to be able to read the small print on the packaging? Thought not.
But my venom is reserved for point 3.
Now I, like most (all?) of you, buy a fair amount of kit. I *always* have the financial side in mind, even when I'm trying to build a super-duper computation server. There's *always* a trade off between price and performance. So why didn't the final grid show a cost/benefit breakdown? Actually this is a gripe I have with a lot of reviews in both printed and online reviews. For example, why not present the cost-per-gig as part of the conclusion in a disk review? Or cost-per-throughput MB, or... you get the idea.
BTW. If anyone knows of a publication/site who *do* present their finding like this - please let us know.
Coincidentally, I was having a conversation about the US approach to (or, rather, retreat from) issues of freedom with a group of Europeans last week, and this was just one of the issues that came up. It was unanimously agreed that the powers of the US secret (and not so secret) police were beginning to become alarming. Their possible future effects on the rest of the world are even more alarming.
Not only does the US have a big say in how the internet is run, they also produce or licence a significant proportion of computer kit today. OK, maybe the US supply to the world market will die the death and other countries will take up the slack; but that's not the issue, is it?
Other countries are trying to follow suit - look at the UK. They have a law called RIP (Regulation of Investigatory Powers) act. This is a misnomer because it is really the HOMUP (Hand Over Massive and Unrestricted Powers) act. Sounds very much like the US Patriot act (BTW, that was a clever name - "If you don't support this act, you can't be a patriot".)
There was a time, not that long ago, when the US prided itself as being the leaders of the free world. Perhaps they should hand the baton over to someone else before they drop it.
"think it would be significantly less effort for you to transition to a new email address. This time, don't give it out to anyone you don't trust. Get a throw away address for that and filter/forward it."
Tried that - doesn't work. Why?
Well, a lot of the people I email use Windows (I know, I know) and they are frequently compromised. The attacker gets their email list and Bingo!
Actually it does work, for a while. But then the rot starts and slowly, but surely the spam mounts up again.
"If you're running raid5 it's probably in an enterprise setup. If so, why aren't you running a dedicated controller?"
'Cos it's cheaper not to, that's why. I've seen plenty of installations that have large disk arrays running under Veritas Volume Manager and the performance is perfectly adequate for the client's needs. Disks are cheap, controllers are not. It's RAID - Redundant Array of INEXPENSIVE Disks - not RAIDVEC (RAID with Very Expensive Controller).
I think this just highlights what a mess the whole issue of net banking/payment is in.
I use several sites for buying goods, but EBay is a good example, so I'll use that:
As noted throughout this discussion, EBay only allows certain payment methods - foremost is Paypal. The others aren't anything like as useful because they are restricted to a geographic location (err.. why?) or just a way of processing credit card payments. Worse still, a lot of sellers seem to be locked into Paypal as well. So, OK, I use Paypal - I don't have any major reasons to worry about that except... Paypal won't allow me to use my (Italian) bank account for payments. Others, such as netteller will, but they in turn won't allow me to transfer funds to Paypal. In fact, even though I have several on-line account facilities, it is very nearly impossible to move funds from one to the other.
What gives? These days I can route data from anywhere to anywhere via the 'net using a plethora of methods; but not money. I wonder if this is deliberate.
I work at a small BioTech company. We use a lot of "obsolete" equipment because a) it's *much* cheaper than the new stuff and b) it does what we want anyway.
This is why we like projects like FreeDOS. For example; we use BioMek 1000's for liquid handling - they cost us about $1,000 each compared with $20,000+ for the newest kit. The problem is that the s/w runs on DOS... do we have DOS installation disks? Nope! Enter FreeDOS - it works fine *and* it runs happily on the old PC host.
Thank you FreeDOS developers - you just saved us a whole load of money which we can put into research instead.
"There even a (if somewhat shaky) DNA test to determine racial descent. I saw it on a TV show once, where they had some school kids find out they had DNA from basically another race. I.e. a black guy turned out to have some asian genes, a white girl with blonde hair turned out to have some black genes etc. Possibly a bullshit test, possibly not."
Determining one's racial ancestry via DNA tests sounds like a good idea, but doesn't work nearly as well as its more vocal proponents make out. Humans have very little variation genetically (possibly why inbreeding is a *very* bad idea for us) and we know very little about what characteristics a caused by what genes. This is mostly due to the fact that there isn't a single sequence that codes for, say, black skin. Rather it's a combination of lots of sequences, some in completely different parts of the genome that add up to a particular characteristic. And that combination, in turn, influences other combinations. Simple, it ain't!
Anyway, the possibility of someone like me (white, blue eyes) having having some black genes isn't all that amazing. In fact I *know* *I* have because my great, great...great grandmother was black (family history - not DNA testing). Maybe that's why I don't burn in the sun so easily:-)
How true. And I'm surprised that anyone is surprised.
It's like the car tax in the UK. It's *supposed* to be for renewing roads, putting up signposts and all sorts of other road-like things. Is it? OF COURSE NOT! I've heard that if the gov spent half the income from this tax on roads, we would have the most brilliant road infrastructure in the world.
Govs like to bring in taxes on any pretext - the stealthier the better. And don't expect anyone but the gov to benefit from it.
Passing the funds to someone else, indeed (mumble, grumble)
I've been an avid follower of space exploration for... well, all of my life. Hell, my father even woke me up at 3 AM so I could watch Armstrong do his stuff (I was eight) - he didn't dare *not* let me watch - the whinging would've been awful.
In those days, youngsters like me *knew* that we would have a base on the Moon in 10 years and another on Mars a few years after that. The excitement!
Oh, dear...
OK. I know now that it was all a "Get there before the Commies", but it *was* done. (BTW. To all you Yanks reading this - I think you guys made the greatest achievement of the human race, to date, happen. The reasons aren't important - you should be very proud).
Now look at it. It's starting again, but this time on many fronts - this isn't the only initiative. I'm eight years old again. The only difference is that I'm too old to play a part.
Are you trying to say Nietzsche was a normal human being? I think not:-)
Anyway, I wasn't using the word "normal" in a moral sense. I meant normal as in the vast majority who have the instincts to nurture and protect children.
To illustrate my point: Theoretically, someone with malformed instincts might be able to supress the actions or even thoughts that accompany this flaw through their morals. Just because they have a moral stance against what their diseased instincts are telling them, doesn't mean that they are *not* normal.
Or, to put it another way: A NORMAL adolescent male (I use male 'cos it's much more pronounced in males) wants sex as often as possible - he's not interested in children because they don't trigger his instincts. He doesn't try to rape every female who comes his way because he has NORMAL morals (and/or a normal understanding of what will happen to him if he's caught). In this case the instinct is normal, but it doesn't *necessarily* result in moral behaviour.
Slashdot Mirror
++One more reason I have observed is that people get used to a particular platform. More often than not, a commercial vendor enters a market first, or even creates the market. So people start using that vendor's products and then it becomes difficult for them to switch and learn something new. Many are satisfied if something just simply works, and they don't want change. In this SIP case, they probably purchase the hardware and software as a bundle.
Yup! I can see that. But why can't OSS businesses offer the same solution.
++This same thing can be said of peoples' reluctance to stop using Windows. Sure, some games don't run on Linux and there are some other drawbacks, but otherwise Linux can serve pretty well in the personal desktop area
You know what? That aspect hadn't even occurred to me, but I bet you just nailed it. It's GAMES. Now I, as a UNIX administrator, don't have many games available to me, but I do remember how an MS administrator told me when DOOM hit the gaming market that his network was suffering badly because everyone was playing DOOM network games. So a large portion of the user community *don't* want to switch because they'd lose access to their favorite games. Someone *PLEASE* tell me I'm being too cynical. On the other hand... I have a single XP system here - why? Well, my *official* excuse is that we need it for a couple of critical applications which will only run on Windows. Unofficially? Err.. There are a couple of games on that system and.. we really need the shit-hot graphics card and 2G RAM and biiiiig processor and monitor because... LOOK - I OWN THE COMPANY - OK? SHADDAP!!
Well, you have some good points, but...
... LINUX ONLY. Proprietary. Want X feature? Only compiles on Linux. User-mode app requires kernel headers. Only supports PC hardware. Requires glibc. Requires gcc. Won't compile on a 64bit platform.
++"Nobody ever got fired when the PIR yielded the reason for the outage was vendor error. The vendor takes the blame, the employee looks clean.
Employee uses unsupported platform with nobody else to take the blame: employee has to suck it up and admit a bad decision."
Or, in the real world: "The reason for the outage was vendor error: employee has to suck it up and admit a bad decision."
++The bells and whistles are often critical functionality. I've seen one place pay up for bounty after bounty to get all their features in open source, however, so that can work.
Err, no. I defined "bells and whistles" as fringe functionality earlier in my post - no cigar. I also pointed out that if a FOSS package won't do what I want, I'll go the proprietary way, so I would certainly not throw money at a dysfunctional FOSS product.
++If my Linux box panics, I get to google for it and eventually give up. If my Sun/Solaris box panics, I get to send a core dump to Sun and annoy them until I get an answer and a patch.
Linux box panics? SOLARIS BOX PANICS? Very rare; however I get your point - I assume you mean O/S or application problems. A few points here: for Linux, I'd google for the correct *forum* and ask there. Funnily enough there are also Solaris forums which can be more helpful than Sun support. In my experience, if the problem proves intractable, an email to the original author of the Linux code usually provides a helpful response within 48 hours. For Solaris I send a full report from the Solaris diagnostic tool (can't remember what it's called, but much more than a core dump). I try not to *annoy* them, because that's usually counter productive.
++"4. It's free, so it can't be worth anything.
Nah, this one has pretty much gone away. Used to be a HUGE amount of this mindset though."
I agree that this is much less of a problem than it was but it's still there more than I would like.
++"FOSS" is a really, really stupid sounding abbreviation. It sucks. Most people in the industry will respond to "Open Source", but they'll either glaze over or burst out laughing if someone says "FOSS". Besides, in a number of industries "OSS" is not free software.
I have to agree. FOSS is not a well recognised acronym and, as you say, it sucks. Also, the little internicene wars between the various licences, definitions (GNU, OSS, etc) are doing very little good for the "Open Source" cause. I can (sort of) understand the "religious" differences between open source licences and the GNU licences (no, they're *not* the same - look it up). I can't help feeling, however, that these disagreements within the community are slowing down the adoption of FOSS in the business community. My view is that as soon as FOSS/OS/Freeware/GNU/etc present a united front, business take-up will improve enormously. Remember that these entities are effectively run by bean-counters and lawyers (spit).
++Windows has no free software culture. It has even less than zero open source culture. A lot of Solaris shops run a lot of open source, but a lot of open source is
You're partially (mostly?) correct. I run a couple of companies which both use Linux and Solaris. I don't really have a problem with using OSS with Solaris, but that's probably because I only want OSS software to supply services on Solaris which Sun *hardware* is especially good at providing. Maybe that's why I can usually find the right applications in the OSS field. I don't use Linux as the O/S on Sun platforms - I use Solaris. Likewise, I don't use KDE/Gnome or anything else on the Sun systems. From my point of view, Suns are for heavy throughput systems - not desktops. I use Linux for desktop systems. (No, I don't use Windows because almost all our system
Or, more puzzlingly (that a word?), how do some companies get away with competing against FOSS products with highly expensive proprietary offerings? I'm assuming that the proprietary solution has the same functionality as the other; maybe some bells and whistles on the fringes, but essentially the same.
They must make their money from licencing fees (and maintenance, but FOSS can do that, too). So why don't customers choose the cheaper option. Don't get me wrong; while I approve of FOSS and use it whenever I can, I won't hestitate to buy a proprietary product if it does what I need and there isn't a viable FOSS alternative.
I'm no expert in this - which is why I'm puzzled. Can anyone tell me (us) why? Is it any combination of the following?
1. "Noone was ever fired for buying IBM" (MS/Cisco/etc).
2. The bells and whistles are what the buyer craves.
3. Proprietary products have better support.
4. It's free, so it can't be worth anything.
5. What's FOSS?
6. We only run Windows (Solaris, whatever).
7. Proprietary products are better "rounded" or "easier to use".
I know that all these have flaws and, sometimes the reason is valid. But overall, I think my question still stands.
BTW. If anyone can think of anything to add to the list - I'd love to hear it.
"something thats so far beyond our current capabilities its like the Wright Brothers talking about building a 747"
Actually, the Wright brothers had a much more difficult challenge. They were doing something for the first time; something that was supposedly impossible. After they had cracked the problem a 747 was simply a matter of time.
If we had decided to simply achieve within "our current capabilities" we wouldn't be that far from banging rocks together. How do you think we push past "current capabilities"?
Sheesh!
I also like Attenborough's presentations - and I liked Steve's. They're like chalk and cheese and you can't *really* compare the two.
Many was the time my wife and I called to the other "Hey, it's that looney Aussie on the telly!" Good fun and educational, too.
Of course his death is not that significant in the greater scheme of things, but then, neither was my Grandmother's. I still miss her.
In a not totally insignificant way, I'll miss him, too.
I've never really had a problem - but then, I'm *not* politically correct.
:) thought it was a "woman's job" to do menial tasks such as make the coffee for the rest of us, despite the fact that their job description was just the same as theirs. I pointed out that *I* made the coffee when it was my turn, I also treated the females exactly the same as I did the males (I *really* don't care), so what was their excuse? They decided to do their share of coffee duty. (OK, saying that the boss should make their coffee, but not the other way round is not a good career move).
:-(
I run two companies; one a computer consultancy and the other is a Biotech research lab (don't ask). Both of them have a mix of male and female (ratio ~ 1:1 and 6:4). My recruitment policy is very simple - "Can you do the job?". I don't care what colour you are, what race you are, what your religion is, what your sexual orientation is, what sex you are, etc. I tell my staff that when they're on the job they're Engineers/Scientists/Whatever - NOT male/female or jewish/christian/muslim or any other thing. And I expect them to behave like professionals. This means I *do* care about friction in the workplace and I enforce the professional approach.
A trivial (but good) example was the issue of who made the coffee. In the early days of the computer consultancy, some males of a "certain ethnic background"
Others have mentioned that going to the bar together is good for team spirit - and so it is! There isn't any problem when the boss says "Time to close up shop. Who wants to go to the pub? I'm buying the first round." And off we went. BTW; in the pub (bar) I'm not the boss anymore.
We have a good, friendly working environment. We don't need nosey, fanatical agencies with incomprehensible (and unworkable) agendas to help us work together. We do that ourselves quite nicely, thank you. All it takes is strong leadership (not just me) and peer pressure to say what is out of line. We're not perfect - the usual rivalries arise, but they don't destroy the team spirit.
It's not "sterile" either. A few years ago two of my employees got married and we were all invited to the wedding. I wasn't asked to be the Best Man though
The great advantage that NetBSD had was its fast and secure network facilities. Unfortunately, for many potential users the problems far outweigh the benefits. And the situation has been getting worse for some time now.
Bye-bye NetBSD, it was good while it lasted.
DISCLAIMER: I am one of those bearded UNIX gurus (28 years experience), but I don't have a beard (Note to self - SHAVE) and I don't live on a mountain (I live in Italy in Porchiano del Monte... damn!)
The point about "rules of thumb" is that they are just that. They are *not* a replacement for a well engineered solution.
I've seen the role of swap space wax and wane over the years and, although its function has changed quite a bit, it still has its uses today.
In the following text I concentrate mostly on storage capacity, not timing (then it became *really* interesting). I have also been a bit free with the technical details over time. (More detail? Finance me to write a book)
In olden days (when a good sysadmin needed to be able to read core dumps in octal) The disparity between RAM (core) and disk space (Fastrand drum, anyone?) was *huge*; eg 2M disk, 16K RAM (about 120:1 - note this figure, it appears again, later). The other important variable was process size (text+data). In those days, programmers went to extreme lengths to reduce the load on RAM. Nonetheless, processes still managed to occupy all available RAM so mechanisms were necessary to manage the overflow. Allocating a portion of disk space was a logical solution when one could sacrifice a small percentage of disk for the ability to run larger programs. Originally these were called "overlays" which just dealt with the text part of the process (the actual program - not the data) but eventually evolved into primitive swapping which dealt with both text and data albeit on a "all or nothing" basis. The important bit is that processes regularly occupied more space than RAM could provide. In these ancient times the "swap" size was carefully calculated, often down to the last word.
Later, both RAM and disk technology started to improve rapidly but not in parallel. RAM was catching up and the new ratio of disk:RAM was more like 50:1 or less. At the same time process size increased, but not to the same extent because programmers were still very aware of memory constraints. Usually a few processes would fit into RAM quite easily. But now real "multitasking" was available and everyone wanted to use it. Multiple processes *wouldn't* fit into RAM but it was possible to cheat by using swap to create the illusion of more RAM than there actually was. This also drove the development of "all or nothing" swapping to something called "demand paging" (BSD something or other; look it up) which tightened the granularity of disk to memory transactions and blurred the distinction between permanent disk store, swap and RAM. So, swap survived because, although RAM capacity had expanded enormously, demands on it had expanded even more. This is the time when the first "rules of thumb" appeared; and they appeared because noone really knew how the process/memory model worked. A good rule of thumb was swap should be between two and three times RAM capacity - it was rough, but it generally worked. There *were* exceptions when the ROT obviously *didn't* work so it was down to trial and error to find out what did. This was always a result of programmers designing programs that weren't really designed for the technology of the day. Some databases and design tools spring to mind here.
Nowadays, we have an entirely new situation. Or do we?
From high-end desktop machines to small/medium sized servers we are looking at a disk:RAM ratio of 120G:1G or 500G:4G and so on. A ratio of about 120:1 - remember that? The big difference between now and the olden days is that the *role* of these computers can differ wildly.
One way of looking at this is that a desktop machine is simply a vastly overpowered server - the same machine could easily deal with hundreds, if not thousands, of server clients but it doesn't need to... just one user. In this case it's very unlikely that it will need to use swap because the number and scale of processes it's required to run are largely known and it's got more RAM than it will likely use. S
"What is generally enough to keep thieves out is a) basic human morality, and b) the law."
You forgot c) which is my preferred solution. Namely:
Three very large dogs who are fanatically devoted to us, well known in the neighbourhood and act appropriately to a threat. This usually involves a "WTF do you think your doing" glare and growl. It's worked with humans so far.*
* Sadly, not with another dog who didn't take the hint.
"If the only thing men wanted was to live their short lives peacefully and raise a family, then nobody would care if he was being watched."
...and looking nervously over your shoulder ALL the time. But you have nothing to hide... do you?
Oh, yeah?
Scenario:
You are having a meal in a restaurant. You aren't doing anything wrong, you have no secrets to hide.
I am standing next to you - not doing anything except watching you. I don't move. I don't speak. I just watch... intently... quietly.
Guaranteed result after 5 mins:
You shout something like, "WTF IS YOUR PROBLEM??!"
But I don't reply. I just stand there. I don't move. I don't speak. I just watch... intently... quietly.
And you think this would be OK?
"Of course the only reason this world would appear unattractive to someone is if he had something to hide. Imagine no lying and no cheating. You do something and you stand by it. Proudly."
...than a thousand words.
Really, you'd be amazed at how even the simplest graphical interpretation of complex data can really show up points of interest. And it's not difficult to see why: Humans' primary sense is visual and we have evolved some seriously complex neural algorithms to interpret visual data.
A simple graph is a case in point. Now take a large amount of complex data and apply just about any process you care to name to present a graphical representation and you can easily see the overall picture.
A very simple example which illustrates statistical clustering. Even with totally random numbers, you *will* find islands of apparently significant populations. This is a common counter-claim to action groups who claim, say, a correlation between mobile 'phone masts and incidents of child leukaemia*. Anyway:
Generate a stream of random numbers and assign a symbol for n = 0.5, display the symbols in a grid and, hey presto! Look at those clusters!
On a more positive note:
We often use graphical representation in our work. This ranges from CTK representations of molecules we're looking at (xlation - pretty pictures with balls and lines) to grid based colour indexed representation of multi-dimentional data sets. In each case the point is to present data in a way that we humans can quickly spot potential areas of interest and get a "feel" for the data we're looking at.
It's all good stuff. (Sometimes very pretty, too)
* Actually, this is a good example of why I'm always wary of purely statistical "proofs". In this case the *science* (ie. proposed mechanisms for this) don't hold up to current understanding.
It's interesting but not (yet) as significant as TFA makes out.
e wsid=47696
These are studies on transgenic mice, so it's more a proof-of-concept rather than clinical trials which will be some way off - mostly due to bureaucracy.
For those who want a quick *scientific-ish* summary:
It is widely believed that a protein called Beta-Amyloid is reponsible synaptic dysfunction in Altzheimer's disease. Another variant (Alpha-Amyloid) also does horrible things to the body such as renal failure and constrictive pericarditis. This often happens as a result of certain auto-immune diseases (which is my speciality).
These tests are based on the accidental discovery that a dysentry drug (PBT-1) has some effects on restoring some cognitive function in patients. The company pursuing this has created a drug which is more specifically targeted towards reducing levels of A-A. And... so far, so good. The mice show greatly reduced A-A levels and they perform better in mazes. I wish them all the best - Altzheimers is a horrible and frightening disease.
For those who would like a fuller summary in non newspaper-speak, try http://www.medicalnewstoday.com/medicalnews.php?n
Yeah! I totally agree!
A few points:
1. What is it with these long drawn-out reviews?
2. And why the huge graphics?
3. And why do they seem to assume that we're not interested in price?
OK, (1) is probably so they can put adverts on each page and get revenue. On the other hand they're not nearly as bad as a certain RP whose first page about a robotic snake contained no fewer than 17 ads. On the gripping hand, this article had about 4 ads per page which works out at 76 ads in total. Right! Definitely about showing ads, then. Good thing I have Adblock on my browser - eat my dirt RP, Mwahahahah!
As for (2), I still don't get it. The images are about 50k each which adds up to.. Ahhh, who cares? A lot more than it should be. Do you *really* want to be able to read the small print on the packaging? Thought not.
But my venom is reserved for point 3.
Now I, like most (all?) of you, buy a fair amount of kit. I *always* have the financial side in mind, even when I'm trying to build a super-duper computation server. There's *always* a trade off between price and performance. So why didn't the final grid show a cost/benefit breakdown? Actually this is a gripe I have with a lot of reviews in both printed and online reviews. For example, why not present the cost-per-gig as part of the conclusion in a disk review? Or cost-per-throughput MB, or... you get the idea.
BTW. If anyone knows of a publication/site who *do* present their finding like this - please let us know.
Ahh. Finished my daily rant - I feel better now.
I had an opportunity to bid for a Viking Lander on eBay. You mean it's bogus?
Hey, are you a prophet? I got my first -1 Troll on that comment.
How true.
Coincidentally, I was having a conversation about the US approach to (or, rather, retreat from) issues of freedom with a group of Europeans last week, and this was just one of the issues that came up. It was unanimously agreed that the powers of the US secret (and not so secret) police were beginning to become alarming. Their possible future effects on the rest of the world are even more alarming.
Not only does the US have a big say in how the internet is run, they also produce or licence a significant proportion of computer kit today. OK, maybe the US supply to the world market will die the death and other countries will take up the slack; but that's not the issue, is it?
Other countries are trying to follow suit - look at the UK. They have a law called RIP (Regulation of Investigatory Powers) act. This is a misnomer because it is really the HOMUP (Hand Over Massive and Unrestricted Powers) act. Sounds very much like the US Patriot act (BTW, that was a clever name - "If you don't support this act, you can't be a patriot".)
There was a time, not that long ago, when the US prided itself as being the leaders of the free world. Perhaps they should hand the baton over to someone else before they drop it.
"think it would be significantly less effort for you to transition to a new email address. This time, don't give it out to anyone you don't trust. Get a throw away address for that and filter/forward it."
Tried that - doesn't work. Why?
Well, a lot of the people I email use Windows (I know, I know) and they are frequently compromised. The attacker gets their email list and Bingo!
Actually it does work, for a while. But then the rot starts and slowly, but surely the spam mounts up again.
And this is appropriate material for a /. story... How exactly?
"If you're running raid5 it's probably in an enterprise setup. If so, why aren't you running a dedicated controller?"
'Cos it's cheaper not to, that's why. I've seen plenty of installations that have large disk arrays running under Veritas Volume Manager and the performance is perfectly adequate for the client's needs. Disks are cheap, controllers are not. It's RAID - Redundant Array of INEXPENSIVE Disks - not RAIDVEC (RAID with Very Expensive Controller).
I think this just highlights what a mess the whole issue of net banking/payment is in.
I use several sites for buying goods, but EBay is a good example, so I'll use that:
As noted throughout this discussion, EBay only allows certain payment methods - foremost is Paypal. The others aren't anything like as useful because they are restricted to a geographic location (err.. why?) or just a way of processing credit card payments. Worse still, a lot of sellers seem to be locked into Paypal as well. So, OK, I use Paypal - I don't have any major reasons to worry about that except... Paypal won't allow me to use my (Italian) bank account for payments. Others, such as netteller will, but they in turn won't allow me to transfer funds to Paypal. In fact, even though I have several on-line account facilities, it is very nearly impossible to move funds from one to the other.
What gives? These days I can route data from anywhere to anywhere via the 'net using a plethora of methods; but not money. I wonder if this is deliberate.
IT'S A MESS!
I work at a small BioTech company. We use a lot of "obsolete" equipment because a) it's *much* cheaper than the new stuff and b) it does what we want anyway.
This is why we like projects like FreeDOS. For example; we use BioMek 1000's for liquid handling - they cost us about $1,000 each compared with $20,000+ for the newest kit. The problem is that the s/w runs on DOS... do we have DOS installation disks? Nope! Enter FreeDOS - it works fine *and* it runs happily on the old PC host.
Thank you FreeDOS developers - you just saved us a whole load of money which we can put into research instead.
"There even a (if somewhat shaky) DNA test to determine racial descent. I saw it on a TV show once, where they had some school kids find out they had DNA from basically another race. I.e. a black guy turned out to have some asian genes, a white girl with blonde hair turned out to have some black genes etc. Possibly a bullshit test, possibly not."
:-)
Determining one's racial ancestry via DNA tests sounds like a good idea, but doesn't work nearly as well as its more vocal proponents make out. Humans have very little variation genetically (possibly why inbreeding is a *very* bad idea for us) and we know very little about what characteristics a caused by what genes. This is mostly due to the fact that there isn't a single sequence that codes for, say, black skin. Rather it's a combination of lots of sequences, some in completely different parts of the genome that add up to a particular characteristic. And that combination, in turn, influences other combinations. Simple, it ain't!
Anyway, the possibility of someone like me (white, blue eyes) having having some black genes isn't all that amazing. In fact I *know* *I* have because my great, great...great grandmother was black (family history - not DNA testing). Maybe that's why I don't burn in the sun so easily
How true. And I'm surprised that anyone is surprised.
It's like the car tax in the UK. It's *supposed* to be for renewing roads, putting up signposts and all sorts of other road-like things. Is it? OF COURSE NOT! I've heard that if the gov spent half the income from this tax on roads, we would have the most brilliant road infrastructure in the world.
Govs like to bring in taxes on any pretext - the stealthier the better. And don't expect anyone but the gov to benefit from it.
Passing the funds to someone else, indeed (mumble, grumble)
I've been an avid follower of space exploration for... well, all of my life. Hell, my father even woke me up at 3 AM so I could watch Armstrong do his stuff (I was eight) - he didn't dare *not* let me watch - the whinging would've been awful.
In those days, youngsters like me *knew* that we would have a base on the Moon in 10 years and another on Mars a few years after that. The excitement!
Oh, dear...
OK. I know now that it was all a "Get there before the Commies", but it *was* done. (BTW. To all you Yanks reading this - I think you guys made the greatest achievement of the human race, to date, happen. The reasons aren't important - you should be very proud).
Now look at it. It's starting again, but this time on many fronts - this isn't the only initiative. I'm eight years old again. The only difference is that I'm too old to play a part.
Are you trying to say Nietzsche was a normal human being? I think not :-)
Anyway, I wasn't using the word "normal" in a moral sense. I meant normal as in the vast majority who have the instincts to nurture and protect children.
To illustrate my point: Theoretically, someone with malformed instincts might be able to supress the actions or even thoughts that accompany this flaw through their morals. Just because they have a moral stance against what their diseased instincts are telling them, doesn't mean that they are *not* normal.
Or, to put it another way: A NORMAL adolescent male (I use male 'cos it's much more pronounced in males) wants sex as often as possible - he's not interested in children because they don't trigger his instincts. He doesn't try to rape every female who comes his way because he has NORMAL morals (and/or a normal understanding of what will happen to him if he's caught). In this case the instinct is normal, but it doesn't *necessarily* result in moral behaviour.
Err... does that make sense?