... I would think that an ATM would be a more complex device than a voting machine.
In a ATM, the crucial work is being done back in a data centre, probably on a mainframe using software signed off by an archangel or higher authority. Much less scope for fuck-ups by the ATM maker. Further, bank presumably own the source to the back-end software, so can inspect it; details of transactions can be verified. Further still, the testing process for the whole system has to involve the bank (or their contractors other than the ATM vendor). It's not just a black box
"Why is it the responsibility of Linux distribution maintainers to provide a means for commericial vendors to package their product? Vendors had to spend money to get certified for other operating systems. How about putting a little work into understanding and using a Linux distribution.
The problem is the multiplicity of packaging systems. If there was one that was always present and always worked, then sure, it would be the vendor's responsibility to use it well.
You might think so, but silicon seems to deal with it OK. When I worked in astronomical instrumentation, we built LN2-cooled CCD cameras with the chips cooled to about -150 deg C (they had heaters in the cryostats to hold this temperature; they went down to about -200 if the heaters were turned off). These things were thermal-cycled many times a year and we rarely lost a chip. Astro CCDs are big chips, albeit very simple compared to CPUs.
Conversely, detectors cooled to liquid helium temperatures are likely to die if cycled up to room temperature a few times.
With the CCD cryostats, the chips were in a vacuum vessel, so condensation wasn't normally a problem; all the water had been sucked out before cooling. If moisture did get in, then they had to be warmed and re-evacuated.
"... Peer review has limitations of course, to thoroughly review an article, one would have to repeat the experiment, which most reviewers (for good reasons) do not do..."
Therefore the repeats of the experiments have to come in work done after the original publication. When some results fail to be confirmed, that's just scientific method doing its thing; it's necessary. If no published results were later found false, something would be wrong.
"But cloud computing is a buzzword for a marketing campaign. It's the newest renaming of renting software as services."
Except when it's not. My project is using rented VMs in a cloud where we control the OS (open) and the infrastructure (Java, open) and the apps (ours, open). We could do the same work on machines we own, if we wanted to buy and manage enough hardware; no lock in.
What actually happens for domestic users when the addresses run out? I get my one, dynamic address at home from the ISP and I guess tomorrow they give that to some other subscriber (DHCP lease seems to be 24 hrs). If there are too few addresses, then what? No more new subscribers; or do they, the ISPs, allow over-subscription and not all customers can get an IP address every day?
They were called, in the UK, trolley buses. Quite a few towns and cities had them, especially those that hadn't bothered with tram networks: conventional buses on road suspension, but with electric motors fed from overhead wires. I think some European cities still use them.
They died out in the UK because of the cost of maintaining the overhead line and, IIRC, the capital cost of the vehicles (few systems => few vehicles ordered => high cost per unit). Therefore, sadly, not every kind of electric bus meets the cheap-infrastructure criterion.
Re:research to application life cycle
on
LHC Success!
·
· Score: 1
50 and 25 years sound suspiciously like two and one human generations respectively. And 50 years is roughly an academic career from start of university education through to retirement. Maybe we only get the new stuff as the old guard are replaced?
There is an argument that gross computing power lessens the near for, and appeal of, "cutting edge mathematics". Researchers are less likely to work for a tricky, analytical solution if they can brute-force a numerical one for less effort.
... Companies need to hold themselves to the contract too...
Yes, I agree. But in the domestic-broadband contracts (in the UK) I have seen, the over-subscription and contention are written in. We need some ISPs to offer better contracts. If you have such a contract I'm very happy for you and please tell us all who, where and how much.
Suppose you write in Perl or Python and you use a lot of 3rd-party libraries. My personal experience is that the modules available for Perl are at a lower, more geekful level than those for Python. In Perl one gets four or five modules off CPAN and spends a painful few days bashing them together into a solution. In Python, one more often gets a canned solution that works immediately.
If that carries over into the corperate domain - and it's a big if - then it makes Perl look expensive. And poorly-trained developers will do better with the canned solutions than with the low-level stuff.
Disclaimer: my experience is coding for scientists with FOSS and may not translate to the corporate domain. And I haven't written very much Python.:-/
Hey, just out of curiosity, what tools do you as a programmer prefer to pay for? The only thing I could imagine paying for personally would be Qt4. I've tried Visual Studio 2008 Professional and it's still humiliated by Eclipse, so I don't think you're talking about that.
The ones I've paid for in the last few years are oXygen, an XML editor; Lazgo Keystore Explorer a GUI programme for managing keystores; and Smartdraw, vector-graphics software for illustrating design documents. I also had access to a project license for a UML system (Together Control Centre, sold by Borland last time I checked).
oXygen I consider to be good value. Keystore Explorer could possibly be replaced by a decent FOSS tool but it's very cheap and handy so I haven't looked. Smartdraw is only for Windows and not useful to me now, but it paid its way in the past. Nowadays, I'd probably manage with the drawing tool in Open Office; back in the day this wasn't practical. Together CC, the only expensive one, is utterly ghastly to use and really deserves to be driven out by FOSS. But last time I looked, maybe 2 years ago now, the FOSS alternatives still weren't as good for what I needed to do.
And what with bad paid-for tools? If your management bought tools which are still usnusable, but paid for them, you would have no problem?
*Grits teeth* No, I don't want commercial tools regardless of quality. I want each tool to be evaluated on its merits, and I want alternatives in each category. Sometimes the commercial tools are good alternatives, sometimes not. If the commercial tools go off the market, then I lose choice. If my organization won't pay for tools, then I lose choice. Both are bad.
I am not in opposition to FOSS. I make my living writing exclusively FOSS; I am privileged in this respect.
The problem here is with management, not free software.
Yes! I agree! See informed, rational decisions, lack of, in other posts of this thread.
So next time my management refuse to buy a $200 tool and I lose a week of working time with an inferior FOSS equivalent that's me saved is it? Even if I have to make up the lost week in unpaid overtime? Good for my soul, maybe.
So, if there were no free tools, and your management had to close up shop and go get jobs working for someone else because the cumulative cost of all these tools was too much for their enterprise to bear, would that make you happier?
No of course not. Listen, I don't want the free software to go away. I want it to fulfil its hype and be as good as the paid-for stuff. In cases where that hasn't happened yet, I want the commercial stuff still to be available.
Why would anyone assume that a commercial tool, or any commercial software would be better than a free one? The reasoning behind the GPL license for example, is that many can add to the content of the work, as long as all can use it for free.
They shouldn't, and they shouldn't make the opposite assumption either. All tools should be evaluated on their merits. My point was that if a commercial tool is sufficiently better, in context, than its FOSS competition, then it should be bought and used.
From a purely rational perspective, I can not see how a commercial business model could produce a better product.
Really? Or is that just rhetoric? Maybe a commercial product could come out better because: it gets more hours of input from an appropriate number of good developers; it gets better QA; its documentation is written by skilled technical authors rather than unskilled coders. Not in all cases, but it's possible.
Just because the GPL allows many top developers to work on a FOSS project it doesn't follow that they will.
Someone making a profit from the work of their programmers can not improve the quality of their work.
I think we're both generalizing and 'if'-fing a little too much. Every case should be examined separately. We can safely assume Qcad is not a real replacement for AutoCAD, whereas OOo will be more than enough for the majority of MSOffice users.
I concur. The problem is that, where I've worked, cases aren't examined on their merits. There is an assumption that a FOSS alternative is always equivalent to the paid-for thing. Got GIMP, not gonna pay for Photoshop, etc.
My experience is that buyers at all levels won't do that when there's a cost-free alternative.
If that were true, most places where employees only use email, web browsing and office software would be installing Linux instead of the almost ubiquitous Windows.
That may be a bad example. For corporate buyers, TCO matters, and many perceive (possibly wrongly) that it's lower for Windows if they already have some Windows infrastructure. For consumers, Windows looks cost-free as the cost is hidden in the price of the machine and most don't know how to buy OS-free hardware.
So next time my management refuse to buy a $200 tool and I lose a week of working time with an inferior FOSS equivalent that's me saved is it? Even if I have to make up the lost week in unpaid overtime? Good for my soul, maybe.
Don't get me wrong, I like FOSS software, but I do need it to work. It's for using, not for looking at. If I need a tool and the FOSS versions are inadequate, then I need the commercial version, at least until the FOSS world catches up.
Bad free tools don't increase my productivity compared to good, paid-for tools. They might increase a society's productivity, and I think that's what your rant was really about. But that doesn't help me as an individual. I'm happy that people who can't pay for tools get cost-free ones, but that shouldn't stop my organization buying better tools when appropriate.
PS: informed, rational decisions are an assumption in free-market economics. The fact that you don't like capitalism doesn't make this untrue, as you seem to imply.
Pure free-market economics assume that the players are making rational informed decisions. In software acquisition, that assumption fails often.
If the more-expensive tool saves time worth more than its cost, then the appropriate free-market choice is to invest. My experience is that buyers at all levels won't do that when there's a cost-free alternative. They'd rather waste time (=money) or lose quality (=money due to cost of fixing later) than spend capital.
Shorter fibres: isn't that what happens to paper after multiple recyclings? I wonder if this is what to do with the recycled conventional paper that's no longer good enough evn for newsprint.
The robust-scanner one, almost certainly. This is likely an easier job than hardening an interactive web-browser. Their robot has no need to execute anything it comes across, so downloaded script needn't be allowed to execute anything, ever. It has no need to render any of the media, so none of the image-library attacks can work. They don't have to keep anything that they scan, so no save-to-disc code. In short, they can maintain exceptionally strong separation between their scanner and its host.
If they were paranoid enough, they could run the robot in a virtual machine and reinstall that after each scanning run. I have no idea if they consider that worthwhile.
Egypt does security theatre routinely. The antiquities sites all have metal detectors on the entrances. They are all turned on and the guards make visitors walk thorugh. The alarms go off for maybe a third of the visitors - coins, belt buckles etc., nothing bad. The attendants never check visitors who set of the alarms. They treat the detectors like some magic amulet that wards off evil; no human intervention needed.
Yes there are. The carpet-weaving industry in the UK still uses card-programmed looms (I have a friend who is employed to load card decks into the machines).
Slashdot Mirror
... I would think that an ATM would be a more complex device than a voting machine.
In a ATM, the crucial work is being done back in a data centre, probably on a mainframe using software signed off by an archangel or higher authority. Much less scope for fuck-ups by the ATM maker. Further, bank presumably own the source to the back-end software, so can inspect it; details of transactions can be verified. Further still, the testing process for the whole system has to involve the bank (or their contractors other than the ATM vendor). It's not just a black box
Did we? I'd always thought that public, lending libraries were a 19th-century sort of thing. Wikipedia reckons that copyright was introduced in the 18th century.
It wasn't Straw who was conned. It was one of his constituency workers (according to the report on C4 news).
The problem is the multiplicity of packaging systems. If there was one that was always present and always worked, then sure, it would be the vendor's responsibility to use it well.
You might think so, but silicon seems to deal with it OK. When I worked in astronomical instrumentation, we built LN2-cooled CCD cameras with the chips cooled to about -150 deg C (they had heaters in the cryostats to hold this temperature; they went down to about -200 if the heaters were turned off). These things were thermal-cycled many times a year and we rarely lost a chip. Astro CCDs are big chips, albeit very simple compared to CPUs.
Conversely, detectors cooled to liquid helium temperatures are likely to die if cycled up to room temperature a few times.
With the CCD cryostats, the chips were in a vacuum vessel, so condensation wasn't normally a problem; all the water had been sucked out before cooling. If moisture did get in, then they had to be warmed and re-evacuated.
Except when it's not. My project is using rented VMs in a cloud where we control the OS (open) and the infrastructure (Java, open) and the apps (ours, open). We could do the same work on machines we own, if we wanted to buy and manage enough hardware; no lock in.
Not dead. Undead.
What actually happens for domestic users when the addresses run out? I get my one, dynamic address at home from the ISP and I guess tomorrow they give that to some other subscriber (DHCP lease seems to be 24 hrs). If there are too few addresses, then what? No more new subscribers; or do they, the ISPs, allow over-subscription and not all customers can get an IP address every day?
They were called, in the UK, trolley buses. Quite a few towns and cities had them, especially those that hadn't bothered with tram networks: conventional buses on road suspension, but with electric motors fed from overhead wires. I think some European cities still use them.
They died out in the UK because of the cost of maintaining the overhead line and, IIRC, the capital cost of the vehicles (few systems => few vehicles ordered => high cost per unit). Therefore, sadly, not every kind of electric bus meets the cheap-infrastructure criterion.
50 and 25 years sound suspiciously like two and one human generations respectively. And 50 years is roughly an academic career from start of university education through to retirement. Maybe we only get the new stuff as the old guard are replaced?
There is an argument that gross computing power lessens the near for, and appeal of, "cutting edge mathematics". Researchers are less likely to work for a tricky, analytical solution if they can brute-force a numerical one for less effort.
Yes, I agree. But in the domestic-broadband contracts (in the UK) I have seen, the over-subscription and contention are written in. We need some ISPs to offer better contracts. If you have such a contract I'm very happy for you and please tell us all who, where and how much.
Suppose you write in Perl or Python and you use a lot of 3rd-party libraries. My personal experience is that the modules available for Perl are at a lower, more geekful level than those for Python. In Perl one gets four or five modules off CPAN and spends a painful few days bashing them together into a solution. In Python, one more often gets a canned solution that works immediately.
If that carries over into the corperate domain - and it's a big if - then it makes Perl look expensive. And poorly-trained developers will do better with the canned solutions than with the low-level stuff.
Disclaimer: my experience is coding for scientists with FOSS and may not translate to the corporate domain. And I haven't written very much Python. :-/
The ones I've paid for in the last few years are oXygen, an XML editor; Lazgo Keystore Explorer a GUI programme for managing keystores; and Smartdraw, vector-graphics software for illustrating design documents. I also had access to a project license for a UML system (Together Control Centre, sold by Borland last time I checked).
oXygen I consider to be good value. Keystore Explorer could possibly be replaced by a decent FOSS tool but it's very cheap and handy so I haven't looked. Smartdraw is only for Windows and not useful to me now, but it paid its way in the past. Nowadays, I'd probably manage with the drawing tool in Open Office; back in the day this wasn't practical. Together CC, the only expensive one, is utterly ghastly to use and really deserves to be driven out by FOSS. But last time I looked, maybe 2 years ago now, the FOSS alternatives still weren't as good for what I needed to do.
*Grits teeth* No, I don't want commercial tools regardless of quality. I want each tool to be evaluated on its merits, and I want alternatives in each category. Sometimes the commercial tools are good alternatives, sometimes not. If the commercial tools go off the market, then I lose choice. If my organization won't pay for tools, then I lose choice. Both are bad.
I am not in opposition to FOSS. I make my living writing exclusively FOSS; I am privileged in this respect.
The problem here is with management, not free software.Yes! I agree! See informed, rational decisions, lack of, in other posts of this thread.
No of course not. Listen, I don't want the free software to go away. I want it to fulfil its hype and be as good as the paid-for stuff. In cases where that hasn't happened yet, I want the commercial stuff still to be available.
They shouldn't, and they shouldn't make the opposite assumption either. All tools should be evaluated on their merits. My point was that if a commercial tool is sufficiently better, in context, than its FOSS competition, then it should be bought and used.
From a purely rational perspective, I can not see how a commercial business model could produce a better product.Really? Or is that just rhetoric? Maybe a commercial product could come out better because: it gets more hours of input from an appropriate number of good developers; it gets better QA; its documentation is written by skilled technical authors rather than unskilled coders. Not in all cases, but it's possible.
Just because the GPL allows many top developers to work on a FOSS project it doesn't follow that they will.
Someone making a profit from the work of their programmers can not improve the quality of their work.Manifestly false.
I think we're both generalizing and 'if'-fing a little too much. Every case should be examined separately. We can safely assume Qcad is not a real replacement for AutoCAD, whereas OOo will be more than enough for the majority of MSOffice users.
I concur. The problem is that, where I've worked, cases aren't examined on their merits. There is an assumption that a FOSS alternative is always equivalent to the paid-for thing. Got GIMP, not gonna pay for Photoshop, etc.If that were true, most places where employees only use email, web browsing and office software would be installing Linux instead of the almost ubiquitous Windows.
That may be a bad example. For corporate buyers, TCO matters, and many perceive (possibly wrongly) that it's lower for Windows if they already have some Windows infrastructure. For consumers, Windows looks cost-free as the cost is hidden in the price of the machine and most don't know how to buy OS-free hardware.
So next time my management refuse to buy a $200 tool and I lose a week of working time with an inferior FOSS equivalent that's me saved is it? Even if I have to make up the lost week in unpaid overtime? Good for my soul, maybe.
Don't get me wrong, I like FOSS software, but I do need it to work. It's for using, not for looking at. If I need a tool and the FOSS versions are inadequate, then I need the commercial version, at least until the FOSS world catches up.
Bad free tools don't increase my productivity compared to good, paid-for tools. They might increase a society's productivity, and I think that's what your rant was really about. But that doesn't help me as an individual. I'm happy that people who can't pay for tools get cost-free ones, but that shouldn't stop my organization buying better tools when appropriate.
PS: informed, rational decisions are an assumption in free-market economics. The fact that you don't like capitalism doesn't make this untrue, as you seem to imply.
Pure free-market economics assume that the players are making rational informed decisions. In software acquisition, that assumption fails often.
If the more-expensive tool saves time worth more than its cost, then the appropriate free-market choice is to invest. My experience is that buyers at all levels won't do that when there's a cost-free alternative. They'd rather waste time (=money) or lose quality (=money due to cost of fixing later) than spend capital.
Doubtless. But inferior, cost-free tools sometimes make better, commercial ones unsellable. That is the tragedy.
Shorter fibres: isn't that what happens to paper after multiple recyclings? I wonder if this is what to do with the recycled conventional paper that's no longer good enough evn for newsprint.
The robust-scanner one, almost certainly. This is likely an easier job than hardening an interactive web-browser. Their robot has no need to execute anything it comes across, so downloaded script needn't be allowed to execute anything, ever. It has no need to render any of the media, so none of the image-library attacks can work. They don't have to keep anything that they scan, so no save-to-disc code. In short, they can maintain exceptionally strong separation between their scanner and its host.
If they were paranoid enough, they could run the robot in a virtual machine and reinstall that after each scanning run. I have no idea if they consider that worthwhile.
Egypt does security theatre routinely. The antiquities sites all have metal detectors on the entrances. They are all turned on and the guards make visitors walk thorugh. The alarms go off for maybe a third of the visitors - coins, belt buckles etc., nothing bad. The attendants never check visitors who set of the alarms. They treat the detectors like some magic amulet that wards off evil; no human intervention needed.
Yes there are. The carpet-weaving industry in the UK still uses card-programmed looms (I have a friend who is employed to load card decks into the machines).