"Overall, we manage to get a 'bullet-proof' patch in about 4-5 weeks (from coding->QA->Build/Packaging->shipment), which I consider not so bad."
It depends on where in that sequence you spend the time. If all the time is going into build/package and shipment then your customers are right to demand a speed-up.
If your QA time was, say, one day, then I would look for a turn-around time (on the critical issues) of 48hrs plus the time to find the root cause and code a fix which could be anything from an hour to a year. If the customers don't know the root cause then they can't expect such a quick response.
Yes, I appeciate that you may need a lot longer than a day for QA. But if QA is a large part of the 4-5 weeks then maybe you need to work on this?
"...all you need to do is get gcc to work on your system..."
Where's the sarcasm tag? "All you need to do" and "get gcc to work on your system" don't belong in the same sentence. Do you really expect people to port a compiler to a new platform just to compile OSS code? How many can do that reliably? How many wuld want to take the time?
"Frankly if this is the best attack you can muster against GNU, then we can sit back and relax, we won."
Frankly, this is why you're not winning. The FOSS stuff doesn't get used when the barriers to adoption are made stupidly high.
You're right: Civ is turn based. However, I'd guess that RT vs. turn-based matters little to the bots; I'm guessing that they are fast enough for the RT aspect not to hinder them.
This tells us more about the game-play balance in AoM than how to approach games in general. I'd be more interested in seeing these bots play CiV 4 where I doubt that neurotic behaviour would triumph.
"Shorting is simply selling what you don't have. Stock certificates are just pieces of paper. And your broker probably has a huge inventory of stock in his portfolio. So when you "short" something you are basically telling your broker that you agree to sell a stock that you don't own. There's no problem with this, since there are rigid mechanisms in place (today) to avoid fraud."
The whole concept sounds fraudulent to me: pretend to sell stock you don't have so as to artificially depress its value so that you can profit without actually investing. It defeats the Market's purpose of promoting investment and makes things unecessarilt unstable.
PS: thanks for the clear explanation of the shorting process.
If you choose to get a telescope, as opposed to binoculars, it's vital to get one that points and tracks well. Otherwise, you spend too long finding targets. (This applies all the way up to 8m-class research telescopes.) Nothing kills the experience as much as not being able to find what you want.
If you get a telescope that you set up at the beginning of the night , as opposed to one permanently fixed in its own enclosure, then you need a drive system that is easy and quick to align. If you have to spend 2 hours each night aligning then it's too tedious. (Again, also true of research-level kit:-/) I believe that drives sold with good makes (Meade? Celestron?) self-align quite well and quickly; but beware of cheaper versions.
If you want to do long, timed exposures, then you either need a drive that tracks outrageously well (sub-arcsecond accuracy for best results) by itself, or you need a guiding system. A manual guiding system is an eyepiece with crosshairs (or a camera) coaxial with the main telescope. You continually tweak the drive motion to keep a bright star centred in the guider while the telescope itself observes something fainter. An autoguider (preferred) is a guide camera + computer that does it for you.
Quality of drive is as important as quality of optics for serious work and photography. It's much more important than the apeture of the telescope. A big telescope that doesn't point or track is useless.
"That's really too bad, because obviously important tech decisions should not be made based on the product name. "
Yes, but the decision isn't made on the name directly, but on the inferred professionalism of the people who chose the name. Lots of products to choose from; not enough time to test them all. Do I bother testing the one that sounds like a joke? Probably not. Incorrect conclusion, but if you don't know a priori that the Ubuntu teram are competent you wouldn't guess it from the name.
My team uses the "bouncy-castle" Java crypto library; it's one of the better FOSS Java products. When we use it in Java-web-start apps the users sometimes get a dialogue saying "Do you want to run this component signed by The Legion of the Bouncy Castle?" They think that's the name of a cracker gang and they think "trojanized" and they click "no". The name hurts customers of BC's customers.
"So yes, I admire your sentiments, but anywhere outside of the best places in Japan, I've never seen them in practice. I've never been to the UK, but I presume they have worse problems than the US given all the surveillance cameras they've felt the need to install in recent years."
Based on your description, and other anecdotes, I'd say that the crime problems in the UK are not quite as bad as in the USA. The cameras are mainly an arse-covering exercise by town and city councils. The councils get complaints from citizens about crime, and are expected to fix things, but they don't run the police forces (even in our largest cities), so can't choose to post more police officers. They post cameras instead.
"I suppose somebody's out there looking to make a weapon out of the thing"
If they do it, it will be the classic SF "disintegrator" weapon. But I doubt that it will be efficient over a macroscopic volume or range. Easier just to pump in kinetic energy until the target comes apart in the traditional way.
You need specialized scanning machines for astronomy. Office equipment doesn't do the job.
The plates have to be scanned in transmission, not reflection (they are photographic negatives).
You have to accurately measure the darkness of the plate in order to deduce the light intensity that fell on it.
Office scanners only approximately measure
the light and dark - enough for visual presentation, not enough to do maths with the result.
My colleagues in the UK had such a scanner. It was ~7 tonnes of metal, glass and electronics (heavy so as
to be very stable), lived in its own building and needed several clever people to keep it running. Building one
of these (or cloning one you already have so as to work faster) could cost a big chunk of the $5M.
The scanner I knew took ~ 30 minutes to scan a plate. For the harvard collection, choose between
one scanner (which they may not have; otherwise why did they wait until now to start the project?) and
a long project with big sallary bill, or multiple scanners, at extra capital cost, and less money for people.
Some human-defined canine behaviour is inherited without training. My parents setters showed gun-dog behaviour despite never having been trained or worked with dogs on a shoot. My partner's West Highland terrier is included to hunt rodents without any training (ratting is the main role for Westies when kept as working dogs).
Seems to me that GPL is becoming like a large, badly factored code-base. It's very hard to see all of what it does in practice, even for experts and specialists; in TFA PJ notes missing an implication of the MS-Novell situation. The only way to be sure about GPL now seems to be to test it - and the only test is a live one, in court, since we don't have legal "evaluation systems".
When code gets like that, we say it's bad code and want to refactor it; or dump it and replace.
When FOSS code gets too hard to understand, then the pool of maintainers dries up and the ESR "bazaar" model doesn't work. If the complication applies also to the users, as in an API, then the software drops out of use. Licences can suffer the same fate.
I read TFA on Selden. He was licencing the patented tech for 0.75% of sales, a rather small burden for a manufacturer. He wasn't using the patent to block use of the tech by competitors. I presume that Ford could have chosen to pay the royalty and the case would never then have come to court. Isn't this how patents are supposed to work?
I understand free software perfectly, thank you, including the rather important point that it only comes with highly selective freedoms. My point was that the pun against "having your cake" conceals one of the truths of the analogy. Socialists offering free beer to workers was a tactic to conjure apparent political support from non-politicized voters. Offering software to businesses gratis is similar; by and large, the businesses don't care about the politics, they just want the "gift". Businesses using free software are not necessarily supports of free software.
That's an option if they were starting now, but they are already using and porting to a new OS is going to be really hard. For TiVO, GPL v3 looks like bait and switch: "yeah you can use our OS for free...oops, now you're committed, you have to give what we want." A commercial company might do this to extract licence money. FSF is doing the same with political capital.
PS: the whole point about the free beer is that it's a political bribe to the workers who want to drink it, not to "keep" it.
Except that it didn't meet her "non-negotiable" requirements: no editing of text config files and h/w works with included drivers and no fiddling. Strange how people set reasonable criteria and then ignore them.
Yes, and in that context the bit in the GP post about high-res displays is relevant. We can't fairly compare print and screen readability until the screen can display text with about 300 pixels per inch and with an entire page on display at once.
The other advantage in print is the lack of distractions. The page isn't cluttered with navigation aids.
But what if AllegedOwner is the little guy and InnocentCompany is vast with top-flight lawers? Surely you don't want to make patent suits only available to rich organizations?
I assume that, under current law, if a patent suit fails then the party sued is awarded legal costs. This seems sufficient.
More importantly, there must be a mechanism for patents to be reassessed after failed suits, especially if the defense submits prior art. Is this piece of the law missing or broken?
It should not be necessary for a party to defend against a patent suit if that patent has previously failed in court due to prior art. The patent should just go away.
"I suppose that will happen the same day as airline pilots report seeing pigs out the window, [...]"
Flying pigs and music promotion: been done. A giant, inflatable pig was flown for the cover art of one of the Floyd albums (Animals, IIRC).
It got loose from its tethers and floated into the Heathrow Airport approaches...
Clearly, iTunes is restricted due to licensing arrangements not of Apple's making (see other posts in this topic). Equally clearly, these restrictions don't benefit Apple: it costs them little to offer extra tracks, even for low-volume sales (also noted in an earlier post in this thread). Therefore, cannot Apple use their commercial leverage to get the licensing changed? The record companies may listen if they think they're going to be trading with the Master Music Archive of all time.
The British Labour party has been leaking credibility since the 1990s; they have failed on so many manifesto promises that nobody much trusts them now; they're fearful for their chance of re-election and prepared to do bad things to save themselves. The Labour cabinet seems to be encouraging a climate of fear -- terrorists, criminals, sex attackers, immigrants, social breakdown -- so that they that can get credit for protecting us. Of course, the "protection" consists in laws that restrict citizens, and the threats are exaggerated. Just routine FUD.
GNU was started as a project when there was relatively little OSS and very little free (as-in-speech) software. It expanded into a kind of moral vacuum, and got adopted by default by many who care about free software. Now, there is FOSS, at some level of development, for just about everything and bringing up new or under-developed parts of the GNU collection is harder; they'd have to fight for mindshare.
If the GNU developers reason like RMS and FSF, then they will choose the course that maximizes use of free software over the course that maximizes GNU and use of free software. They won't add or revive projects that compete with established FOSS as that confuses new users. They might work less on GNU products and more on other FOSS projects. In an extreme case, they might wind up GNU altogether and putr their effort elsewhere. They can't go that far, of course, as the pervasive GNU products (e.g. compilers, GNU bits of GNU/Linux) have to be maintained.
Alternatively, if GNU carry on with "unnecessary" projects like HURD, or if they under-resource their critical, legacy products, then one can conclude that they're splitting away from FSF.
Slashdot Mirror
It depends on where in that sequence you spend the time. If all the time is going into build/package and shipment then your customers are right to demand a speed-up.
If your QA time was, say, one day, then I would look for a turn-around time (on the critical issues) of 48hrs plus the time to find the root cause and code a fix which could be anything from an hour to a year. If the customers don't know the root cause then they can't expect such a quick response.
Yes, I appeciate that you may need a lot longer than a day for QA. But if QA is a large part of the 4-5 weeks then maybe you need to work on this?
Where's the sarcasm tag? "All you need to do" and "get gcc to work on your system" don't belong in the same sentence. Do you really expect people to port a compiler to a new platform just to compile OSS code? How many can do that reliably? How many wuld want to take the time?
Frankly, this is why you're not winning. The FOSS stuff doesn't get used when the barriers to adoption are made stupidly high.
You're right: Civ is turn based. However, I'd guess that RT vs. turn-based matters little to the bots; I'm guessing that they are fast enough for the RT aspect not to hinder them.
This tells us more about the game-play balance in AoM than how to approach games in general. I'd be more interested in seeing these bots play CiV 4 where I doubt that neurotic behaviour would triumph.
The whole concept sounds fraudulent to me: pretend to sell stock you don't have so as to artificially depress its value so that you can profit without actually investing. It defeats the Market's purpose of promoting investment and makes things unecessarilt unstable.
PS: thanks for the clear explanation of the shorting process.
If you choose to get a telescope, as opposed to binoculars, it's vital to get one that points and tracks well. Otherwise, you spend too long finding targets. (This applies all the way up to 8m-class research telescopes.) Nothing kills the experience as much as not being able to find what you want.
If you get a telescope that you set up at the beginning of the night , as opposed to one permanently fixed in its own enclosure, then you need a drive system that is easy and quick to align. If you have to spend 2 hours each night aligning then it's too tedious. (Again, also true of research-level kit :-/) I believe that drives sold with good makes (Meade? Celestron?) self-align quite well and quickly; but beware of cheaper versions.
If you want to do long, timed exposures, then you either need a drive that tracks outrageously well (sub-arcsecond accuracy for best results) by itself, or you need a guiding system. A manual guiding system is an eyepiece with crosshairs (or a camera) coaxial with the main telescope. You continually tweak the drive motion to keep a bright star centred in the guider while the telescope itself observes something fainter. An autoguider (preferred) is a guide camera + computer that does it for you.
Quality of drive is as important as quality of optics for serious work and photography. It's much more important than the apeture of the telescope. A big telescope that doesn't point or track is useless.
"That's really too bad, because obviously important tech decisions should not be made based on the product name. "
Yes, but the decision isn't made on the name directly, but on the inferred professionalism of the people who chose the name. Lots of products to choose from; not enough time to test them all. Do I bother testing the one that sounds like a joke? Probably not. Incorrect conclusion, but if you don't know a priori that the Ubuntu teram are competent you wouldn't guess it from the name.
My team uses the "bouncy-castle" Java crypto library; it's one of the better FOSS Java products. When we use it in Java-web-start apps the users sometimes get a dialogue saying "Do you want to run this component signed by The Legion of the Bouncy Castle?" They think that's the name of a cracker gang and they think "trojanized" and they click "no". The name hurts customers of BC's customers.
"So yes, I admire your sentiments, but anywhere outside of the best places in Japan, I've never seen them in practice. I've never been to the UK, but I presume they have worse problems than the US given all the surveillance cameras they've felt the need to install in recent years."
Based on your description, and other anecdotes, I'd say that the crime problems in the UK are not quite as bad as in the USA. The cameras are mainly an arse-covering exercise by town and city councils. The councils get complaints from citizens about crime, and are expected to fix things, but they don't run the police forces (even in our largest cities), so can't choose to post more police officers. They post cameras instead.
"It's hardly coincidental that coffee and tea caught on in Europe just as the first factories were bringing in the industrial revolution."
Cute idea, but actually coffee was established by c.1650AD and tea by maybe 100years later. Maybe it took everybody a while to wake up.
From the researcher's point of view:
Step 3 doesn't have to involve selling technology.
"I suppose somebody's out there looking to make a weapon out of the thing"
If they do it, it will be the classic SF "disintegrator" weapon. But I doubt that it will be efficient over a macroscopic volume or range. Easier just to pump in kinetic energy until the target comes apart in the traditional way.
So it's good if a FOSS developer makes an ugly-but-useful clone of something, but bad when MS do it?
You need specialized scanning machines for astronomy. Office equipment doesn't do the job.
My colleagues in the UK had such a scanner. It was ~7 tonnes of metal, glass and electronics (heavy so as to be very stable), lived in its own building and needed several clever people to keep it running. Building one of these (or cloning one you already have so as to work faster) could cost a big chunk of the $5M.
The scanner I knew took ~ 30 minutes to scan a plate. For the harvard collection, choose between one scanner (which they may not have; otherwise why did they wait until now to start the project?) and a long project with big sallary bill, or multiple scanners, at extra capital cost, and less money for people.
Some human-defined canine behaviour is inherited without training. My parents setters showed gun-dog behaviour despite never having been trained or worked with dogs on a shoot. My partner's West Highland terrier is included to hunt rodents without any training (ratting is the main role for Westies when kept as working dogs).
Seems to me that GPL is becoming like a large, badly factored code-base. It's very hard to see all of what it does in practice, even for experts and specialists; in TFA PJ notes missing an implication of the MS-Novell situation. The only way to be sure about GPL now seems to be to test it - and the only test is a live one, in court, since we don't have legal "evaluation systems".
When code gets like that, we say it's bad code and want to refactor it; or dump it and replace.
When FOSS code gets too hard to understand, then the pool of maintainers dries up and the ESR "bazaar" model doesn't work. If the complication applies also to the users, as in an API, then the software drops out of use. Licences can suffer the same fate.
I read TFA on Selden. He was licencing the patented tech for 0.75% of sales, a rather small burden for a manufacturer. He wasn't using the patent to block use of the tech by competitors. I presume that Ford could have chosen to pay the royalty and the case would never then have come to court. Isn't this how patents are supposed to work?
I understand free software perfectly, thank you, including the rather important point that it only comes with highly selective freedoms. My point was that the pun against "having your cake" conceals one of the truths of the analogy. Socialists offering free beer to workers was a tactic to conjure apparent political support from non-politicized voters. Offering software to businesses gratis is similar; by and large, the businesses don't care about the politics, they just want the "gift". Businesses using free software are not necessarily supports of free software.
That's an option if they were starting now, but they are already using and porting to a new OS is going to be really hard. For TiVO, GPL v3 looks like bait and switch: "yeah you can use our OS for free...oops, now you're committed, you have to give what we want." A commercial company might do this to extract licence money. FSF is doing the same with political capital.
PS: the whole point about the free beer is that it's a political bribe to the workers who want to drink it, not to "keep" it.
Except that it didn't meet her "non-negotiable" requirements: no editing of text config files and h/w works with included drivers and no fiddling. Strange how people set reasonable criteria and then ignore them.
Yes, and in that context the bit in the GP post about high-res displays is relevant. We can't fairly compare print and screen readability until the screen can display text with about 300 pixels per inch and with an entire page on display at once.
The other advantage in print is the lack of distractions. The page isn't cluttered with navigation aids.
But what if AllegedOwner is the little guy and InnocentCompany is vast with top-flight lawers? Surely you don't want to make patent suits only available to rich organizations?
I assume that, under current law, if a patent suit fails then the party sued is awarded legal costs. This seems sufficient.
More importantly, there must be a mechanism for patents to be reassessed after failed suits, especially if the defense submits prior art. Is this piece of the law missing or broken? It should not be necessary for a party to defend against a patent suit if that patent has previously failed in court due to prior art. The patent should just go away.
"I suppose that will happen the same day as airline pilots report seeing pigs out the window, [...]"
Flying pigs and music promotion: been done. A giant, inflatable pig was flown for the cover art of one of the Floyd albums (Animals, IIRC). It got loose from its tethers and floated into the Heathrow Airport approaches...
This problem is an opportunity.
Clearly, iTunes is restricted due to licensing arrangements not of Apple's making (see other posts in this topic). Equally clearly, these restrictions don't benefit Apple: it costs them little to offer extra tracks, even for low-volume sales (also noted in an earlier post in this thread). Therefore, cannot Apple use their commercial leverage to get the licensing changed? The record companies may listen if they think they're going to be trading with the Master Music Archive of all time.
The British Labour party has been leaking credibility since the 1990s; they have failed on so many manifesto promises that nobody much trusts them now; they're fearful for their chance of re-election and prepared to do bad things to save themselves. The Labour cabinet seems to be encouraging a climate of fear -- terrorists, criminals, sex attackers, immigrants, social breakdown -- so that they that can get credit for protecting us. Of course, the "protection" consists in laws that restrict citizens, and the threats are exaggerated. Just routine FUD.
Hitler did something similar in the 1930s.
GNU was started as a project when there was relatively little OSS and very little free (as-in-speech) software. It expanded into a kind of moral vacuum, and got adopted by default by many who care about free software. Now, there is FOSS, at some level of development, for just about everything and bringing up new or under-developed parts of the GNU collection is harder; they'd have to fight for mindshare.
If the GNU developers reason like RMS and FSF, then they will choose the course that maximizes use of free software over the course that maximizes GNU and use of free software. They won't add or revive projects that compete with established FOSS as that confuses new users. They might work less on GNU products and more on other FOSS projects. In an extreme case, they might wind up GNU altogether and putr their effort elsewhere. They can't go that far, of course, as the pervasive GNU products (e.g. compilers, GNU bits of GNU/Linux) have to be maintained.
Alternatively, if GNU carry on with "unnecessary" projects like HURD, or if they under-resource their critical, legacy products, then one can conclude that they're splitting away from FSF.