OK mate, I've read enough of your BS to know there's no point arguing with stupid people: they beat you down to their level and then win from experience.
Hey APK, why are you going around misusing my username? I've worked hard on Slashdot to build my reputation and you come along and spout garbage and start flame-wars for no reason at all. Just do a brother a favour, back off and leave me alone. Deal?
Cheers, I'll take a second look. Unfortunately we have multiple people across lots of varied platforms, which is what drew us to Hangouts in the first place.
I realise that Hangouts is still here, it's just the API that's being discontinued. But as others have pointed out, it probably means the site is going to be canned at some point in the near future.
The main problem is that they claim "the problem" with the impact is that we can't find any evidence or residue from Theia. Well, a lot of the models show that the impacter's core (which would have already differentiated and have a metal core) mostly goes into the core of the earth, and the silicate portion mixes with the bulk silicate earth (BSE). The moon is mostly formed from the earth's mantel, not from the impactor, and this is why the moon has the same composition as the BSE. The problem is working out why there is little to no isotopic fractionation between the moon and the earth, since conventional wisdom suggests a hot violent process such as the giant impact would have resulted in such fractionation.
The most useful part of the article is the suggestion that we go to venus. The similarities and differences in isotopic fractionation between the moon and the earth can only really be interpreted with a greater understanding of isotopic variation among the terrestrial planets, and failed planets (e.g. Vesta) in the asteroid belt.
For more information I recommend reading this paper which is fortunately funded to be open access.
I applaud your criticism of management co-processors (Intel vPro, Intel ME, Intel AMT, AMD PSP) as while they may have a place in the enterprise world (assuming those IT techs can secure them properly) they are an anathema to home users. I don't support the idea that we all have to be lumped with these back-doors to our systems and believe we should all be able to either choose CPUs that lack them, or disable them entirely (motherboard jumper anyone?) as we wish. Of course I'm voting with my wallet, but the options are shrinking year by year...
Can you expand on why you think Intel is the lesser of two evils here?
For those unsure of what this is about, there is a discussion here and some really good info here. But look out for Leah Rowe at that last link: that bitch be crazy!! [Citation]
Cam Frog looks bloody awful. I can tell from their tiered service plans, which I can't even copy/paste here since their website is horrendously broken.
Sorry, a few of us use Linux, so it's a no-go anyway...
But for family get togethers they are perfect. My family tried most video conferencing programs and eventually settled on Hangouts as the easiest for the oldies to handle. Google screwed us with the new G+ interface last year, as they dropped (or hid it so well none of us could find it) the Hangouts tools and we had to migrate everyone to hangouts.google.com to get going again.
Anyone able to suggest an alternative?
We need it to be cross platform, as we have a dozen family members on 3 continents using different systems, and "good enough" for us to see the babies, tell some jokes, and just catch up. Suggestions more than welcome!
Databases are piss-easy these days. Any competent programmer can code up a custom web site with DB support, forms, basic customer-service tools, as well as helpful information for the visitors, in about 5 days (max). That's with no 3rd party libraries.
Agree completely. Surely contract law dictates that since the government can't legally break the law they can't circumvent FOI legislation? Hence, the NDA is void?
2014 invasion of Ukraine 2008 invasion of Georgia 1979 invasion of Afghanistan 1968 invasion of Czechoslovakia
US Invasions
2003 invasion of Iraq 2001 invasion of Afghanistan 1994 invasion of Haiti 1991 invasion of Iraq 1989 invasion of Panama 1983 invasion of Grenada
What about (out of interest) invasions by Israel? Do we count these as US? Israel is the largest beneficiary of US foreign aid, and certainly appears for all intents and purposes to be a US proxy way in the middle east.
2014 invasion of Gaza 2008 invasion of Gaza 2006 invasion of Lebanon 1982 invasion of Lebanon 1978 invasion of Lebanon 1967 invasion of Egypt 1967 invasion Syria 1967 invasion of Jordan
I'm sure the naysayers will say it's convenient that I cut off at 66 and missed some Russian activity in eastern Europe, but the US was active in the DR in 65, Cuba in 61, North Korea in 51 and 50, so actually I didn't go back far enough. Certainly, reviewing the past 100 years then without a doubt the US is a clear antagonist on the global stage. To say otherwise is to be revisionist and to ignore history, which over the ages many smarter than I have advised against doing.
We may be talking at crossed-purposes, but Kodak very much did "miss the boat". Can you buy a "Kodak" brand digital camera today? If so, is it well regarded as a market leader for quality? That's what Kodak used to have in the film industry, and they lost it all due to poor management and market strategy. They even invented the first "digital camera"...what idiots.
"Do you really think that the world would be a better place if the US simply disengaged across the board? Should Japan and Korea be the only entities in the front line dealing with China's territorial expansionism?"
Ha ha, very funny. As it happens, and not that you care, but I gave up helping people to "be popular" over a decade ago. Best thing I ever did. Helping people is a gift, not an obligation, and when you give freely you become free yourself. Try it sometime, it will make you a much better person.
I went to set up a iPod Touch for a friend the other day. Can't believe that in addition to asking for a lot of personal identifying information (date of birth, address, email, three security questions quite similar to those used in banking, etc...) the Apple ID account creation process *forces* you to enter a credit card to sign up. What possible justification has Apple made to substantiate this requirement? All I want to do is install a few programs to make the little thing more useful for my friend. Even though this item was a Christmas present, I advised her to return it and tell her other friend that it's just a shiny paperweight.
"Multithreading you suggest (that I do albeit only on things benefitting concurrent ops) can backfire - how?"
Any one process is not universally aware of the load from other processes on the system. By over-optimizing you have tweaked your code execution to be as fast as possible under the exact same conditions at the time of optimization: this relates to CPU workload, memory bandwidth, memory fragmentation, slower IO to storage, and even bus communications such as streaming 10Gbit/s LAN packets through to memory and/or disk, for example.
All I'm saying is that when that code is executed on a different system under different circumstances you no longer have any claim that your over-optimized solution will run any better. In general, I would argue (from experience, no citations here) that it will run worse.
As such, and as frustrating as it sounds, for a general purpose program, designed to be run on millions (or more) work stations, you are best leaving the OS-interaction tweaks to the OS developers, and focus instead on your application logic.
You're a smart guy APK, though obviously with a few screws loose, and I think you know all this and are probably arguing from the same position.
The only reason I've bothered to reply to you here is in case some young developers are wondering what all the fuss is about and may even learn a little by reading through our argument.
It sounds good, but you run the risk of over-optimization, which means your code runs fast on your system, but there is no guarantee that it will do so on any general system. In other words: your users may be seeing worse performance than you are.
APK, my friend, the whole point is you can't. The non-deterministic nature of multi-threaded tasks makes any sort of general optimization impossible. I think you know this, since you somewhat allude to it in your post. Any given parallel task can only be made more efficient in parallel through domain specific optimizations. This is why we don't have a magic "parallel" tick box in the compiler options. For anyone else who's reading, and is interested in these kinds of things, it all comes down to how you can partition the work amongst processes/threads and then how much communication has to occur between the threads to get through the workload. Some tasks can be partitioned really easily and benefit trivially from parallel processing. Others (most?...we need a CS PhD to comment here) show a reasonable trade off between task partitioning and inter-process communications. While others, the "trouble makers" require so much inter-process communications that a parallel task can actually be slower than a linear task.
The whole root cause of this tirade with APK was simply about whether making the current form (on Windows? It wouldn't actually matter) have real-time priority on the CPU would make it run any faster. All I did was point out that most well-written parallel tasks run the actual work-load on background threads, and any attempt to give the current form thread priority would also have to apply the same priority to the entire process and even then, the benefit would be marginal. APK actually agrees with me, since he points out that he tried this and found no benefit, as expected: "& I suggested it as I tried it - no gain"
So I think we're in agreement, which is good, but APK wants to harass me for some reason, possibly for my handle, who knows...Happy New Year everyone!
Slashdot Mirror
OK mate, I've read enough of your BS to know there's no point arguing with stupid people: they beat you down to their level and then win from experience.
Have a nice life...
Hey APK, why are you going around misusing my username? I've worked hard on Slashdot to build my reputation and you come along and spout garbage and start flame-wars for no reason at all. Just do a brother a favour, back off and leave me alone. Deal?
Cheers, I'll take a second look. Unfortunately we have multiple people across lots of varied platforms, which is what drew us to Hangouts in the first place.
I realise that Hangouts is still here, it's just the API that's being discontinued. But as others have pointed out, it probably means the site is going to be canned at some point in the near future.
I read the article.
The main problem is that they claim "the problem" with the impact is that we can't find any evidence or residue from Theia. Well, a lot of the models show that the impacter's core (which would have already differentiated and have a metal core) mostly goes into the core of the earth, and the silicate portion mixes with the bulk silicate earth (BSE). The moon is mostly formed from the earth's mantel, not from the impactor, and this is why the moon has the same composition as the BSE. The problem is working out why there is little to no isotopic fractionation between the moon and the earth, since conventional wisdom suggests a hot violent process such as the giant impact would have resulted in such fractionation.
The most useful part of the article is the suggestion that we go to venus. The similarities and differences in isotopic fractionation between the moon and the earth can only really be interpreted with a greater understanding of isotopic variation among the terrestrial planets, and failed planets (e.g. Vesta) in the asteroid belt.
For more information I recommend reading this paper which is fortunately funded to be open access.
This needs more attention.
Unfortunately, the circus has now died down, and most of the enraged posters above will not read this far down to be informed. Sigh :(
I applaud your criticism of management co-processors (Intel vPro, Intel ME, Intel AMT, AMD PSP) as while they may have a place in the enterprise world (assuming those IT techs can secure them properly) they are an anathema to home users. I don't support the idea that we all have to be lumped with these back-doors to our systems and believe we should all be able to either choose CPUs that lack them, or disable them entirely (motherboard jumper anyone?) as we wish. Of course I'm voting with my wallet, but the options are shrinking year by year...
Can you expand on why you think Intel is the lesser of two evils here?
For those unsure of what this is about, there is a discussion here and some really good info here. But look out for Leah Rowe at that last link: that bitch be crazy!! [Citation]
Cam Frog looks bloody awful. I can tell from their tiered service plans, which I can't even copy/paste here since their website is horrendously broken.
Sorry, a few of us use Linux, so it's a no-go anyway...
But for family get togethers they are perfect. My family tried most video conferencing programs and eventually settled on Hangouts as the easiest for the oldies to handle. Google screwed us with the new G+ interface last year, as they dropped (or hid it so well none of us could find it) the Hangouts tools and we had to migrate everyone to hangouts.google.com to get going again.
Anyone able to suggest an alternative?
We need it to be cross platform, as we have a dozen family members on 3 continents using different systems, and "good enough" for us to see the babies, tell some jokes, and just catch up. Suggestions more than welcome!
Databases are piss-easy these days. Any competent programmer can code up a custom web site with DB support, forms, basic customer-service tools, as well as helpful information for the visitors, in about 5 days (max). That's with no 3rd party libraries.
Seconded.
Agree completely. Surely contract law dictates that since the government can't legally break the law they can't circumvent FOI legislation? Hence, the NDA is void?
But what is your point? Yes, Russia invaded Ukraine.
Now let's review the past 50 years: 1966 - 2016 (inclusive) of US vs Russian invasions.
Russian (and USSR) Invasions
2014 invasion of Ukraine
2008 invasion of Georgia
1979 invasion of Afghanistan
1968 invasion of Czechoslovakia
US Invasions
2003 invasion of Iraq
2001 invasion of Afghanistan
1994 invasion of Haiti
1991 invasion of Iraq
1989 invasion of Panama
1983 invasion of Grenada
What about (out of interest) invasions by Israel? Do we count these as US? Israel is the largest beneficiary of US foreign aid, and certainly appears for all intents and purposes to be a US proxy way in the middle east.
2014 invasion of Gaza
2008 invasion of Gaza
2006 invasion of Lebanon
1982 invasion of Lebanon
1978 invasion of Lebanon
1967 invasion of Egypt
1967 invasion Syria
1967 invasion of Jordan
I'm sure the naysayers will say it's convenient that I cut off at 66 and missed some Russian activity in eastern Europe, but the US was active in the DR in 65, Cuba in 61, North Korea in 51 and 50, so actually I didn't go back far enough. Certainly, reviewing the past 100 years then without a doubt the US is a clear antagonist on the global stage. To say otherwise is to be revisionist and to ignore history, which over the ages many smarter than I have advised against doing.
"Kodak did not miss photo boat"
We may be talking at crossed-purposes, but Kodak very much did "miss the boat". Can you buy a "Kodak" brand digital camera today? If so, is it well regarded as a market leader for quality? That's what Kodak used to have in the film industry, and they lost it all due to poor management and market strategy. They even invented the first "digital camera"...what idiots.
"About 50% of them are discarded due to incomplete information or the voter being ineligible."
Well there's your problem right there!! And you call that democracy? Who gets to make those decisions? Sounds absolutely ripe for abuse!!
"Do you really think that the world would be a better place if the US simply disengaged across the board? Should Japan and Korea be the only entities in the front line dealing with China's territorial expansionism?"
Yes I do.
"in fact the way Linux is now(the people involved) I probably trust it less than windows at this point."
Care to elaborate?
I don't trust Canonical, and I'm wary (but not zealously opposed to) of systemd. Is there anything else shifty going on that should be discussed?
Ha ha, very funny. As it happens, and not that you care, but I gave up helping people to "be popular" over a decade ago. Best thing I ever did. Helping people is a gift, not an obligation, and when you give freely you become free yourself. Try it sometime, it will make you a much better person.
"the Russians hacked a couple of people at the Democratic Party (maybe) and embarrassed the hell out of them"
And even this has not been substantiated to the public by the US intelligence agencies.
I went to set up a iPod Touch for a friend the other day. Can't believe that in addition to asking for a lot of personal identifying information (date of birth, address, email, three security questions quite similar to those used in banking, etc...) the Apple ID account creation process *forces* you to enter a credit card to sign up. What possible justification has Apple made to substantiate this requirement? All I want to do is install a few programs to make the little thing more useful for my friend. Even though this item was a Christmas present, I advised her to return it and tell her other friend that it's just a shiny paperweight.
"they don't impress me @ all - they're just men too man"
Agree 100%!!!
And what can we do? Hope it doesn't degrade into WW3?
"Multithreading you suggest (that I do albeit only on things benefitting concurrent ops) can backfire - how?"
Any one process is not universally aware of the load from other processes on the system. By over-optimizing you have tweaked your code execution to be as fast as possible under the exact same conditions at the time of optimization: this relates to CPU workload, memory bandwidth, memory fragmentation, slower IO to storage, and even bus communications such as streaming 10Gbit/s LAN packets through to memory and/or disk, for example.
All I'm saying is that when that code is executed on a different system under different circumstances you no longer have any claim that your over-optimized solution will run any better. In general, I would argue (from experience, no citations here) that it will run worse.
As such, and as frustrating as it sounds, for a general purpose program, designed to be run on millions (or more) work stations, you are best leaving the OS-interaction tweaks to the OS developers, and focus instead on your application logic.
You're a smart guy APK, though obviously with a few screws loose, and I think you know all this and are probably arguing from the same position.
The only reason I've bothered to reply to you here is in case some young developers are wondering what all the fuss is about and may even learn a little by reading through our argument.
Happy NYE!
It sounds good, but you run the risk of over-optimization, which means your code runs fast on your system, but there is no guarantee that it will do so on any general system. In other words: your users may be seeing worse performance than you are.
Hi APK, see my other post. I think we're in agreement, which is cool. Have a happy new year mate, all the best,
NotAPK
APK, my friend, the whole point is you can't. The non-deterministic nature of multi-threaded tasks makes any sort of general optimization impossible. I think you know this, since you somewhat allude to it in your post. Any given parallel task can only be made more efficient in parallel through domain specific optimizations. This is why we don't have a magic "parallel" tick box in the compiler options. For anyone else who's reading, and is interested in these kinds of things, it all comes down to how you can partition the work amongst processes/threads and then how much communication has to occur between the threads to get through the workload. Some tasks can be partitioned really easily and benefit trivially from parallel processing. Others (most? ...we need a CS PhD to comment here) show a reasonable trade off between task partitioning and inter-process communications. While others, the "trouble makers" require so much inter-process communications that a parallel task can actually be slower than a linear task.
The whole root cause of this tirade with APK was simply about whether making the current form (on Windows? It wouldn't actually matter) have real-time priority on the CPU would make it run any faster. All I did was point out that most well-written parallel tasks run the actual work-load on background threads, and any attempt to give the current form thread priority would also have to apply the same priority to the entire process and even then, the benefit would be marginal. APK actually agrees with me, since he points out that he tried this and found no benefit, as expected: "& I suggested it as I tried it - no gain"
So I think we're in agreement, which is good, but APK wants to harass me for some reason, possibly for my handle, who knows...Happy New Year everyone!