Slightly OT: If power efficiency is important to you, then a Vega 56 might be a better choice than the RX590. Slightly lower TBP and slightly better performance at the same time. It is more expensive though.
The game shown in the keynote was a new, unreleased version of Devil May Cry (DMC 5). That makes it hard to make comparisons. I could not find any other source that shows off DMC5 performance. I guess we will know more in a month, when the Vega VII is released and testers can compare it to the RTX 2080 in existing games.
Max out your character with the loot and they just need to raise the enemy health count so you can repeat the grind for new loot to kill stronger enemies.
Repeat forever, or until the community gets bored.
Slightly OT, but you can kill a game that way. Skyforge for instance. It is nicely made and even gets new content from time to time, but every two months there is a new invasion where the level cap is increased by 10, while the mobs gain proportionally in health. Then it is grinding time again just to keep your effective power level.
By now Skyforge is down in the Steam charts to around 170 average players and 300 peak players. I wonder if MyCom still make any profit from this. I stopped playing myself last August, because I was fed up with the same content going up in level again and again.
With the open source "shim" support might even last longer than for Windows. Because you can keep the old blob and only adapt the interface kernel-side. But obviously, you will be stuck with the functionality of the last blob that Nvidia released. Including whatever bugs it still had at that point.
Besides, with about 30% more "raw" computing power and twice the memory, the Vega VII might actually beat the RTX 2080 long-term. Short term I agree that the RTX 2080 will probably bring slightly higher FPS, but AMD has a history of gaining a few percent over the years. What has been more critical in the past was memory requirements. Several years in the future, I guess the RTX 2080 will have real problems with its RAM size, while the Vega VII is still happily chugging along.
like the "copy this to your wall or facebook will publish all your private data due to privacy change!" thing has been making rounds _again_ just this week.
If you use Facebook at all, this is a significant risk anyway. SCNR
Tesla owner walks up to Bob's big old Ford 4x4 and tells Bob "I bet I can tow your Ford over 50 feet even if you have the brakes applied". Bob might agree, being sure of winning. Now we have a contest where nobody gets shot. We might even get a YouTube video out of it:)
There were some more recent attempts too. https://en.wikipedia.org/wiki/Steam_car. I vaguely remember an article about the Saab steam car. IIRC, they had a prototype with decent performance and running quite silent. Fuel consumption was a bit higher than in a conventional ICE though.
Well, many of the people driving the car would have loved to buy it from General Motors. But GM preferred to have them scrapped at the end of the leasing period. So this one seems to be a case of politically wanted (by GM itself) failure.
Now we go into round two with much more capable cars by multiple manufacturers. The Chevrolet Bolt and the Hyundai Kona Electric come to mind. And of course, the Tesla Model 3. Not exactly the cheapest cars, but much more serious competition this time than the GM EV.
If China wants to keep their research private, they can always not publish in traditional ways at all, but keep access to their research limited to Chinese researchers. In that case, maybe the Chinese government would pay for the infrastructure. And publishers like Elsevier would get exactly zero money out of it. But that is somewhat off topic. The question here is if we (in the western world) let commercial publishers control access to the results of research. Or if we make sure it gets released in some Open Access model. I'm in favor of the latter.
In this particular case, using a "terminator gene" might indeed be a good thing. If plants with that boosted photosynthesis get in the wild, it might give them enough of an evolutionary advantage that they spread across the world and upset the ecosystems.
Of course, someone will probably fuck it up and let some of these escape...
Actually, I believe that one. Some people really cling to their Windows XP, but if you can choose between Windows 8 and Windows 8.1, I see no reason not to prefer 8.1.
Note that I don't include myself in any of these. For me it is Windows 7 as long as it's supported, then probably Xubuntu.
Well it used to be a point in favor of Slashdot that they would not censor stuff. I happily accepted conspiracy theories and GNAA posts as price for that.
But now, perhaps I should stop reading and posting on Slashdot. The lack of crazy conspiracy theories suggests that the allegations of censorship are true:(
At least in Germany, the sum insured is much too small. Something like 2.5 billion Euros mandatory sum insured, and anything beyond that is up to the operator.
If we consider Chernobyl and Fukushima, an appropriate sum insured would be at least 200 billion. The last estimate I read about the total damage from Fukushima was around 190 billion, and I think insurance should cover the sums that "typically" happen in a worst-case scenario. We had two of those so far, so they are not impossible.
Now a group of major insurance companies might be able to collectively pay that sort of payout and offer insurance. But the premiums would probably make nuclear power quite unattractive.
Slightly OT: They had some help from Microsoft there, Windows used to be better but has gone downhill in the last years.
Say what you will about pre-Win10 OSes from MS, they promised their 10 years (security) support from release and mostly kept that promise. I'm typing this on an older Win7 machine, and it still gets security updates. The biggest (and intentional) flaw is that new processor generations don't get these updates at all for Win7. Also, there are relatively few compatibility problems with the updates in Win7.
I strongly suspect that making IE tied into the OS was for political reasons. More exactly, in the 2001 antitrust lawsuit Microsoft claimed that IE was inextricably linked to the OS. Someone (sorry, can't remember who) debunked that and showed a version of Win9x that ran just fine without IE.
I guess that Microsoft decided to make IE and Windows really entangled after that, so they would be not be caught with the same lie again. But design-wise, that is a bad idea. More modular and less entangled code is easier to maintain and tends to have less bugs. I guess the current problems are a late consequence of the politics-driven design decisions of 20 years ago.
I think it was more about the price point. Netbooks were aimed at a market where a "regular" Windows Home license would have hurt competitiveness quite a bit due to its price. So the makers of early netbooks went "hey, lets use Linux to cut the cost". Cue Microsoft making an even more crippled version of Windows XP and selling it cheap enough for the netbook market. User inertia did the rest.
Microsoft reacted to the NetBook fad by releasing a dirt cheap version of XP, limited to screen resolutions and memory sizes typically found in a netbook. Microsoft might not have earned much with that, but they successfully beat back the threat of a major Linux invasion in that market.
Yep. It was Windows RT. And the reason it failed is because it was confused with Windows. It couldn't run normal 32 bit Windows apps.
I think the reason was not just confusion: If 32 bit Windows apps don't run, then Microsoft is throwing away its probably greatest advantage, the lots of compatible software in the wild. Without that, they will at best be able to match the existing app stores of Google and Apple. And latecomers without some serious advantage don't get far.
Can't be licensing management alone, because it does not get easier than (open source) LibreOffice: Install as many as you want, never worry about having enough licenses. There are some governments that seem to be bribed by Microsoft (The state of Nordrhein-Westfalen in Germany is now dumping a perfectly working Linux environment for Windows).
Back on- topic too, that MS is using the customer channel of Win10 to QA the enterprise channel should be obvious from the update policies alone. Users of Win10 Home cannot postpone updates at all, so they do involuntary QA for the pro edition. Same for Home and Pro for Enterprise.
Arguably the bad guy continuing at all after the fake proxy request was bad OPSEC by itself. Detecting a trap should tell you that there is probably some police agency out to catch you.
Slashdot Mirror
Slightly OT:
If power efficiency is important to you, then a Vega 56 might be a better choice than the RX590. Slightly lower TBP and slightly better performance at the same time. It is more expensive though.
The game shown in the keynote was a new, unreleased version of Devil May Cry (DMC 5). That makes it hard to make comparisons. I could not find any other source that shows off DMC5 performance. I guess we will know more in a month, when the Vega VII is released and testers can compare it to the RTX 2080 in existing games.
Max out your character with the loot and they just need to raise the enemy health count so you can repeat the grind for new loot to kill stronger enemies.
Repeat forever, or until the community gets bored.
Slightly OT, but you can kill a game that way. Skyforge for instance. It is nicely made and even gets new content from time to time, but every two months there is a new invasion where the level cap is increased by 10, while the mobs gain proportionally in health. Then it is grinding time again just to keep your effective power level.
By now Skyforge is down in the Steam charts to around 170 average players and 300 peak players. I wonder if MyCom still make any profit from this. I stopped playing myself last August, because I was fed up with the same content going up in level again and again.
With the open source "shim" support might even last longer than for Windows. Because you can keep the old blob and only adapt the interface kernel-side. But obviously, you will be stuck with the functionality of the last blob that Nvidia released. Including whatever bugs it still had at that point.
Besides, with about 30% more "raw" computing power and twice the memory, the Vega VII might actually beat the RTX 2080 long-term. Short term I agree that the RTX 2080 will probably bring slightly higher FPS, but AMD has a history of gaining a few percent over the years.
What has been more critical in the past was memory requirements. Several years in the future, I guess the RTX 2080 will have real problems with its RAM size, while the Vega VII is still happily chugging along.
like the "copy this to your wall or facebook will publish all your private data due to privacy change!" thing has been making rounds _again_ just this week.
If you use Facebook at all, this is a significant risk anyway.
SCNR
Make it a bet, that's more polite anyway.
Tesla owner walks up to Bob's big old Ford 4x4 and tells Bob "I bet I can tow your Ford over 50 feet even if you have the brakes applied". Bob might agree, being sure of winning. Now we have a contest where nobody gets shot. We might even get a YouTube video out of it :)
There were some more recent attempts too. https://en.wikipedia.org/wiki/Steam_car.
I vaguely remember an article about the Saab steam car. IIRC, they had a prototype with decent performance and running quite silent. Fuel consumption was a bit higher than in a conventional ICE though.
Well, many of the people driving the car would have loved to buy it from General Motors. But GM preferred to have them scrapped at the end of the leasing period.
So this one seems to be a case of politically wanted (by GM itself) failure.
Now we go into round two with much more capable cars by multiple manufacturers. The Chevrolet Bolt and the Hyundai Kona Electric come to mind. And of course, the Tesla Model 3. Not exactly the cheapest cars, but much more serious competition this time than the GM EV.
If China wants to keep their research private, they can always not publish in traditional ways at all, but keep access to their research limited to Chinese researchers. In that case, maybe the Chinese government would pay for the infrastructure. And publishers like Elsevier would get exactly zero money out of it.
But that is somewhat off topic. The question here is if we (in the western world) let commercial publishers control access to the results of research. Or if we make sure it gets released in some Open Access model. I'm in favor of the latter.
In this particular case, using a "terminator gene" might indeed be a good thing. If plants with that boosted photosynthesis get in the wild, it might give them enough of an evolutionary advantage that they spread across the world and upset the ecosystems.
Of course, someone will probably fuck it up and let some of these escape...
Actually, I believe that one.
Some people really cling to their Windows XP, but if you can choose between Windows 8 and Windows 8.1, I see no reason not to prefer 8.1.
Note that I don't include myself in any of these. For me it is Windows 7 as long as it's supported, then probably Xubuntu.
Bullshit.
From ibm.com: https://www.ibm.com/it-infrastructure/z/os/linux-tested-platforms.
I don't think it gets any bigger than one of those mainframes. Except for some supercomputers in research, but many of those are running Linux too.
Well it used to be a point in favor of Slashdot that they would not censor stuff. I happily accepted conspiracy theories and GNAA posts as price for that.
But now, perhaps I should stop reading and posting on Slashdot. The lack of crazy conspiracy theories suggests that the allegations of censorship are true :(
In some very specific cases, they even appear cost-efficient now. Such as South Australia's Tesla battery: https://www.theguardian.com/technology/2018/sep/27/south-australias-tesla-battery-on-track-to-make-back-a-third-of-cost-in-a-year
I expect that cases like that become more common as battery technology improves.
At least in Germany, the sum insured is much too small. Something like 2.5 billion Euros mandatory sum insured, and anything beyond that is up to the operator.
If we consider Chernobyl and Fukushima, an appropriate sum insured would be at least 200 billion. The last estimate I read about the total damage from Fukushima was around 190 billion, and I think insurance should cover the sums that "typically" happen in a worst-case scenario. We had two of those so far, so they are not impossible.
Now a group of major insurance companies might be able to collectively pay that sort of payout and offer insurance. But the premiums would probably make nuclear power quite unattractive.
Oops, my logic was upside down here:
Windows 10 made it more difficult to support an OS _worse_ than MS
I mean how can you support an OS _worse_ than MS?
Slightly OT:
They had some help from Microsoft there, Windows used to be better but has gone downhill in the last years.
Say what you will about pre-Win10 OSes from MS, they promised their 10 years (security) support from release and mostly kept that promise. I'm typing this on an older Win7 machine, and it still gets security updates. The biggest (and intentional) flaw is that new processor generations don't get these updates at all for Win7. Also, there are relatively few compatibility problems with the updates in Win7.
I strongly suspect that making IE tied into the OS was for political reasons. More exactly, in the 2001 antitrust lawsuit Microsoft claimed that IE was inextricably linked to the OS. Someone (sorry, can't remember who) debunked that and showed a version of Win9x that ran just fine without IE.
I guess that Microsoft decided to make IE and Windows really entangled after that, so they would be not be caught with the same lie again. But design-wise, that is a bad idea. More modular and less entangled code is easier to maintain and tends to have less bugs. I guess the current problems are a late consequence of the politics-driven design decisions of 20 years ago.
I think it was more about the price point. Netbooks were aimed at a market where a "regular" Windows Home license would have hurt competitiveness quite a bit due to its price.
So the makers of early netbooks went "hey, lets use Linux to cut the cost". Cue Microsoft making an even more crippled version of Windows XP and selling it cheap enough for the netbook market. User inertia did the rest.
Microsoft reacted to the NetBook fad by releasing a dirt cheap version of XP, limited to screen resolutions and memory sizes typically found in a netbook.
Microsoft might not have earned much with that, but they successfully beat back the threat of a major Linux invasion in that market.
Yep. It was Windows RT. And the reason it failed is because it was confused with Windows. It couldn't run normal 32 bit Windows apps.
I think the reason was not just confusion:
If 32 bit Windows apps don't run, then Microsoft is throwing away its probably greatest advantage, the lots of compatible software in the wild.
Without that, they will at best be able to match the existing app stores of Google and Apple. And latecomers without some serious advantage don't get far.
You CAN compile .NET/C# for the "x86" target platform. Which means a Win32 program.
Can't be licensing management alone, because it does not get easier than (open source) LibreOffice:
Install as many as you want, never worry about having enough licenses.
There are some governments that seem to be bribed by Microsoft (The state of Nordrhein-Westfalen in Germany is now dumping a perfectly working Linux environment for Windows).
Back on- topic too, that MS is using the customer channel of Win10 to QA the enterprise channel should be obvious from the update policies alone. Users of Win10 Home cannot postpone updates at all, so they do involuntary QA for the pro edition. Same for Home and Pro for Enterprise.
Arguably the bad guy continuing at all after the fake proxy request was bad OPSEC by itself. Detecting a trap should tell you that there is probably some police agency out to catch you.
Something like that happened with Open Office.
https://en.wikipedia.org/wiki/LibreOffice#ooo-build,_Go-oo_and_Oracle
That was certainly not the only reason, as Open Office was also stagnating from a development POV. But I'm sure it contributed.