Re:Some traditional solutions to monitoring...
on
Employee Monitoring
·
· Score: 1
Interesting idea, but it wouldn't work. A machine can't work out what a human is doing (and thence whether or not he is "working") -- it doesn't have a brain.
To pick on the CPU example: what constitutes "work", exactly? Not a 100% CPU graph: that probably just means that the employee in question has his feet up and is waiting for a build job to complete. Or has sussed out the system, and is running a low priority CPU eating process in order to fool the SNMP thingy into thinking he's "working" all the time. Regular small spikes, indicating typing, clicking, saving, etc? Can be faked just as easily.
You could install a screen recorder -- the equivalent of CCTV for the desktop. That would require a lot more resources, and thence be more expensive. Fooling it would be harder, but still possible in various ways. (The how is left as an exercise for the reader. I conducted a security review of one once...)
In my opinion (as a security specialist, not as a lawyer, mind you), none of this evidence would be admissable in a court of law: if a company fired someone based solely on evidence provided by things like this, the employee in question could sue for unfair dismissal with a solid chance of winning. Therefore, it has negative value to the company -- it's useful only for generating pointless monitoring work, and for harassing employees and making them feel unhappy.
Re:A lot of the waste is a matter of opportunity
on
Employee Monitoring
·
· Score: 1
I have only an anecdote as a retort to this, it'll have to do. My point seems blindingly obvious, and I suspect my situation is not uncommon.
My productivity would *plummet* without Internet access at work. I spend a lot of time looking up information on the web -- mostly online documentation. If I didn't have that connection, I would be constantly purchasing paper documentation and industry journals, and so would all my peers -- at great overall cost to the company, without even considering the additional cost of the delay incurred every time I needed to wait for a new publication to arrive. (Simply put, product releases would not come out on time.)
In addition, no internet access would kill my ability to work remotely, because I need to be able to access my own systems at my desk remotely in order to be productive while out of the office.
I live in the UK. I used to get cable internet service from Virgin Media (the only cable provider in the country, because they bought up all the others). I would *love* to have had the quality of service that you guys above are complaining about from Comcast et al. Understand that Virgin Media works great until it breaks. Up to 50Mbps wherever you are, low latency, dropouts rare. When it breaks though, getting it fixed is a nightmare. And it *will* break. They don't keep track of what models of modems they've given people; they never send existing customers new hardware to replace obsolete models; they change the wire protocols without notice; they push broken firmware updates.
Tech support is outsourced to India. It's manned 24-7, but wait time is at least half an hour at all times. The "people" at the other end of that phone line are barely more sentient than M-x doctor. Diverge from their script, even the tiniest bit, and they'll tell you you're not supported and hang up on you. To get through their script, you must either lie to them or unplug every single piece of gear you have except for a Windows PC connected directly to your cable modem. You then spend half an hour having them tell you to unplug and re-plug all the connectors and reboot it five times. At the end of their script there's still a 50% chance that they'll tell you your PC must be broken and just hang up on you, rather than agree to do anything about it at all.
If you're lucky, you'll get sent an "engineer". He won't have a 4 hour window of arrival -- oh no, it's all day, any time between 8am and 6pm, and his best trick is arriving at 9am THE FOLLOWING MORNING. When he arrives, he's woefully underprepared, with only about a third of the equipment he ought to have (he will complain about this). He will fiddle with your modem, attach a meter contraption to the cable, and possibly change the little widget they fit inline with the cable to make up for the signal strength being too high. If you're unlucky and this does not work, he'll spend a few minutes using *your* phone to ring someone and explain to them that he doesn't understand what's going on, he'll noncommittally say "they'll look into it", and he'll leave. If you want to chase up (and thence have a hope that they'll sort things out), it's back on the phone to India, but the goon at the other end doesn't seem to understand the concept of records -- so you're back to square one!
Last year I was unlucky, and had a problem that was slightly non trivial. I counted. After three visits by these "engineers", SEVEN hours on the phone to India, one whole week waiting for second level support to ring back -- and they rang while I had something on the boil on the kitchen, I asked them to call me back in ten minutes, I never heard from them again -- they still had no idea what was wrong. After a month of no service despite constant chasing I rang the sales line, and cancelled, and told them precisely why. My call got escalated immediately, and the manager offered to send along one of the engineers who handle their much more expensive business service to take a look, but in a further two weeks' time; I cancelled my contract anyway, but accepted the engineer appointment since it was free.
Seven weeks after my connection had originally broken, and one week after I had DSL fitted -- slow, but with real support (www.aaisp.net.uk -- they're very good) -- the proper engineer arrived, picked up my cable modem, fiddled with it for a couple of minutes, and said "yeah, there's a return path fault on the modem. I can replace it if you'd like." I spent some time staring at him open mouthed before I managed to explain to him why I wouldn't like him to do that. I think he was pretty shocked at the quality of service I'd received.
However, when you install Ubuntu, if you find that there isn't a package for a particular program within the standard repositories, you can go and compile it from source and run it. Heck, you can even add third party repositories.
On iPhone, if there isn't a package for a particular program, you can't run it, unless you jailbreak the phone (at which point Apple starts actively trying to brick it for you.)
They're commodity displays. TN film is cheap and good enough for many people. If you want quality, there is a decent selection of IPS panel equipped monitors on the market.
There are good sites that compile information about what monitors use what technology, such as www.tftcentral.co.uk.
Anecdotally, the situation is not as bad as "4-6 weeks". I have a 5850. I pre-ordered it a couple of days before launch. I got it the following week.
According to the web forum of the retailer I shopped with (overclockers.co.uk) the stock has been trickling in in small shipments. If the shipments are never quite large enough to finish off the retailer's pre-order book, the item may never appear as "in stock" on the website (giving the impression there aren't any around at all), even though people who order are actually getting them reasonably promptly.
That *is* disappointing. As a Windows user who's never knowingly had a virus, and certainly never had my Windows machine's performance degraded by a virus, the first thing I'd look for in any AV solution before I install it is "will I notice a performance degradation"?
Big missed opportunity for Microsoft to show up the other vendors whose programs slow Windows down so notoriously.
I think I'll keep on taking the risk of having no AV...
Automated Windows Update is something you want on most systems anyway. The DRM section looks like it's just calling standard Windows DRM functionality as necessary, not adding more. What's wrong with Silverlight, aside from eating a little bit of (cheap) disk space?
Where did the "not authorized to talk about MSE" bit come from? I admit I've only skimmed the material but I couldn't see that.
You say "Core Duo" rather than "Core 2 Duo". There's a significant difference.
Amongst many other things the Core Duo has a weakness in SIMD floating-point performance that the Core 2 Duo fixes. I'd expect the latter chip to perform much more similarly to the Turion X2 in your benchmark.
DSL plans are always up-sold to the maximum possible speeds you could get under optimal line conditions.
Cable doesn't appear to be, though. I have NTL's top cable broadband service. It's rated 10Mbps downstream and I regularly see my download rates peg at the full 10Mbps. I don't know what the official upstream speed rating is but it appears to be about 600kbps (not great, but all services appear to be that asymmetric).
And I live in the outskirts of Cambridge, not somewhere that might be "special" like Westminster...;-)
I hope I'm remembering this correctly, I _think_ it's the "title" tag I'm thinking of (can't be bothered to go to the effort of checking...)
I used to develop a big scary web application. This web application used title tags to include descriptions of things which were sometimes quite long. After being localized into French they came out very long, and Firefox truncated them. The testers duly filed a bug report. It was assigned to me and I researched and discovered that apparently "title" is meant for short summary text only and should not include an essay; Firefox is behaving correctly. I removed the descriptions from the title tags and put them somewhere else; bug resolved.
Qwantz should use something else instead. I believe "alt" might be considered correct, except the browser doesn't have to display it unless you disable images...
As another poster said, it's a BBFC rating of 18 in the UK, which is stronger than an ELSPA rating. But it's not "functionally equivalent" to the "AO" rating in the US, because all the UK games stores happily stock 18-rated games. (Doom 3 got an 18 from the BBFC too and they all stocked that, for instance.) It's more similar to an "M" rating, except for the 1 year age difference.
The "put it in another room" technique is one I use too and is worth considering. A 10-metre DVI cable and a USB repeater cable are much cheaper than one of these cases:)
I'd note that in my experience, the Nvidia driver's RenderAccel option is OK on generations NV2x and later (GeForce3, GeForce4 Ti, GeForce FX, Quadro FX, etc) but dodgy and prone to causing crashes on NV1x (GeForce4 MX, some Quadro NVS such as the ones in all the recent cheapo Dell workstations my company has bought us, grr). In fact, I believe it's documented that running KDE 3.4 with an NV1x GPU with RenderAccel enabled will cause an instant X server crash. Check your GPUs...
Not strictly true. Whilst only Bill Gates can afford a *television* that can handle 1080p, there are several (comparatively) affordable LCD monitors on the market that can (the bigger Apple Cinema Displays, the BenQ FP231W, for instance -- I have one of the latter myself *g*)
Hopefully it will be possible to use an HDMI-to-DVI adapter to plug the PS3 into one of those and get the full 1080p resolution:)
Re:Massive processor, not much for graphics though
on
The Xbox 360 Unveiled
·
· Score: 1
Aha! Another person to pour cold water onto!
I very much doubt the 48 figure refers to pipelines as seen in current PC GPUs -- if it did, the GPU would have something like 700 million transistors, that's just improbable!
I believe ATI's new R5x0 architecture uses "unified shaders" or something (re-uses the same units for pixel and vertex operations). That 48 figure is probably a shader unit count or something. It's better than the current generation, but unlikely to make it 3x faster: perhaps 2x faster, at the outside.
The Nvidia G70 (to be announced Real Soon Now(TM)) is rumoured to have 2x the performance of the current 6800 series, as well...
New game engines (e.g. Unreal Engine 3) are already being built with multi core chips for PCs in mind, because the next gen consoles need this and they like to share the codebase.:)
To mention Unreal Engine 3 again, this engine uses the Novodex physics API which is to be hardware accelerated on PC by the AGEIA PhysX chip. Which looks like being a lot more powerful than the Xbox360's CPU for that particular task. (I've lost the link...)
Expect games released for Xbox360 *and* PC to perform damn well on a multi-core PC:)
Indeed. Shameless plug time, but I was under the impression that The Elder Scrolls IV: Oblivion (http://www.elderscrolls.com/) was going to be one of the launch titles for this thing. I'm most disappointed there appears to have been no mention of it, it looks much better than the stuff they did come up with...
Slashdot Mirror
Interesting idea, but it wouldn't work. A machine can't work out what a human is doing (and thence whether or not he is "working") -- it doesn't have a brain.
To pick on the CPU example: what constitutes "work", exactly? Not a 100% CPU graph: that probably just means that the employee in question has his feet up and is waiting for a build job to complete. Or has sussed out the system, and is running a low priority CPU eating process in order to fool the SNMP thingy into thinking he's "working" all the time. Regular small spikes, indicating typing, clicking, saving, etc? Can be faked just as easily.
You could install a screen recorder -- the equivalent of CCTV for the desktop. That would require a lot more resources, and thence be more expensive. Fooling it would be harder, but still possible in various ways. (The how is left as an exercise for the reader. I conducted a security review of one once...)
In my opinion (as a security specialist, not as a lawyer, mind you), none of this evidence would be admissable in a court of law: if a company fired someone based solely on evidence provided by things like this, the employee in question could sue for unfair dismissal with a solid chance of winning. Therefore, it has negative value to the company -- it's useful only for generating pointless monitoring work, and for harassing employees and making them feel unhappy.
I have only an anecdote as a retort to this, it'll have to do. My point seems blindingly obvious, and I suspect my situation is not uncommon.
My productivity would *plummet* without Internet access at work. I spend a lot of time looking up information on the web -- mostly online documentation. If I didn't have that connection, I would be constantly purchasing paper documentation and industry journals, and so would all my peers -- at great overall cost to the company, without even considering the additional cost of the delay incurred every time I needed to wait for a new publication to arrive. (Simply put, product releases would not come out on time.)
In addition, no internet access would kill my ability to work remotely, because I need to be able to access my own systems at my desk remotely in order to be productive while out of the office.
(Yes, I am a software developer.)
Rant mode on, but it's on topic.
I live in the UK. I used to get cable internet service from Virgin Media (the only cable provider in the country, because they bought up all the others). I would *love* to have had the quality of service that you guys above are complaining about from Comcast et al.
Understand that Virgin Media works great until it breaks. Up to 50Mbps wherever you are, low latency, dropouts rare. When it breaks though, getting it fixed is a nightmare. And it *will* break. They don't keep track of what models of modems they've given people; they never send existing customers new hardware to replace obsolete models; they change the wire protocols without notice; they push broken firmware updates.
Tech support is outsourced to India. It's manned 24-7, but wait time is at least half an hour at all times. The "people" at the other end of that phone line are barely more sentient than M-x doctor. Diverge from their script, even the tiniest bit, and they'll tell you you're not supported and hang up on you. To get through their script, you must either lie to them or unplug every single piece of gear you have except for a Windows PC connected directly to your cable modem. You then spend half an hour having them tell you to unplug and re-plug all the connectors and reboot it five times. At the end of their script there's still a 50% chance that they'll tell you your PC must be broken and just hang up on you, rather than agree to do anything about it at all.
If you're lucky, you'll get sent an "engineer". He won't have a 4 hour window of arrival -- oh no, it's all day, any time between 8am and 6pm, and his best trick is arriving at 9am THE FOLLOWING MORNING. When he arrives, he's woefully underprepared, with only about a third of the equipment he ought to have (he will complain about this). He will fiddle with your modem, attach a meter contraption to the cable, and possibly change the little widget they fit inline with the cable to make up for the signal strength being too high. If you're unlucky and this does not work, he'll spend a few minutes using *your* phone to ring someone and explain to them that he doesn't understand what's going on, he'll noncommittally say "they'll look into it", and he'll leave. If you want to chase up (and thence have a hope that they'll sort things out), it's back on the phone to India, but the goon at the other end doesn't seem to understand the concept of records -- so you're back to square one!
Last year I was unlucky, and had a problem that was slightly non trivial. I counted. After three visits by these "engineers", SEVEN hours on the phone to India, one whole week waiting for second level support to ring back -- and they rang while I had something on the boil on the kitchen, I asked them to call me back in ten minutes, I never heard from them again -- they still had no idea what was wrong. After a month of no service despite constant chasing I rang the sales line, and cancelled, and told them precisely why. My call got escalated immediately, and the manager offered to send along one of the engineers who handle their much more expensive business service to take a look, but in a further two weeks' time; I cancelled my contract anyway, but accepted the engineer appointment since it was free.
Seven weeks after my connection had originally broken, and one week after I had DSL fitted -- slow, but with real support (www.aaisp.net.uk -- they're very good) -- the proper engineer arrived, picked up my cable modem, fiddled with it for a couple of minutes, and said "yeah, there's a return path fault on the modem. I can replace it if you'd like." I spent some time staring at him open mouthed before I managed to explain to him why I wouldn't like him to do that. I think he was pretty shocked at the quality of service I'd received.
Never, ever, ever use Virgin Media.
However, when you install Ubuntu, if you find that there isn't a package for a particular program within the standard repositories, you can go and compile it from source and run it. Heck, you can even add third party repositories.
On iPhone, if there isn't a package for a particular program, you can't run it, unless you jailbreak the phone (at which point Apple starts actively trying to brick it for you.)
They're commodity displays. TN film is cheap and good enough for many people. If you want quality, there is a decent selection of IPS panel equipped monitors on the market.
There are good sites that compile information about what monitors use what technology, such as www.tftcentral.co.uk.
I agree about the laptops though. :(
"Never attribute to malice that which can be adequately explained by stupidity."
As a security guy, I like to counter that by paraphrasing Arthur C. Clarke:
Sufficiently advanced stupidity is indistinguishable from malice.
This is precisely why I stopped building ships in bottles. At the end of the day, all you've done is move some atoms around on a planet somewhere.
If the availability situation now is worse than it was at launch, that's distressing.
There may be a technical reason for the supply problems. Take with a pinch of salt, since this is Charlie Demerjian:
http://www.semiaccurate.com/2009/11/16/ati-58xx-parts-delayed-bit-more/
I suspect that if you're in the market for a high end video card now, you're still better off ordering a 58xx than waiting for Fermi.
Anecdotally, the situation is not as bad as "4-6 weeks". I have a 5850. I pre-ordered it a couple of days before launch. I got it the following week.
According to the web forum of the retailer I shopped with (overclockers.co.uk) the stock has been trickling in in small shipments. If the shipments are never quite large enough to finish off the retailer's pre-order book, the item may never appear as "in stock" on the website (giving the impression there aren't any around at all), even though people who order are actually getting them reasonably promptly.
That *is* disappointing. As a Windows user who's never knowingly had a virus, and certainly never had my Windows machine's performance degraded by a virus, the first thing I'd look for in any AV solution before I install it is "will I notice a performance degradation"?
Big missed opportunity for Microsoft to show up the other vendors whose programs slow Windows down so notoriously.
I think I'll keep on taking the risk of having no AV...
The license agreement, for reference:
http://www.microsoft.com/security_essentials/eula.aspx#mainNav
Automated Windows Update is something you want on most systems anyway. The DRM section looks like it's just calling standard Windows DRM functionality as necessary, not adding more. What's wrong with Silverlight, aside from eating a little bit of (cheap) disk space?
Where did the "not authorized to talk about MSE" bit come from? I admit I've only skimmed the material but I couldn't see that.
That's the funniest thing I've read for ages. 20 minutes later and I'm still in hysterics. AC, I salute you.
A TGV derailed at high speed on the LGV Nord line in 1993.
;)
http://en.wikipedia.org/wiki/LGV_Nord#History
I was waiting for another train at the Gare du Nord station in Paris at the time. I remember lots of cancellations and a long delay.
Use Kubuntu, it comes in a nice colour, no brown in sight :)
You say "Core Duo" rather than "Core 2 Duo". There's a significant difference.
Amongst many other things the Core Duo has a weakness in SIMD floating-point performance that the Core 2 Duo fixes. I'd expect the latter chip to perform much more similarly to the Turion X2 in your benchmark.
DSL plans are always up-sold to the maximum possible speeds you could get under optimal line conditions.
;-)
Cable doesn't appear to be, though. I have NTL's top cable broadband service. It's rated 10Mbps downstream and I regularly see my download rates peg at the full 10Mbps.
I don't know what the official upstream speed rating is but it appears to be about 600kbps (not great, but all services appear to be that asymmetric).
And I live in the outskirts of Cambridge, not somewhere that might be "special" like Westminster...
I hope I'm remembering this correctly, I _think_ it's the "title" tag I'm thinking of (can't be bothered to go to the effort of checking...)
I used to develop a big scary web application. This web application used title tags to include descriptions of things which were sometimes quite long. After being localized into French they came out very long, and Firefox truncated them. The testers duly filed a bug report. It was assigned to me and I researched and discovered that apparently "title" is meant for short summary text only and should not include an essay; Firefox is behaving correctly. I removed the descriptions from the title tags and put them somewhere else; bug resolved.
Qwantz should use something else instead. I believe "alt" might be considered correct, except the browser doesn't have to display it unless you disable images...
As another poster said, it's a BBFC rating of 18 in the UK, which is stronger than an ELSPA rating. But it's not "functionally equivalent" to the "AO" rating in the US, because all the UK games stores happily stock 18-rated games. (Doom 3 got an 18 from the BBFC too and they all stocked that, for instance.) It's more similar to an "M" rating, except for the 1 year age difference.
The "put it in another room" technique is one I use too and is worth considering. A 10-metre DVI cable and a USB repeater cable are much cheaper than one of these cases :)
Why was this modded off topic? Hellooooo, moderators. The topic is about IBM PPC chips, not Apple! The poster has a good point...
I'd note that in my experience, the Nvidia driver's RenderAccel option is OK on generations NV2x and later (GeForce3, GeForce4 Ti, GeForce FX, Quadro FX, etc) but dodgy and prone to causing crashes on NV1x (GeForce4 MX, some Quadro NVS such as the ones in all the recent cheapo Dell workstations my company has bought us, grr). In fact, I believe it's documented that running KDE 3.4 with an NV1x GPU with RenderAccel enabled will cause an instant X server crash. Check your GPUs...
Not strictly true. Whilst only Bill Gates can afford a *television* that can handle 1080p, there are several (comparatively) affordable LCD monitors on the market that can (the bigger Apple Cinema Displays, the BenQ FP231W, for instance -- I have one of the latter myself *g*)
:)
Hopefully it will be possible to use an HDMI-to-DVI adapter to plug the PS3 into one of those and get the full 1080p resolution
Aha! Another person to pour cold water onto!
I very much doubt the 48 figure refers to pipelines as seen in current PC GPUs -- if it did, the GPU would have something like 700 million transistors, that's just improbable!
I believe ATI's new R5x0 architecture uses "unified shaders" or something (re-uses the same units for pixel and vertex operations). That 48 figure is probably a shader unit count or something. It's better than the current generation, but unlikely to make it 3x faster: perhaps 2x faster, at the outside.
The Nvidia G70 (to be announced Real Soon Now(TM)) is rumoured to have 2x the performance of the current 6800 series, as well...
New game engines (e.g. Unreal Engine 3) are already being built with multi core chips for PCs in mind, because the next gen consoles need this and they like to share the codebase. :)
:)
To mention Unreal Engine 3 again, this engine uses the Novodex physics API which is to be hardware accelerated on PC by the AGEIA PhysX chip. Which looks like being a lot more powerful than the Xbox360's CPU for that particular task. (I've lost the link...)
Expect games released for Xbox360 *and* PC to perform damn well on a multi-core PC
Indeed. Shameless plug time, but I was under the impression that The Elder Scrolls IV: Oblivion (http://www.elderscrolls.com/) was going to be one of the launch titles for this thing. I'm most disappointed there appears to have been no mention of it, it looks much better than the stuff they did come up with...