If they're running a helpdesk, those systems typically do not provide such granular data that the cost of supporting specific applications is reliably easy to do.
Actually most helpdesk systems, both commercial and OSS do, (or can if they're set up properly), provide very granular data which can be used to produce very pretty graphs that make PHBs feel all warm and fuzzy.
That said, as with any metrics, you have to understand the data you're trying to capture and as you so rightly pointed out the inability to perform this sort of thing is a problem that's endemic.
What they should have done is tracked a year's worth of MS-Office support issues to get a good baseline, then done the same for the OpenOffice implementation. OO would have a spike in the beginning of course but after the initial settling period it wouldn't be too hard to compare average number of issues and time to resolution.
Has anybody reading this here actually done this kind of migration and have numbers they can post? [enabling troll/flame filter now]
it was almost impossible to work out what open-source was actually costing
Sounds like there's a disconnect between the IT staff and the business side of the house. Any CIO worth their salt would have had before-and-after metrics to compare.
But the iPhone is from Apple, of course it would ask for a Mac address! Heck, they should be glad it didn't ask for a Mac-II address, things would be twice as bad! (You can do the math for a Mac-IIcx:)
but several phones can bring down the network? seems very vulnerable. Is there anything AP can do to just ignore the rogue requests?
It's probably related to Cisco's built in defense mechanisms. By default if a Cisco AP detects what it thinks is an attack it will go offline for awhile. The problem is that in the real world there are buggy chipsets and drivers that will trigger this so one usually ends up disabling them in self-defense. As a specific example there is an Intel WLAN chipset present in many older laptops that will randomly resend packets. An AP configured with default settings will shut off for exactly 60 seconds after it sees a couple of those as it thinks a replay attack is being used against it.
There are several different attack vectors detected and timers associated. But I would think a university would already know all about this and have them configured correctly but if not then yeah, a couple of rogue devices can bring the whole shootin' match down. (To be fair Cisco isn't the only AP vendor that this can happen to).
I'd be willing to bet that each offending iPhone may have been first connected to a home wireless router or gateway, and it may automatically and repeatedly be trying to reconnect to it again
Not unless somebody hacked it to run WinCE, that kind of bizarreness is strictly Windows-land...
Incorrect. That is more of something that an electrical engineer or someone doing low-level programming will have to worry about. All the binary fun with xors, ands, ors, shifts, etc., are all unrelated to the sort of math of Alan Turing.
Come back to the discussion once you understand what a Turing Machine is.
As someone who's been in this industry for over twenty years in a wide range of positions and responsibilities consulting on well over a thousand projects for hundreds of companies of all sizes and industries, and knowing over a dozen programming languages, I feel I have the background to not only comment somewhat knowledgeably, but to do so in language that can be understood by most people, not just CompSci majors. I would recommend you re-read my parent post with a little more open-minded attitude...:)
Anybody else notice how many people completely missed the bit about "walk into the office"?
Fart jokes aside there's quite a few posts about "waking up" activities, or maybe a lot of people just sleep at the office...?
is the same as writing litterature with a programming language.
The reason computer science is so heavily influenced by math is the binary architecture that every piece of hardware is designed around. Every real world problem, right down to choosing the color of a font, has to be translated into the digital world by algorithmic approximation - a lot of math! The problem is that it is this very abstraction that makes computers so "flexible" in what they can do. Analog computers existed many years ago but they could only ever be built for a single purpose.
Unfortunately(?) it is much easier to design and mass produce something which is based on a finite lowest common denominator (bits) than it is to do so based on the continuum that a non-digital solution would require.
That said, who's to say that a beautiful painting rendered in Gimp/PhotoShop isn't a program of sorts? Certainly it has input, (from the original creator), and output, (its effect on us), and the "code" can be modified to change both!
No kidding, heck most companies can't even turn AP/AR around in 22 days, and that's part of their core business! Something out of the ordinary routine like this is certainly going to take longer...
(Maybe KDawson misread it as 22 months?:)
Re:I'm buying.. Friday.
on
All Things iPhone
·
· Score: 3, Interesting
The paltry 200 txt msgs standard to each plan is also annoying (so now I'll have to drop $10 or $20 per phone for extra). But even still, I'm firmly in the demographic that is willing to pay extra for the phone, the service, just for the UI (and non-crashyness) that Apple will bring to the table.
As something of an Apple fanboi myself, I'm actually going to wait awhile, maybe a really long while, for the simple reason that the data rate with EDGE absolutely sucks and Cingular's signal coverage is pitiful. I love the UI, I've been drooling over this thing since I first heard about it, but I'm not about to drop $500 on a phone I can't actually use productively in the real world. It's almost painful to watch this unfold; I so wish Apple had gone with Sprint or Verizon or almost anybody except Cingular (well, okay, T-Mobile would have been worse). And with a 5-year exclusivity no less!?! Egads, wtf is the Jobster smoking?
Companies have websites with dead links that don't get caught for years too. People make mistakes under any/all circumstances, not just involving SPF records. Sounds like these guys weren't exactl on the ball though, any automated system should have some sort of reporting/monitoring with a live human on the other end.
Interesting to hear that AOL is using Barracudas now, I knew they were switching but didn't know who won the contract.
FYI - see my post above regarding Barracuda's SPF bouncebacks, that's been taken care of recently.
are responsible for some 40% of the shite in my mailbox
For what it's worth one of the updates recently disables bouncebacks triggered by SPF check. They heard that complaint from a lot of folks and finally got around to fixing it.
The problem is that things change, servers get moved, new websites get created, and if the SPF records aren't updated to reflect those changes... Yes, it's possible that all the other DNS records could get updated and just that one be ignored but it's pretty unlikely.
...then some emails are going to go missing. A problem like that could go undetected for months, and that bears a cost. Having had occasion to mistype an IP in an SPF record once or twice I can assure you that it does not go undetected for months. A few hours perhaps, (pft, more like minutes). There are enough ISP's and private spam filters using SPF now that if there's a glitch it shows up pretty damn quick in the form of an irate user, "I was able to email this person yesterday, ZOMGWTFBBQ?!?"
Seagate's purchase of Maxtor was purely for intellectual property reasons. Maxtor owned a couple of patents on technology for exceptionally high speed data transfer within the drive. DoveBid Auctions has had decommissioned Maxtor assembly line equipment auctions constantly for several months now - Seagate is decommissioning them all. (But in the next year or two look for Seagate drives to get even faster!:)
Therefore, if 2 threads...for an app that spends the vast majority of its time in user interaction, then the company will find it uneconomic
And again, that's a specific case, (or, rather, class of cases), which cannot be used as a broad-stroke brush in the big picture. The original submitter's question remains valid for a great deal of "back end" operations. To be fair though, your reference to the economics of such development being the sticking point still tends to hold true. (Though if a compiler/interpreter could be made smart enough those economic hurdles could quite easily vanish.)
With single-thread tasks on a computer with more than two cores and only one user, the user's attention span becomes the bottleneck.
While that is certainly quite be true for specific cases applying it as a general rule is ridiculous.
(And it's STILL irrelevant to the topic at hand...:)
apps are minimized, they are blocking, not running That would be a pretty poorly written app!:)
Blocking...blocking...blocking...blocking None of these should be truly be "blocking". Unless you're using the term to mean something else...?
But is the total CPU utilization among cores regularly greater than 100%? By definition, no.:) But I understand what you mean, and if I'm doing something compute intensive like a compile yes, one CPU will usually be totally maxed and the other will usually be at about 75%-100% depending on what all else I'm trying to do at the same time.
is parsing and laying out a single HTML document easily parallelized? Within a single application, no, of course not, that's what this whole discussion is about. But overall, across applications, yes indeed! My web browser can be loading pages on one monitor (I have multiple screens) while I'm typing away in the word processor on another, and if I look at the CPU affinity for these apps there's a 50-50 chance they will be on different cores. (Obviously if they are on the same core then it's just task switching and they're not really running in parallel.)
Now if you want to nit-pick and say they are still not truly in parallel because the video updates to the screen(s) come down a single bus and they have to take turns well then that's true, and is also true of a ton of other IO functions, but it's also not really related to the parallel processing question. (Be nice if there was a solution for that though!)
Slashdot Mirror
If they're running a helpdesk, those systems typically do not provide such granular data that the cost of supporting specific applications is reliably easy to do.
Actually most helpdesk systems, both commercial and OSS do, (or can if they're set up properly), provide very granular data which can be used to produce very pretty graphs that make PHBs feel all warm and fuzzy.
That said, as with any metrics, you have to understand the data you're trying to capture and as you so rightly pointed out the inability to perform this sort of thing is a problem that's endemic.
What they should have done is tracked a year's worth of MS-Office support issues to get a good baseline, then done the same for the OpenOffice implementation. OO would have a spike in the beginning of course but after the initial settling period it wouldn't be too hard to compare average number of issues and time to resolution.
Has anybody reading this here actually done this kind of migration and have numbers they can post? [enabling troll/flame filter now]
Obviously they're not fully backwards-compatible, hardware wise, or they'd also be asking for Lisa addresses..
:)
They probably know Lisa's married now and not giving out her address any more.
In the end we too settled on Microsoft. It's just the logical choice.
:)
Says the AC...
it was almost impossible to work out what open-source was actually costing
Sounds like there's a disconnect between the IT staff and the business side of the house. Any CIO worth their salt would have had before-and-after metrics to compare.
No wireless hardware requests a MAC address.
:)
But the iPhone is from Apple, of course it would ask for a Mac address! Heck, they should be glad it didn't ask for a Mac-II address, things would be twice as bad!
(You can do the math for a Mac-IIcx
but several phones can bring down the network? seems very vulnerable. Is there anything AP can do to just ignore the rogue requests?
It's probably related to Cisco's built in defense mechanisms. By default if a Cisco AP detects what it thinks is an attack it will go offline for awhile. The problem is that in the real world there are buggy chipsets and drivers that will trigger this so one usually ends up disabling them in self-defense. As a specific example there is an Intel WLAN chipset present in many older laptops that will randomly resend packets. An AP configured with default settings will shut off for exactly 60 seconds after it sees a couple of those as it thinks a replay attack is being used against it.
There are several different attack vectors detected and timers associated. But I would think a university would already know all about this and have them configured correctly but if not then yeah, a couple of rogue devices can bring the whole shootin' match down. (To be fair Cisco isn't the only AP vendor that this can happen to).
I'd be willing to bet that each offending iPhone may have been first connected to a home wireless router or gateway, and it may automatically and repeatedly be trying to reconnect to it again
Not unless somebody hacked it to run WinCE, that kind of bizarreness is strictly Windows-land...
Heh, not to reply to myself, but to reply to myself, this is the quote at the bottom of the page right now:
"A word to the wise: a credentials dicksize war is usually a bad idea on the net." (David Parsons in c.o.l.development.system, about coding in C.)
Rather apt I think!
Incorrect. That is more of something that an electrical engineer or someone doing low-level programming will have to worry about. All the binary fun with xors, ands, ors, shifts, etc., are all unrelated to the sort of math of Alan Turing.
:)
Come back to the discussion once you understand what a Turing Machine is.
As someone who's been in this industry for over twenty years in a wide range of positions and responsibilities consulting on well over a thousand projects for hundreds of companies of all sizes and industries, and knowing over a dozen programming languages, I feel I have the background to not only comment somewhat knowledgeably, but to do so in language that can be understood by most people, not just CompSci majors. I would recommend you re-read my parent post with a little more open-minded attitude...
the first thing I do in the morning
Anybody else notice how many people completely missed the bit about "walk into the office"?
Fart jokes aside there's quite a few posts about "waking up" activities, or maybe a lot of people just sleep at the office...?
is the same as writing litterature with a programming language.
The reason computer science is so heavily influenced by math is the binary architecture that every piece of hardware is designed around. Every real world problem, right down to choosing the color of a font, has to be translated into the digital world by algorithmic approximation - a lot of math! The problem is that it is this very abstraction that makes computers so "flexible" in what they can do. Analog computers existed many years ago but they could only ever be built for a single purpose.
Unfortunately(?) it is much easier to design and mass produce something which is based on a finite lowest common denominator (bits) than it is to do so based on the continuum that a non-digital solution would require.
That said, who's to say that a beautiful painting rendered in Gimp/PhotoShop isn't a program of sorts? Certainly it has input, (from the original creator), and output, (its effect on us), and the "code" can be modified to change both!
Do you have any particular reason for believing increased mass (in a car with the same cross-sectional area) reduces fuel economy?
Not much for highway but probably a measurable difference for city driving - it takes more fuel to accelerate an increased mass.
This is craziness...in 22 days
:)
No kidding, heck most companies can't even turn AP/AR around in 22 days, and that's part of their core business! Something out of the ordinary routine like this is certainly going to take longer...
(Maybe KDawson misread it as 22 months?
The paltry 200 txt msgs standard to each plan is also annoying (so now I'll have to drop $10 or $20 per phone for extra). But even still, I'm firmly in the demographic that is willing to pay extra for the phone, the service, just for the UI (and non-crashyness) that Apple will bring to the table.
As something of an Apple fanboi myself, I'm actually going to wait awhile, maybe a really long while, for the simple reason that the data rate with EDGE absolutely sucks and Cingular's signal coverage is pitiful. I love the UI, I've been drooling over this thing since I first heard about it, but I'm not about to drop $500 on a phone I can't actually use productively in the real world. It's almost painful to watch this unfold; I so wish Apple had gone with Sprint or Verizon or almost anybody except Cingular (well, okay, T-Mobile would have been worse). And with a 5-year exclusivity no less!?! Egads, wtf is the Jobster smoking?
I reject such bounce-backs at SMTP time.
:)
So how do you know when it's a legitimate bounce-back? Or do you just reject all bounce-backs out of hand? (Sure hope that's a home network!
Companies have websites with dead links that don't get caught for years too. People make mistakes under any/all circumstances, not just involving SPF records. Sounds like these guys weren't exactl on the ball though, any automated system should have some sort of reporting/monitoring with a live human on the other end.
Interesting to hear that AOL is using Barracudas now, I knew they were switching but didn't know who won the contract.
FYI - see my post above regarding Barracuda's SPF bouncebacks, that's been taken care of recently.
are responsible for some 40% of the shite in my mailbox
For what it's worth one of the updates recently disables bouncebacks triggered by SPF check. They heard that complaint from a lot of folks and finally got around to fixing it.
The problem is that things change, servers get moved, new websites get created, and if the SPF records aren't updated to reflect those changes...
...then some emails are going to go missing. A problem like that could go undetected for months, and that bears a cost.
Yes, it's possible that all the other DNS records could get updated and just that one be ignored but it's pretty unlikely.
Having had occasion to mistype an IP in an SPF record once or twice I can assure you that it does not go undetected for months. A few hours perhaps, (pft, more like minutes). There are enough ISP's and private spam filters using SPF now that if there's a glitch it shows up pretty damn quick in the form of an irate user, "I was able to email this person yesterday, ZOMGWTFBBQ?!?"
Been done before...
And the article linked to in the parent's comment is much more detailed and informative! This is what should have been submitted instead!
Maxtor (Seagate?) drives are junk.
:)
Seagate's purchase of Maxtor was purely for intellectual property reasons. Maxtor owned a couple of patents on technology for exceptionally high speed data transfer within the drive. DoveBid Auctions has had decommissioned Maxtor assembly line equipment auctions constantly for several months now - Seagate is decommissioning them all. (But in the next year or two look for Seagate drives to get even faster!
Therefore, if 2 threads...for an app that spends the vast majority of its time in user interaction, then the company will find it uneconomic
And again, that's a specific case, (or, rather, class of cases), which cannot be used as a broad-stroke brush in the big picture. The original submitter's question remains valid for a great deal of "back end" operations. To be fair though, your reference to the economics of such development being the sticking point still tends to hold true. (Though if a compiler/interpreter could be made smart enough those economic hurdles could quite easily vanish.)
With single-thread tasks on a computer with more than two cores and only one user, the user's attention span becomes the bottleneck.
:)
While that is certainly quite be true for specific cases applying it as a general rule is ridiculous.
(And it's STILL irrelevant to the topic at hand...
human attention span...has nothing to do with parallel programming. How the heck did we get on that topic?!? :)
apps are minimized, they are blocking, not running :)
:) But I understand what you mean, and if I'm doing something compute intensive like a compile yes, one CPU will usually be totally maxed and the other will usually be at about 75%-100% depending on what all else I'm trying to do at the same time.
That would be a pretty poorly written app!
Blocking...blocking...blocking...blocking
None of these should be truly be "blocking". Unless you're using the term to mean something else...?
But is the total CPU utilization among cores regularly greater than 100%?
By definition, no.
is parsing and laying out a single HTML document easily parallelized?
Within a single application, no, of course not, that's what this whole discussion is about. But overall, across applications, yes indeed! My web browser can be loading pages on one monitor (I have multiple screens) while I'm typing away in the word processor on another, and if I look at the CPU affinity for these apps there's a 50-50 chance they will be on different cores. (Obviously if they are on the same core then it's just task switching and they're not really running in parallel.)
Now if you want to nit-pick and say they are still not truly in parallel because the video updates to the screen(s) come down a single bus and they have to take turns well then that's true, and is also true of a ton of other IO functions, but it's also not really related to the parallel processing question. (Be nice if there was a solution for that though!)