Yes, if your goal is to maximize the line-count of your software.
LOC is a "perfectly good metric" for a manager who has no understanding of software engineering and needs to identify the coder(s) who do the bulk of the work. Is it flawed? Yes: proper code re-use libraries will completely ruin that measurement. But when your goal is to manage a group of people without knowing what they do, it's a great metric! (Now, I personally question why any manager would consider themselves competent while they don't understand the job of the people they manage, but I am led to understand a great many people do not consider this a problem at all.)
Problem is, some systems are inherently cranky, you've been handed crappy code and not enough time to rewrite it, or the faults are entirely out of your control.
And if you look at the title of this article ("The Perils of Ramming Products Down IT's Throat") and think of that not just in terms of purchasing, but also in the "no budget to upgrade" light, you might like the idea of a metric that will point out the "inherently cranky" systems. How often does management ignore IT when they say "it will cost you $50K to upgrade this but over the next three years return about $150K in productivity gains" but don't have 'real numbers'? This is a first step in creating those sorts of measurements, and that power shouldn't be ignored.
Yes, I am fully aware of the fact that LOC is a lousy, easily gamed metric. On the other hand, if you are trying to measure which of your programmers does the most work, LOC or check-ins per day isn't a bad start (just don't make it your final and only metric).
Likewise, the idea of using a combination of system down-time and administrative attention may be an easily gamed, faulty measurement of which systems perform "better" — but as a way to pinpoint which systems need overhauling and updating, it's a start.
The issue with IT is that nobody can really measure how well something new (or old) is doing.
I am not (currently) a IT admin, but with about 3 minutes of thinking I came up with what should be a metric at least as good as LOC is for programmers, if not better. Given that this is a first-iteration idea, I'm sure there is some nasty "bug" in it but I offer it up anyway.
Take a page from all of those number-of-nines reliability measurements and institute a measurement of "unmonitored performance" — the amount of time the server ran without needing administrative attention in the past week, month, and year, expressed as a percentage of that amount of time. All maintenance time — resulting in loss of service or not — is deducted from this measurement. For activities which result in loss of service, the entire length of loss of services is deducted against the measurement (even if it does not require active work during that period, like a reboot) — after all, if the server is not performing its' job, it is not unmonitored performance time.
The use of three time periods to measure in will help even out the "blip" which may occur with some unusual outage, helps to detail trends with particular servers and services, and pinpoint exact "bottleneck" services which consume a disparate amount of administrative hours.
There, now please tell me why this is not a fair measurement.
No business modely directly supports the creation of free items, digital or otherwise. But because you cannot directly make money off of an activity does not make it worthless (otherwise buying advertising would be a losing proposition).
There are several economic models that support "free" as an indirect means of doing business; here are three examples:
one can subsidize the creation of the product (such as many websites do with ad space)
"cross subsidize" the free item with sales of another (the King Gillette "sell blades, not razors" model)
use the "free" item as a loss leader to sell some other service (as many businesses do with promotional items)
In short: just because you are in the widget making business, doesn't mean you have to directly sell your widgets to the public. Any transaction that involves you getting money in return for giving people widgets (even if someone other than the final recipent pays) can be a profit making business model.
For that matter, as part of my "boycott", should I stop singing Happy Birthday?
Yes; as an alternative, may I suggest the "Birthday Dirge":
(Sung to the tune of "The Volga Boatmen")
Happy Birthday! <thud!> Happy Birthday! <thud!>
Birthdays come but once a year
Marking time as Death draws near
Happy Birthday! <thud!> Happy Birthday! <thud!>
(Alternate verses include:
It's your birthday never fear / You'll be dead this time next year
So you've aged another year / With each one your death draws near
Doom, destruction, and despair / People dying everywhere
Were I sitting in your shoes / I'd go out and sing the blues
etc. etc. etc.)
But the information in the book didn't "cost next to nothing to produce"
This is a fallacy when talking about pricing. Yes, the information in the book cost something to produce an initial "creative cost" — however, once you have created your original "proof" or "manuscript" copy, you don't pay that creative cost again when producing the next copy (or the next 10,000 identical copies) to that work. This is true whether the product is a book or digital file; however, physical books have additional costs (in paper and ink, as well as transportation) associated with their production. There you have to continue paying to create additional copies: so is it any wonder that piracy of the digital product is rampant when the barriers to reproduce the work illegally are so low (~$0)?
the cost of producing the information contained in either is still the same and what you are paying for is the information.
Agreed, but I (and IMO the grandparent poster) both just ask that the businesses selling these games recognize that selling a digital file is cheaper than selling a physical product, and as a result buying the digital file should be cheaper too. The other two big RPG players (Steve Jackson Games and White Wolf) both do exactly that, so why can't TSR/WoTC?
Is it wrong that my first response to seeing these achievements is "Good, now I have an excuse to avoid posting on Slashdot more often"?
I *like* the fact that my achievement score is so low. Here's to not raising it....
What are you talking about? It's not a theory, it's a definition. One specific definition out of several, I would imagine.
The substitution principle is a litte more than just "one definition out of several": in part, it is the definition for what inheritance means, in any object oriented programming language. Here's the Wikipedia version of the Liskov Substituion Principle: Let q(x) be a property provable about objects x of type T. Then q(y) should be true for objects y of type S where S is a subtype of T.
You may notice this doesn't say anyhing about behavior directly; any program behavior referenced is considerd "a property provable about objects x of type T". For those following along at home, let's reword this slightly to talk about classes: Let q(x) be a property provable about objects x of class T. Then q(y) should be true for objects y of type S where S is a subclass of T.
Now let's simplify this by taking out some of the formalism in the definition of terms and use a familiar, C++-style reference: Let T.q be a public function or variable for objects of class T. Then S.q should exist for objects of class S where S is a subclass of T.
Hey, this is looking familiar; in fact, that sounds like the typical definition of class inheritance to me! Sure, I may have cut out anything that references program behavior, but we'll get to that next.
Moreover, nothing in it says "OBEY ME OR YOU'RE BUGGED"; it's a very good guideline for sensible program design, but it has nothing to do with the completely independent definition of a bug in the general sense. Of course, you can always *make* the LSP part of your design criteria so that violations of it do in fact constitute bugs for your project, but that's a different matter.
If you limit the substitution principle to only referring to program behavior, then what you say is true — it isn't a limiting constraint on how you define classes and subclasses. I also must agree the principle does not define what a bug is in any manner at all; instead, it is concerned with formal proof there are no bugs. If that is all you care about, try this phrasing: Given a class T and its' subclass S, you cannot formally verify all programs referencing members of class S as a type of T if any public behavior of class S differs from class T.
So yes, I will recognize there are bug-free programs out there that may have classes which are not "proper subtypes" of each other; however, the principle only states that you cannot FORMALLY PROVE that such programs contain no bugs, which is quite a different question.
People always say this, and I never know what they mean.
No more journalists? Seriously? No more people who go find out news and then write about it?
No, that's not what the grandparent said. He said jouralism "outside the structure of traditional media" -- ie. journalists who do not work for a news station or paper directly.
Think of it as 10,000 Clark Kents and Lois Lanes, typing away on their blog (rather than at their cubicle in the Daily Planet), and the "headlines" being the articles selected by some sort of Google News style aggregator.
Just from reading this article, I feel like it's the lead in to a Monty Python sketch.
"And in this cage," (displays empty cage), "we can see the rare Caribbean poisonous shrew, which jumps out and injects its' prey with venom. Bites are instantly fatal, so we have to use extreme..."
[A brown blur crosses the screen and attacks the speaker's face]
"AAAAAAARRRRRRRGH!"
You have a constitutional right to infringe copyrights?
Or maybe you have a constitutional right to use the University's network in any way you wish?
No, but I do have a constitutional right to face my accuser in court... while this law appears to be little more than a way to accuse an entire campus of malfeasance at once.
But on the other hand, you don't tend to buy a $400 refurbished machine directly from Lenovo or Dell — there's no profit in it — you go to Ebay, Craigslist, freecycle....
On those sites, I'm sure you can very easily get a used Intel-based Macintosh for $400 or less.
What possible legal grounds could a political party - a private organization - have for forcing a state to do anything?
How about simply because the law says so.
America was founded on the principal that no man lives above the law; that every man lives under the same rules. If this is to be more than just hot air or pretty words, it actually has to be upheld by the courts. Nobody, not even the people in office and seats of power, should be exempt from the law -- otherwise our country is no different from any tin-pot dictatorship, where the laws are applied or ignored on a whim.
Please vote. A null vote (no candidate selected) says you're fine with the way things are run, and from your comment it's clear that you aren't. If you want to see change in the political arena, vote for one of the third party candidates -- Green, Constitutional, Libertarian, Independent -- just pick one.
Like the Adult Swim bump says, "We don't care who you vote for, just vote."
If you can't understand what a self signed certificate is, you shouldn't be accepting them.
If you can't rebuild a gasoline powered engine, you shouldn't be operating any vehicles that utilize them.
Apples and oranges, you say? No - not to 99% of the population.
We're technically minded people — the type who want to know how certificate signing operates — while 99% of the world wants it to "just work". While I would agree with your unspoken argument that trust is significant enough to warrant forcing everyone to know the mechanisms, for most people there is as much interest in understanding the "engineering" aspects of that as there is in understanding internal combustion engines.
If you have the type of job you need to be on-call for, you obviously have much more value to your employer than a burger-flipper, and the worker should be looking to enhance that, not to sue for after-hours work.
Yes, and I look at a pre-declared, known on-call rotation schedule as being an "enhancement" to the employment relationship. If you (by which I also mean "any employer") want me on-call, I require to know when I am on call and will want appropriate compensation. If you (as the employer) consider such things irrelevant and dismiss the request — even when required by law — then I would consider suing over the issue. THAT is what the article is describing.
I would be unwilling to work in a job with hazardous working conditions (at least not without reporting them to OSHA) — or are you trying to tell me that employment law designed to protect physical health is more vital than one to protect mental health?
You're not interacting with a person [online], you're interacting with your own imagination, seeded with a few select facts or fictions from someone else.
In that case, since I am not in the habit of arguing with myself, I see no need to rebut the obvious fallacies of your argument — or perhaps you meant something else by "not interacting with a person, you're interacting with your own imagination"?
Slashdot Mirror
I think it might be a reference to the first computer programmer, Countess Ada Lovelace.
LOC is a "perfectly good metric" for a manager who has no understanding of software engineering and needs to identify the coder(s) who do the bulk of the work. Is it flawed? Yes: proper code re-use libraries will completely ruin that measurement. But when your goal is to manage a group of people without knowing what they do, it's a great metric! (Now, I personally question why any manager would consider themselves competent while they don't understand the job of the people they manage, but I am led to understand a great many people do not consider this a problem at all.)
And if you look at the title of this article ("The Perils of Ramming Products Down IT's Throat") and think of that not just in terms of purchasing, but also in the "no budget to upgrade" light, you might like the idea of a metric that will point out the "inherently cranky" systems. How often does management ignore IT when they say "it will cost you $50K to upgrade this but over the next three years return about $150K in productivity gains" but don't have 'real numbers'? This is a first step in creating those sorts of measurements, and that power shouldn't be ignored.
Yes, I am fully aware of the fact that LOC is a lousy, easily gamed metric. On the other hand, if you are trying to measure which of your programmers does the most work, LOC or check-ins per day isn't a bad start (just don't make it your final and only metric).
Likewise, the idea of using a combination of system down-time and administrative attention may be an easily gamed, faulty measurement of which systems perform "better" — but as a way to pinpoint which systems need overhauling and updating, it's a start.
I am not (currently) a IT admin, but with about 3 minutes of thinking I came up with what should be a metric at least as good as LOC is for programmers, if not better. Given that this is a first-iteration idea, I'm sure there is some nasty "bug" in it but I offer it up anyway.
Take a page from all of those number-of-nines reliability measurements and institute a measurement of "unmonitored performance" — the amount of time the server ran without needing administrative attention in the past week, month, and year, expressed as a percentage of that amount of time. All maintenance time — resulting in loss of service or not — is deducted from this measurement. For activities which result in loss of service, the entire length of loss of services is deducted against the measurement (even if it does not require active work during that period, like a reboot) — after all, if the server is not performing its' job, it is not unmonitored performance time.
The use of three time periods to measure in will help even out the "blip" which may occur with some unusual outage, helps to detail trends with particular servers and services, and pinpoint exact "bottleneck" services which consume a disparate amount of administrative hours.
There, now please tell me why this is not a fair measurement.
In short: just because you are in the widget making business, doesn't mean you have to directly sell your widgets to the public. Any transaction that involves you getting money in return for giving people widgets (even if someone other than the final recipent pays) can be a profit making business model.
Yes; as an alternative, may I suggest the "Birthday Dirge":
(Sung to the tune of "The Volga Boatmen")
Happy Birthday! <thud!> Happy Birthday! <thud!>
Birthdays come but once a year
Marking time as Death draws near
Happy Birthday! <thud!> Happy Birthday! <thud!>
(Alternate verses include:
It's your birthday never fear / You'll be dead this time next year
So you've aged another year / With each one your death draws near
Doom, destruction, and despair / People dying everywhere
Were I sitting in your shoes / I'd go out and sing the blues
etc. etc. etc.)
This is a fallacy when talking about pricing. Yes, the information in the book cost something to produce an initial "creative cost" — however, once you have created your original "proof" or "manuscript" copy, you don't pay that creative cost again when producing the next copy (or the next 10,000 identical copies) to that work. This is true whether the product is a book or digital file; however, physical books have additional costs (in paper and ink, as well as transportation) associated with their production. There you have to continue paying to create additional copies: so is it any wonder that piracy of the digital product is rampant when the barriers to reproduce the work illegally are so low (~$0)?
Agreed, but I (and IMO the grandparent poster) both just ask that the businesses selling these games recognize that selling a digital file is cheaper than selling a physical product, and as a result buying the digital file should be cheaper too. The other two big RPG players (Steve Jackson Games and White Wolf) both do exactly that, so why can't TSR/WoTC?
Is it wrong that my first response to seeing these achievements is "Good, now I have an excuse to avoid posting on Slashdot more often"? I *like* the fact that my achievement score is so low. Here's to not raising it....
The substitution principle is a litte more than just "one definition out of several": in part, it is the definition for what inheritance means, in any object oriented programming language. Here's the Wikipedia version of the Liskov Substituion Principle:
Let q(x) be a property provable about objects x of type T. Then q(y) should be true for objects y of type S where S is a subtype of T.
You may notice this doesn't say anyhing about behavior directly; any program behavior referenced is considerd "a property provable about objects x of type T". For those following along at home, let's reword this slightly to talk about classes:
Let q(x) be a property provable about objects x of class T. Then q(y) should be true for objects y of type S where S is a subclass of T.
Now let's simplify this by taking out some of the formalism in the definition of terms and use a familiar, C++-style reference:
Let T.q be a public function or variable for objects of class T. Then S.q should exist for objects of class S where S is a subclass of T.
Hey, this is looking familiar; in fact, that sounds like the typical definition of class inheritance to me! Sure, I may have cut out anything that references program behavior, but we'll get to that next.
If you limit the substitution principle to only referring to program behavior, then what you say is true — it isn't a limiting constraint on how you define classes and subclasses. I also must agree the principle does not define what a bug is in any manner at all; instead, it is concerned with formal proof there are no bugs. If that is all you care about, try this phrasing:
Given a class T and its' subclass S, you cannot formally verify all programs referencing members of class S as a type of T if any public behavior of class S differs from class T.
So yes, I will recognize there are bug-free programs out there that may have classes which are not "proper subtypes" of each other; however, the principle only states that you cannot FORMALLY PROVE that such programs contain no bugs, which is quite a different question.
Proof, please; you are contesting an award-winning theory, and I for one side with prevailing theory until further evidence is provided.
Heads on spikes is sooo 20th century. This is the 31st century... we'll dessicate them and grind their corpses into a fine pink powder.
(Rattles Can)
Torgo's Excecutive Powder, anyone?
No, that's not what the grandparent said. He said jouralism "outside the structure of traditional media" -- ie. journalists who do not work for a news station or paper directly.
Think of it as 10,000 Clark Kents and Lois Lanes, typing away on their blog (rather than at their cubicle in the Daily Planet), and the "headlines" being the articles selected by some sort of Google News style aggregator.
I seem to recall from previous discussions that WOW players would be OK with this if they could PvP attack the taxman....
Just from reading this article, I feel like it's the lead in to a Monty Python sketch.
"And in this cage," (displays empty cage), "we can see the rare Caribbean poisonous shrew, which jumps out and injects its' prey with venom. Bites are instantly fatal, so we have to use extreme..."
[A brown blur crosses the screen and attacks the speaker's face] "AAAAAAARRRRRRRGH!"
In Soviet Russia, new smiley overlords profit??? from YOU! Oh, wait....
No, but I do have a constitutional right to face my accuser in court... while this law appears to be little more than a way to accuse an entire campus of malfeasance at once.
Not from Apple, no.
But on the other hand, you don't tend to buy a $400 refurbished machine directly from Lenovo or Dell — there's no profit in it — you go to Ebay, Craigslist, freecycle....
On those sites, I'm sure you can very easily get a used Intel-based Macintosh for $400 or less.
Okay, so why not tie it down to some floating platform deep in the Pacific?
Nothing said that "tied down" meant to the ground.
How about simply because the law says so.
America was founded on the principal that no man lives above the law; that every man lives under the same rules. If this is to be more than just hot air or pretty words, it actually has to be upheld by the courts. Nobody, not even the people in office and seats of power, should be exempt from the law -- otherwise our country is no different from any tin-pot dictatorship, where the laws are applied or ignored on a whim.
Please vote. A null vote (no candidate selected) says you're fine with the way things are run, and from your comment it's clear that you aren't. If you want to see change in the political arena, vote for one of the third party candidates -- Green, Constitutional, Libertarian, Independent -- just pick one.
Like the Adult Swim bump says, "We don't care who you vote for, just vote."
If you can't rebuild a gasoline powered engine, you shouldn't be operating any vehicles that utilize them.
Apples and oranges, you say? No - not to 99% of the population.
We're technically minded people — the type who want to know how certificate signing operates — while 99% of the world wants it to "just work". While I would agree with your unspoken argument that trust is significant enough to warrant forcing everyone to know the mechanisms, for most people there is as much interest in understanding the "engineering" aspects of that as there is in understanding internal combustion engines.
This should go before my second paragraph.
Yes, and I look at a pre-declared, known on-call rotation schedule as being an "enhancement" to the employment relationship. If you (by which I also mean "any employer") want me on-call, I require to know when I am on call and will want appropriate compensation. If you (as the employer) consider such things irrelevant and dismiss the request — even when required by law — then I would consider suing over the issue. THAT is what the article is describing.
I would be unwilling to work in a job with hazardous working conditions (at least not without reporting them to OSHA) — or are you trying to tell me that employment law designed to protect physical health is more vital than one to protect mental health?
In that case, since I am not in the habit of arguing with myself, I see no need to rebut the obvious fallacies of your argument — or perhaps you meant something else by "not interacting with a person, you're interacting with your own imagination"?