My favourite was one I saw last year, a requirement for 5 years Java/J2EE (okay)and 5 years C#/dot NET (eh!). Apart from being difficult to have 5 years experience in something that came out in 2002, I'm not sure that I would want to work for a company with this bad a grasp of skills management.
I think you're right about the market, and about how people only need a few weeks to get up to speed on new stuff (it's not brain surgery is it!) but the crunch is always with the contractors. Trying to stay ahead of the game is tough as you end up in a catch 22 where people will only hire you if you already have experience in something
The fact that you've found good people is probably more a reflection of your abililty as a manager. Your time "at the coalface" gives you an insight into how to hire someone that might not have the skills now, but would be fantastic with a little training. With too many managers, that's a risk they can't take as they just can't see potential, so they fall back on proven knowledge - experience.
One of the things that always troubles me with the Outsourcing debate is how it regards IT and software development as an entity in itself, rather than one that must deal with others. By this I mean both dealing with the business you are in and also the other departments in your company. By making IT a commodity, it can be offshored or outsourced easily. When it's a specialty, that becomes difficult to impossible.
If you are developing a piece of medical software such as an EEG recorder, you need to have some understanding of the science of EEGs and the medical background in which they are used. Likewise, a piece of financial software requires detailed knowledge of financial systems and the rules and regulations that govern them. This sort of knowledge keeps the development "in-house" and keeps you employed. I do agree that simple development jobs should be done by the most efficient and appropriate people, normally either recent grads or outsourced developers. I mean, you wouldn't waste the Technical Architects time getting them to write basic code.
Someone looking for a career in IT needs to be constantly challenging themselves by learning new skills, and not always IT related ones so that your specialty keeps you needed. IT has never been an industry that rewards those that keep still (hell, if it did I would still be bashing out BASIC on my Vic 20!) but those that stay ahead of the game. Do this and you will have a career.
Or even better, default installations of the more popular OS's and Web servers (you know who you are) so that these security professionals-to-be get a taste of the real world!
Once they're handled this, then step it up to a fully patched and locked down version.
Whatever we think he should have done, if this story is true his actions are unprofessional. The ban on University servers acknowledges that they could be compromised with some effect on services, so to recommend to test it on unknown thirdparties is just saying "not in my backyard".
Wonderful post, although I'm not sure whether it should be modded "Troll (first class)" or "Joke (variety: subtle)" because it sure as hell will be misunderstood most people - much like evolu.... oh stop it!.
Delicate genius is a rare and beautiful flower, and this is of the genus Rafflesia. Slashdot barely deserves this post. Sir, I salute you.
of course you're not going to see banks running on it right now... duh:-)
You might be suprised. I work at a tier one investment bank and they are surprisingly flexible in their approach to new technologies if it will save them money. For small projects such as informational microsites, we have a certain amount of freedom, especially for prototypes. This is why we are currently using Cocoon from jakarta for one project (Verdict - awesomely flexible but a little slow) which is certainly not a standard platform. In this environment, RoR hasn't made the cut yet and I'm not sure that the "Convention over Configuration" aspect of it will work for larger projects where a certain amount of fine tuning is required.
Of course for the core business stuff, stability is the number one and they will throw more money than you can believe at a problem to solve it.
But in the Java world, pretty much all the frameworks ARE FOSS. Whether you choose the established players: Struts, WebWork, JSF, Spring, Hibernate, or the newer ones such as Groovy on Rails, all are FOSS.
I feel it's one of the benefits of Java is that the code "ecosystem" that has grown up around it is mostly FOSS, even extending to the Application Servers such as JBoss.
C# is a great language too, but you pay to use it, especially on larger systems (sorry, I don't include the Mono project for various reasons).
Okay, it's a nice list but come on! They aren't exactly heavy hitters!
All joking aside, I'll take more notice when I see a site dealing with serious traffic involving database transactions and other services. Else RoR will always been seen as a bit of a hobbyists plaything suitable for proto-typing or small sites only.
Basically, wake me up when the current gods are superceeded.
Which in the UK would be 15 quid for an album with probably 3 or 4 good tracks and the rest as fillers, so there's no way that I'd be buying another copy. Frankly, the Recording Industry Ass. of America are having a laugh.
I'd rather back up all my CDs (and of course rip them to my iPod), leave the originals at home and put the copy in my car/take to work, so should one get damaged/be stolen I haven't lost anything. At no point have I engaged in file trading or selling of pirated copies, I've merely protected my (costly) investment. Now that's fair use, and to complain that it's not is why the Recording Industry Ass. of America are labelled "Pigopolists".
Exactly how do we measure this level of certainty?
Statistics. Tests such as Chi-squared, Students T-Test and Mann-Whitney amongst others. So it is your option 3; Some sort of unbiased, impartial set of conditions.
I know you don't mean to be facetious about the coin through the table comment but that's what it comes across as. Your example displays either a distinct lack of familiarity with science or a general disdain. Any one who has been involved in science knows that with the standard scientific method you must do statistics in your experiments. This makes the difference between saying "there's an observable difference" which is okay, and "there is a statistically significant difference" which is your proof that somehting important is going on. This is one of the things that makes science science. It is an objective decision on what is proof and what is not, not some subjective opinion.
This is true for science as far apart as drug trials to cosmology.
"Proven" in this context can be regarded as being shorthand for "proven to a statistically significant degree of certainty" where the level of certainty is somewhere in the region of greater than 95% or more. "Absolute prove" really only exists in Mathematics, and saying the longhand version of proven as above isn't helpful to readability, clarity of thought, or speed of communication.
But where you are correct is that people will (willfully) misunderstand what "proven" represents either to support a view (as in ID and the word "theory") or just to prevent them from having to think.
As someone who's just finished a PhD, this is a hot topic for me, where words such as "proven", "theory", "shown", and "hypothesis" have both a scientific and colloquial meaning which many people seem to forget.
It implies a source of the bang that makes Creationists salivate.
How so? By admitting that we don't understand how it came about and what caused it? For someone to think that this supports creationism, there are two issues. Firstly, this is a "God of the Gaps" argument. This is just a statement about their disbelief that science will ever provide explanations for everything so they fall back to their default position, "God did it" which still explains or proves nothing. Secondly, "what occured before the Big Bang" cannot be answered with the creationist position of "God did" as the immediate response is "what occured before God?". The creationist might say that God has always been present which is no more or less valid that saying that the Big Bang has always been present (Big Bang --> Big Crunch --> Big Bang.... etc) so neither position has been shown to be any more valid.
So to say that the Big Bang is no longer popular with the evolutionist debate crowd, you must be referering to the sophists who debate for fun as opposed to scientists/evolutionists who still very much believe in the Big Bang.
I hear this a fair bit from people who dive into development without any prior research. Normally it's not their fault but their managers. Java is a mature language and so are the surrounding applications: servers, frameworks and IDEs. Confusion only exists in people with no experience with Java development, and is one reason that experienced Java developers are worth so much: their ability to recognise the correct choice from the many.
You're arguing that going the default route is better than taking a little bit of time at the start to find out which option is better for you. This way leads to shoehorning a ill-suited product into your project, so more time and more money.
This default option taking could be called the "Trabant hypothesis". The chioce you are given is the correct one because it's the only one available!
Well maybe it's so that those of us who don't regularly try out "bleeding-edge" products can learn what is good and what is bad.
While I take eveything I read on/. with a pinch of salt, I would expect the posters to be among the more technically literate. So a post like this is a great heads up to me as a Web Developer who also has input into supported platforms that neither Vista or IE7 are ready for consideration let alone usage.
And as for the various people saying that a beta product is supposed to have bugs, I'm guessing they are firm believers in the "release early, release often, patch always" school of programming. For MS, having stated that security is their main focus, this amounts to them saying: "Our systems and bug tracking is so bad we need you to tell us what's wrong". Releasing a beta version is an opportunity to display new features that may or may not work completely, not to display the inherent vulnerability of a product.
Actually no I don't think so. Anatomically standardised temporal lobectomies have been done since the 60's and is still used today. Other surgeries such as the fantastically named Multiple SubPial Transection are however new and would not have been available, but the parent was refering to epilepsy surgery in general.
While I'm not a medic I have spent 7 years at University studying Neurophysiology, I believe that the only anti-epilepsy drugs(AEDs)in the above list is Neurontin (gabapentin) which doesn't have that bad a side effect profile. All the others are for other psychiatric conditions, not neurophysiological ones.
I really feel for you having to take all these pills, you must rattle when you walk but the side effects of AEDs are really quite small when compared to their method of action and their widespread use and I would hate to think that people might not take their medication as they thought that AEDs had a similar profile to the other above drugs.
Yes. A PhD called "Investigation into the pathophysiology of epilepsy by electrical stimulation of the human cortex via intracranial electrodes implanted in patients undergoing pre-operative assessment for possible epilepsy surgery" submitted to King's College London. It's still awaiting viva but it covers all this topic.
More available is Engels CD Rom "Epilepsy: The Comprehensive Cd-Rom" is considered the gold standard for Epilepsy abd seizure information. What you want are the topics of recurrent inhibition, and temporal and spatial summation.
The original operation was a bilateral anterior temporal lobectomy. The patients subsequent anterograde amnesia was the result of both hippocampi being removed. Further tests, using anaesthetic not surgery, revealed that amnesia did not occur as long as one hippocampus could support memory function. Unilateral anterior temporal lobectomies were the result and are now the commonest surgical treatment for epilepsy and are very much accepted practice.
So you think that intention doesn't matter? The idea was to cure this persons dibilatating epilepsy, not investigate the function of the anterior hippocampus. It was an experimental procedure as his epilepsy was life-threateningly bad. Check SUDEP in google.
I believe it did. He was thought to have bilateral temporal lobe eiplepsy, the removal of which cured or at least reduced his seizures markedly but left him with severve anterograde amnesia.
Modern therapies for brain disorders are often highly dangerous, extremely toxic to the rest of the body, notorious for side-effects, often addictive, and many are poorly studied with completely unknown long-term consequences
And what do you base this comment on? Modern therapies are rarely dangerous (felbamate being the only modern therapy I would have said was dangerous and that is restricted), have few side effects especially compared to their action, aren't addictive, and are very intensively studied with long term effects based on the duration of their use. Surgery can also be fantastic for those with medically refractive epilepsy and with an assessment period of about 18 months can produce effects that are superior to drugs.
I think your post is either a troll or you are really quite ignorant about epilepsy treatment. I cannot for the life of me understand why you are currently rated +3 interesting.
Disclaimer: I don't work for any drug or surgical products companies
Actually as someone who has just finished a PhD in Neurophysiology I feel I may be a little better placed to comment than your average/. reader
Complex partial seizures originating in the temporal lobe have one of the best success rates in epilepsy surgery, but surgery is only offered to patients whos epilepsy is medically refractive (cannot be controlled by drugs) and affects their life in such as way that they would strongly benefit from surgery. Temporal lobe epilepsy is most often caused by mesial temporal or hippocampal sclerosis, this means that that part of the brain has become scarred and shrunk and this damage is causing the seizures. So this part of the brain supports a minimal amount of function. As your seizures are probably well controlled by drugs, you would never have been offered a surgical option.
we still don't know enough about the way it works to reliably fix problems that the brain itself cannot handle.
That's correct to a certain extent, but we do know a lot more and one of them is how to avoid causing the sort of condition that HM suffers.
It might be a bit off topic but at some point you have to draw the line on the constant misspelling and grammatical errors you see on Slashdot. Not to flame but if you can't do it in a thread on sliding educational values, when can you do it?
lose v. loose
it's v. its
grammar v. grammer
their v. there v. they're
may be v. maybe
Its teeth grindingly annoying as it's not difficult to distinguish between them and normally due to laziness, unlike typos caused by fast typing.
FileUpload from jakarta would be used for that as it's pretty much the de facto choice. There are other options like javazoom but it's an easy add-on. The apparent lack of a file upload control is probably reflective of the fact that Java allows multiple ways to do the same thing (which can be very confusing to someone new to the field) where MS products tend to have all the features included whether you use them or not. The JSF user community is not that small and is growing (although there are about 10x as many Struts jobs as JSF jobs currently in the UK) so it's one to take seriously.
The back button is an issue with JSF and is another reason why I use Struts for java web apps alongside it's easier integration with JSTL and greater maturity.
I prefer Struts to JSF as it's more mature and you can use more JSTL in it (JSF has issues with ). The JSTL fmt tags can also be used for formatting, including dates and this is recommended by the Struts team which is why they probably didn't both with a competing struts tag to do the same thing.
I'm also with you on the fact that what framework you use depends more on what language you are happiest coding in and what your company uses. Comparisons like this are a bit of fluff as you can say anything depending on your setup.
A brief decription of the object/class and then simple comments on any methods. That's a minimum but I would also go for single line comments for conceptually difficult peices of code that you know you will forget in a couple of weeks. Not overly rigorous but easy enough that people do follow it.
A good model for me would be the Java SDK docs and the javadocs tool but that's just me.
Slashdot Mirror
My favourite was one I saw last year, a requirement for 5 years Java/J2EE (okay)and 5 years C#/dot NET (eh!). Apart from being difficult to have 5 years experience in something that came out in 2002, I'm not sure that I would want to work for a company with this bad a grasp of skills management.
I think you're right about the market, and about how people only need a few weeks to get up to speed on new stuff (it's not brain surgery is it!) but the crunch is always with the contractors. Trying to stay ahead of the game is tough as you end up in a catch 22 where people will only hire you if you already have experience in something
The fact that you've found good people is probably more a reflection of your abililty as a manager. Your time "at the coalface" gives you an insight into how to hire someone that might not have the skills now, but would be fantastic with a little training. With too many managers, that's a risk they can't take as they just can't see potential, so they fall back on proven knowledge - experience.
One of the things that always troubles me with the Outsourcing debate is how it regards IT and software development as an entity in itself, rather than one that must deal with others. By this I mean both dealing with the business you are in and also the other departments in your company. By making IT a commodity, it can be offshored or outsourced easily. When it's a specialty, that becomes difficult to impossible.
If you are developing a piece of medical software such as an EEG recorder, you need to have some understanding of the science of EEGs and the medical background in which they are used. Likewise, a piece of financial software requires detailed knowledge of financial systems and the rules and regulations that govern them. This sort of knowledge keeps the development "in-house" and keeps you employed. I do agree that simple development jobs should be done by the most efficient and appropriate people, normally either recent grads or outsourced developers. I mean, you wouldn't waste the Technical Architects time getting them to write basic code.
Someone looking for a career in IT needs to be constantly challenging themselves by learning new skills, and not always IT related ones so that your specialty keeps you needed. IT has never been an industry that rewards those that keep still (hell, if it did I would still be bashing out BASIC on my Vic 20!) but those that stay ahead of the game. Do this and you will have a career.
Or even better, default installations of the more popular OS's and Web servers (you know who you are) so that these security professionals-to-be get a taste of the real world!
Once they're handled this, then step it up to a fully patched and locked down version.
Whatever we think he should have done, if this story is true his actions are unprofessional. The ban on University servers acknowledges that they could be compromised with some effect on services, so to recommend to test it on unknown thirdparties is just saying "not in my backyard".
Wonderful post, although I'm not sure whether it should be modded "Troll (first class)" or "Joke (variety: subtle)" because it sure as hell will be misunderstood most people - much like evolu.... oh stop it!.
Delicate genius is a rare and beautiful flower, and this is of the genus Rafflesia. Slashdot barely deserves this post. Sir, I salute you.
of course you're not going to see banks running on it right now... duh:-)
You might be suprised. I work at a tier one investment bank and they are surprisingly flexible in their approach to new technologies if it will save them money. For small projects such as informational microsites, we have a certain amount of freedom, especially for prototypes. This is why we are currently using Cocoon from jakarta for one project (Verdict - awesomely flexible but a little slow) which is certainly not a standard platform. In this environment, RoR hasn't made the cut yet and I'm not sure that the "Convention over Configuration" aspect of it will work for larger projects where a certain amount of fine tuning is required.
Of course for the core business stuff, stability is the number one and they will throw more money than you can believe at a problem to solve it.
But in the Java world, pretty much all the frameworks ARE FOSS. Whether you choose the established players: Struts, WebWork, JSF, Spring, Hibernate, or the newer ones such as Groovy on Rails, all are FOSS.
I feel it's one of the benefits of Java is that the code "ecosystem" that has grown up around it is mostly FOSS, even extending to the Application Servers such as JBoss.
C# is a great language too, but you pay to use it, especially on larger systems (sorry, I don't include the Mono project for various reasons).
Okay, it's a nice list but come on! They aren't exactly heavy hitters!
All joking aside, I'll take more notice when I see a site dealing with serious traffic involving database transactions and other services. Else RoR will always been seen as a bit of a hobbyists plaything suitable for proto-typing or small sites only.
Basically, wake me up when the current gods are superceeded.
"Affordable prices"
Which in the UK would be 15 quid for an album with probably 3 or 4 good tracks and the rest as fillers, so there's no way that I'd be buying another copy. Frankly, the Recording Industry Ass. of America are having a laugh.
I'd rather back up all my CDs (and of course rip them to my iPod), leave the originals at home and put the copy in my car/take to work, so should one get damaged/be stolen I haven't lost anything. At no point have I engaged in file trading or selling of pirated copies, I've merely protected my (costly) investment. Now that's fair use, and to complain that it's not is why the Recording Industry Ass. of America are labelled "Pigopolists".
Exactly how do we measure this level of certainty?
Statistics. Tests such as Chi-squared, Students T-Test and Mann-Whitney amongst others. So it is your option 3; Some sort of unbiased, impartial set of conditions.
I know you don't mean to be facetious about the coin through the table comment but that's what it comes across as. Your example displays either a distinct lack of familiarity with science or a general disdain. Any one who has been involved in science knows that with the standard scientific method you must do statistics in your experiments. This makes the difference between saying "there's an observable difference" which is okay, and "there is a statistically significant difference" which is your proof that somehting important is going on. This is one of the things that makes science science. It is an objective decision on what is proof and what is not, not some subjective opinion.
This is true for science as far apart as drug trials to cosmology.
Not really.
"Proven" in this context can be regarded as being shorthand for "proven to a statistically significant degree of certainty" where the level of certainty is somewhere in the region of greater than 95% or more. "Absolute prove" really only exists in Mathematics, and saying the longhand version of proven as above isn't helpful to readability, clarity of thought, or speed of communication.
But where you are correct is that people will (willfully) misunderstand what "proven" represents either to support a view (as in ID and the word "theory") or just to prevent them from having to think.
As someone who's just finished a PhD, this is a hot topic for me, where words such as "proven", "theory", "shown", and "hypothesis" have both a scientific and colloquial meaning which many people seem to forget.
It implies a source of the bang that makes Creationists salivate.
How so? By admitting that we don't understand how it came about and what caused it? For someone to think that this supports creationism, there are two issues. Firstly, this is a "God of the Gaps" argument. This is just a statement about their disbelief that science will ever provide explanations for everything so they fall back to their default position, "God did it" which still explains or proves nothing. Secondly, "what occured before the Big Bang" cannot be answered with the creationist position of "God did" as the immediate response is "what occured before God?". The creationist might say that God has always been present which is no more or less valid that saying that the Big Bang has always been present (Big Bang --> Big Crunch --> Big Bang .... etc) so neither position has been shown to be any more valid.
So to say that the Big Bang is no longer popular with the evolutionist debate crowd, you must be referering to the sophists who debate for fun as opposed to scientists/evolutionists who still very much believe in the Big Bang.
I hear this a fair bit from people who dive into development without any prior research. Normally it's not their fault but their managers. Java is a mature language and so are the surrounding applications: servers, frameworks and IDEs. Confusion only exists in people with no experience with Java development, and is one reason that experienced Java developers are worth so much: their ability to recognise the correct choice from the many.
You're arguing that going the default route is better than taking a little bit of time at the start to find out which option is better for you. This way leads to shoehorning a ill-suited product into your project, so more time and more money.
This default option taking could be called the "Trabant hypothesis". The chioce you are given is the correct one because it's the only one available!
What is the point of this post?
Well maybe it's so that those of us who don't regularly try out "bleeding-edge" products can learn what is good and what is bad.
While I take eveything I read on /. with a pinch of salt, I would expect the posters to be among the more technically literate. So a post like this is a great heads up to me as a Web Developer who also has input into supported platforms that neither Vista or IE7 are ready for consideration let alone usage.
And as for the various people saying that a beta product is supposed to have bugs, I'm guessing they are firm believers in the "release early, release often, patch always" school of programming. For MS, having stated that security is their main focus, this amounts to them saying: "Our systems and bug tracking is so bad we need you to tell us what's wrong". Releasing a beta version is an opportunity to display new features that may or may not work completely, not to display the inherent vulnerability of a product.
Actually no I don't think so. Anatomically standardised temporal lobectomies have been done since the 60's and is still used today. Other surgeries such as the fantastically named Multiple SubPial Transection are however new and would not have been available, but the parent was refering to epilepsy surgery in general.
I really feel for you having to take all these pills, you must rattle when you walk but the side effects of AEDs are really quite small when compared to their method of action and their widespread use and I would hate to think that people might not take their medication as they thought that AEDs had a similar profile to the other above drugs.
Yes. A PhD called "Investigation into the pathophysiology of epilepsy by electrical stimulation of the human cortex via intracranial electrodes implanted in patients undergoing pre-operative assessment for possible epilepsy surgery" submitted to King's College London. It's still awaiting viva but it covers all this topic.
More available is Engels CD Rom "Epilepsy: The Comprehensive Cd-Rom" is considered the gold standard for Epilepsy abd seizure information. What you want are the topics of recurrent inhibition, and temporal and spatial summation.
The original operation was a bilateral anterior temporal lobectomy. The patients subsequent anterograde amnesia was the result of both hippocampi being removed. Further tests, using anaesthetic not surgery, revealed that amnesia did not occur as long as one hippocampus could support memory function. Unilateral anterior temporal lobectomies were the result and are now the commonest surgical treatment for epilepsy and are very much accepted practice.
So you think that intention doesn't matter? The idea was to cure this persons dibilatating epilepsy, not investigate the function of the anterior hippocampus. It was an experimental procedure as his epilepsy was life-threateningly bad. Check SUDEP in google.
I believe it did. He was thought to have bilateral temporal lobe eiplepsy, the removal of which cured or at least reduced his seizures markedly but left him with severve anterograde amnesia.
Modern therapies for brain disorders are often highly dangerous, extremely toxic to the rest of the body, notorious for side-effects, often addictive, and many are poorly studied with completely unknown long-term consequences
And what do you base this comment on? Modern therapies are rarely dangerous (felbamate being the only modern therapy I would have said was dangerous and that is restricted), have few side effects especially compared to their action, aren't addictive, and are very intensively studied with long term effects based on the duration of their use. Surgery can also be fantastic for those with medically refractive epilepsy and with an assessment period of about 18 months can produce effects that are superior to drugs.
I think your post is either a troll or you are really quite ignorant about epilepsy treatment. I cannot for the life of me understand why you are currently rated +3 interesting.
Disclaimer: I don't work for any drug or surgical products companies
Complex partial seizures originating in the temporal lobe have one of the best success rates in epilepsy surgery, but surgery is only offered to patients whos epilepsy is medically refractive (cannot be controlled by drugs) and affects their life in such as way that they would strongly benefit from surgery. Temporal lobe epilepsy is most often caused by mesial temporal or hippocampal sclerosis, this means that that part of the brain has become scarred and shrunk and this damage is causing the seizures. So this part of the brain supports a minimal amount of function. As your seizures are probably well controlled by drugs, you would never have been offered a surgical option.
we still don't know enough about the way it works to reliably fix problems that the brain itself cannot handle.
That's correct to a certain extent, but we do know a lot more and one of them is how to avoid causing the sort of condition that HM suffers.
Hell no I disagree!
It might be a bit off topic but at some point you have to draw the line on the constant misspelling and grammatical errors you see on Slashdot. Not to flame but if you can't do it in a thread on sliding educational values, when can you do it?
Its teeth grindingly annoying as it's not difficult to distinguish between them and normally due to laziness, unlike typos caused by fast typing.
FileUpload from jakarta would be used for that as it's pretty much the de facto choice. There are other options like javazoom but it's an easy add-on. The apparent lack of a file upload control is probably reflective of the fact that Java allows multiple ways to do the same thing (which can be very confusing to someone new to the field) where MS products tend to have all the features included whether you use them or not. The JSF user community is not that small and is growing (although there are about 10x as many Struts jobs as JSF jobs currently in the UK) so it's one to take seriously.
The back button is an issue with JSF and is another reason why I use Struts for java web apps alongside it's easier integration with JSTL and greater maturity.
I prefer Struts to JSF as it's more mature and you can use more JSTL in it (JSF has issues with ). The JSTL fmt tags can also be used for formatting, including dates and this is recommended by the Struts team which is why they probably didn't both with a competing struts tag to do the same thing.
I'm also with you on the fact that what framework you use depends more on what language you are happiest coding in and what your company uses. Comparisons like this are a bit of fluff as you can say anything depending on your setup.
A brief decription of the object/class and then simple comments on any methods. That's a minimum but I would also go for single line comments for conceptually difficult peices of code that you know you will forget in a couple of weeks. Not overly rigorous but easy enough that people do follow it.
A good model for me would be the Java SDK docs and the javadocs tool but that's just me.