It all really depends. Ultimately I feel that you will learn to write code that explains itself to a large degree. There are two times I ask them to make comments.
1.) The first time to comment is when you are doing something you know are not standard. Bitwise shifts to multiply, things you know are sub-optimal etc. Exceptional things that are not done how they shold be done, or are typically done.
2.) When you want to explain how a relatively large block of code is working, or an entire classes purposes. Explaining a method or a class. Or a block of code to give an idea of flow.
Another thing that I feel improves comment quality is to force them to be complete sentences. Try and keep the sentences brief and complete so that it clearly illustrates what you are trying to say without repeating. The "Do not Repeat Yourself" DRY principle applies to written communication as well as code.
If you always refer back to these simple rules you will generally have code that is easier to maintain and that the comments don't go out of date on as quickly. I really think you could write an entire book on commenting code and naming variables and organizing code flow. Good Luck and happy coding!
I will agree that developers go through phases in how and when they comment, but having someone lay out guidelines and telling you when to comment and giving you a few concerete examples makes a huge difference. For me its always about being a good leader and doing what I say. I have seen programmers come from mediocre/average just following a few simple "rules". Commenting is a huge part of it. Because if you can comment at an abstract/high level you can prove you know what the code should be doing. And if you write comments before you write the code....;-)
You fail to understand the synergies of the technologies and how they infuse RAD developers with powerful tools to let them get their jobs done. *frowns* I will stop now. *hangs his head in shame* But its allt rue I tell you!
Not my community college. The intro to CS class has you in C++. Its a serious program that Georgia Tech readily accepts folks with 2yr degrees from. I grew up on C and network programming. Then 4GL web application programming languages. Kind of the reverse of the scenario the OP is referencing. The 4GLs are good as long as you apply good development concepts everywhere you go. For instance I use a well built MVC framework and lots of OO concepts and the occasional design pattern in a 4GL language that lives ontop of a J2EE compliant web application server many people would probably just discard offhand.
The best way I have ever heard someone explain how you can tell of an interface is good is if you look at it and think, "Of course. How else would you do it?". The design is natural and flows well for your average *skilled* programmer. If someone is very taltented but no one else can understand their designs then the code is not maintainable. Code that is not maintainable is not good. Creating unmaintainable code is something BAD programmers do. So you should consider that fella brilliantly BAD;) The strongest strength anyone can have is realizing their limits and working around or past them. If someone never realizes they are limited they will never grow;)
Not to mention if you write "correct" and cohesive code you can easily optimize it.
As long as you don't over architect or go to extremes of abstraction your can do some very well abstracted code that pretty nicely separates all of the things that change from the things that stay the same. When you create this kind of separation optimization becomes easier and easier.
Good (simple) examples are creating a generic interface to sort. You may start off with a simple insertion sort (why bother with heap/quick if you don't need it?). Anyone can understand insertion sort and its fine for relatively small sets of data. Then you expand the scope of the program and realize you need a better sorting algorithm. No problem! You go in and change the behind the scenes sort. Your interface to sort did not change and your program happily works just how it did. You apply this very same idea to as much of the software you write as is reasonable. Separate out data models and logic and displays. Keep each part of the system doing one task and doing it well.
If the system is written good enough you don't even need anything but high level comments explaining the "big picture" of what is going on. I hate comments that are like the following:
/* Set foo to 10 */ int foo = 10;
Well thank you captain obvious!
Much better is an explanation over a block of 5-10 lines giving you an idea of what you are trying to achieve. Comment any thing that is not clear, like if your using bitwise shifts to multiply and divide, for example.
I thought I did just admit it and move on saying that my argument as presented was pointless;) How much more of an admission do you expect on slashdot?
The argument as I presented it is pointless. I was just trying to relay something I read, but I shouldn't have because: A.) It wasn't a valid comparison B.) The way I presented was more opinion than anything. So I will just say what I am thinking. I don't know that I can enlighten you about anything. I really only have my opinion based on simple observations and some studies I read that pointed out this correlation (whatever it proves, IF nothing).
Quite honestly I don't agree with the life style. It is not for me. I am biased towards "traditional" family values (its to bad christians have hijacked this terminology to have religious overtones). I have friends who are gay. They KNOW I don't agree with what they do. But I would never for a second interfere with someone elses pursuit of fundamental happiness and freedom. As long as what they do does not hamper anyone but the parties that decided to be involved.
AIDS is way to prevalent in our country given our degree of development. Our healthcare system is already straining due to people who can't pay their bills. Eventually the really sick can't work, can't afford insurance. If this really does cost tax dollars then its a bigger issue. Its pretty simple and it goes beyond anything but the simple question of "does this lifestyle burden society beyond what is currently expected and planned for?" If the answer is "no" then I am ok with it and I am only against it in opinion. If the answer is ultimately found to be "yes", then there needs to be additional efforts taken to help these people understand. And it goes WAY beyond gays/lesbians. Its any kind of promiscuous sex that can result in transmission/spread of STDs. Or really just any choices people actively make that result in others being responsible for their poor choices.
So I just don't want to go into this with a complete rubber seal of approval. People make choices, and some of the time those choices have consequences that may affect other people.
No. I did read it, but that doesn't mean what I remember is correct since its not something I even really am concerned with. People are free to do what they want. I was just pointing out that the medical communities consider it an "at risk", or "riskier than average" lifestyle.
FWIW, thats all I am saying. Honestly, anything that does not directly interfere with my pursuit of life, liberty and happiness is fine by me. I am just making a point that there are more STDs out there in gay/lesbian communities.
I am bad for using the term "Marriage". I mean commitment to a partner for life. I have no clue and I don't have time to dig up the stats so take what I am saying at face value. But from what I recall reading the average "lifelong commited" monogamous relationship among gays don't last very long. Its part of why STDs are more prevalent and the medical folks label it an "risky" lifestyle. Not that I agree with any of it, but assuming that is a fact its an interesting one.
Their weddings may be more fun because the average length of a gay/lesbian marriage is much shorter than your typical hetero marriage. Though most hetero marriages are pretty sad affairs too.
My wedding was at a golf course with lots of beer for those who wanted it. It was quite fun:-)
Its not about rewriting code, its just about getting a compiler that can translate your code to "MISL" that will run seamlessly on both platforms.
The real deal isn't just having the language standardized. It really is kinda meaningless to have the language standardized in my opinion. It is something, but such a small something that its like a "SMALL NEGATIVE" that approaches 0;) You also need to have complete compatability with the.NET runtime to make it truly seamless. IE: Compile C++ to MSIL and have it run on the.NET and Mono frameworks. And, some day, you could have all Win32 API calls translated into platform independent calls. That , theoreticall, breaks your coupling to Win32 since the Mono framework can then just take the IL and it should be able to process the IL and render graphics calls and Win32 calls without ever actually knowing they were Win32...... I think we all know how likely it is that having a working Win32 API, or compiler than can generically interpret and translate Win32 is though.
As long as MS controls MSIL it controls the platform. It will still be NICE to use Mono as it is nice to use Java and C# when they suit your needs. But the dream of compatability is going to be just that, a dream.
I do think you have a good point about some languages being good at different tasks. I think its even better to just look at computation from a more generic perspective to get a true picture. Computer programs are written to solve problems for the most part. Most problems will have a "problem domain" so to speak. The problem, and problem domain, are key factors in choosing a language.
System Automation - For a quick and easy automation task on a server I administer I choose a light-weight and dynamic language that lets me very expressively accomplish some task.
Big Business - Automating a 10,000 person brick and mortar company with virtually no technology. Things like ERP, CRM, and B2B (god forgive me for the buzzwords) are what are on your mind then. You need to integrate payrolls, manage sales, and integrate with other businesses and resources out there. Your problem domain is so large that multiple languages and tools will be used to fill the gap. And.. depending on your business you may even get into domain specific languages *or* invent new domain specific models of computation that are better at solving your particular problem.
So you can look at the most simple thing to a very large task and see that there are enermous differences in scope and scale, yet we are solving a problem, or a series of problems. Each problem and facet of the problem must be addressed and solved (correctly). Some of the time your model and business domain may not fit anything very well . Some of the time it will.
We have learned that OO programming tends to work reasonably well at modelling most problems in the business domain. As a system administrator you know that something like Perl or Python is quite good at solving problems in the system administration domain. As an embedded programmer you know that assembler can be used to solve problems when efficiency is absolutely required... but the point is that if you take things one level hire and look at the domain of the problem you are solving certain types of tools tend to work better than others. By knowing the TYPES of tools out there, not just one languages implementation of a concept, you can remain more flexible. If the problem is large enough and has enough scope, using a domain specific, possibly in house created/project specific, language to model the problem can be acceptable. It all just depends on the problem you are solving:)
Its on a mainstream news outlet. How bad do you think they can possibly be. The whole point was that they were completely digitally removed to the point where if you just saw them as pictures you would have NO idea.
Many internal, or administrative, web applications target Internet explorer to do fancier DHTML etc. without worrying about writing standard compliant code;)
Get it done;) Its just an internal application, what does it matter?
Thats why there is a new movement to design lightweight Java frameworks that are small enough and agile enough to move with the times.
Check out Spring and Hibernate. They are easy enough to learn that its not like a MAJOR time investment to get your work done (as compared to say EJB). At the end of the day you learn what you need to accomplish A B and C. The frameworks are written well, are easy to understand, and attempt to have as loose a relationship to your application as possible.
A winning combination in my opinion. Something Java has sorely lacked. In all of my J2EE developments I found it cumbersome, but tolerable after many man hours of investment in reading documentation and experimenting. Then I found Spring and Hibernate. I spent a few hours playing around and I was doing things that had me thinking, "Well....". But then I moved on to writing more code and it just worked. Maybe its a fad, maybe not, but at least it works for me and it was damn easy to do.
Search for the text "1978" and it takes you right to his 1978 house of representatives run. I have read about this from multiple sources, and this link is encarta. Seems credible enough to me.
Interesting that most business applications that get work done from day to day are basically nothing but database operations. Sure there is a lot of processing and such in some businesses, but it is mostly about tracking and managing data for analysis or improvement of business process.
Its mostly things you would never want as a home user. Things like access to the administrative file shares IE: \\server\c$. Other networking things, no IIS, etc.
Slashdot Mirror
THe ??? is taking the idea and turning it into the real working product that does something useful people will pay you for ;-)
Jeremy
(doing what I say as in.. practice what I preach :)
It all really depends. Ultimately I feel that you will learn to write code that explains itself to a large degree. There are two times I ask them to make comments.
;-)
1.) The first time to comment is when you are doing something you know are not standard. Bitwise shifts to multiply, things you know are sub-optimal etc. Exceptional things that are not done how they shold be done, or are typically done.
2.) When you want to explain how a relatively large block of code is working, or an entire classes purposes. Explaining a method or a class. Or a block of code to give an idea of flow.
Another thing that I feel improves comment quality is to force them to be complete sentences. Try and keep the sentences brief and complete so that it clearly illustrates what you are trying to say without repeating. The "Do not Repeat Yourself" DRY principle applies to written communication as well as code.
If you always refer back to these simple rules you will generally have code that is easier to maintain and that the comments don't go out of date on as quickly. I really think you could write an entire book on commenting code and naming variables and organizing code flow. Good Luck and happy coding!
I will agree that developers go through phases in how and when they comment, but having someone lay out guidelines and telling you when to comment and giving you a few concerete examples makes a huge difference. For me its always about being a good leader and doing what I say. I have seen programmers come from mediocre/average just following a few simple "rules". Commenting is a huge part of it. Because if you can comment at an abstract/high level you can prove you know what the code should be doing. And if you write comments before you write the code....
Jeremy
You fail to understand the synergies of the technologies and how they infuse RAD developers with powerful tools to let them get their jobs done. *frowns* I will stop now. *hangs his head in shame* But its allt rue I tell you!
Jeremy
Not my community college. The intro to CS class has you in C++. Its a serious program that Georgia Tech readily accepts folks with 2yr degrees from. I grew up on C and network programming. Then 4GL web application programming languages. Kind of the reverse of the scenario the OP is referencing. The 4GLs are good as long as you apply good development concepts everywhere you go. For instance I use a well built MVC framework and lots of OO concepts and the occasional design pattern in a 4GL language that lives ontop of a J2EE compliant web application server many people would probably just discard offhand.
Jeremy
Not such a brilliant coder then ehh?
;) The strongest strength anyone can have is realizing their limits and working around or past them. If someone never realizes they are limited they will never grow ;)
The best way I have ever heard someone explain how you can tell of an interface is good is if you look at it and think, "Of course. How else would you do it?". The design is natural and flows well for your average *skilled* programmer. If someone is very taltented but no one else can understand their designs then the code is not maintainable. Code that is not maintainable is not good. Creating unmaintainable code is something BAD programmers do. So you should consider that fella brilliantly BAD
Jeremy
Not to mention if you write "correct" and cohesive code you can easily optimize it.
/* Set foo to 10 */
:)
As long as you don't over architect or go to extremes of abstraction your can do some very well abstracted code that pretty nicely separates all of the things that change from the things that stay the same. When you create this kind of separation optimization becomes easier and easier.
Good (simple) examples are creating a generic interface to sort. You may start off with a simple insertion sort (why bother with heap/quick if you don't need it?). Anyone can understand insertion sort and its fine for relatively small sets of data. Then you expand the scope of the program and realize you need a better sorting algorithm. No problem! You go in and change the behind the scenes sort. Your interface to sort did not change and your program happily works just how it did. You apply this very same idea to as much of the software you write as is reasonable. Separate out data models and logic and displays. Keep each part of the system doing one task and doing it well.
If the system is written good enough you don't even need anything but high level comments explaining the "big picture" of what is going on. I hate comments that are like the following:
int foo = 10;
Well thank you captain obvious!
Much better is an explanation over a block of 5-10 lines giving you an idea of what you are trying to achieve. Comment any thing that is not clear, like if your using bitwise shifts to multiply and divide, for example.
Strive for simplicity
Jeremy
I thought I did just admit it and move on saying that my argument as presented was pointless ;) How much more of an admission do you expect on slashdot?
The argument as I presented it is pointless. I was just trying to relay something I read, but I shouldn't have because: A.) It wasn't a valid comparison B.) The way I presented was more opinion than anything. So I will just say what I am thinking. I don't know that I can enlighten you about anything. I really only have my opinion based on simple observations and some studies I read that pointed out this correlation (whatever it proves, IF nothing).
Quite honestly I don't agree with the life style. It is not for me. I am biased towards "traditional" family values (its to bad christians have hijacked this terminology to have religious overtones). I have friends who are gay. They KNOW I don't agree with what they do. But I would never for a second interfere with someone elses pursuit of fundamental happiness and freedom. As long as what they do does not hamper anyone but the parties that decided to be involved.
AIDS is way to prevalent in our country given our degree of development. Our healthcare system is already straining due to people who can't pay their bills. Eventually the really sick can't work, can't afford insurance. If this really does cost tax dollars then its a bigger issue. Its pretty simple and it goes beyond anything but the simple question of "does this lifestyle burden society beyond what is currently expected and planned for?" If the answer is "no" then I am ok with it and I am only against it in opinion. If the answer is ultimately found to be "yes", then there needs to be additional efforts taken to help these people understand. And it goes WAY beyond gays/lesbians. Its any kind of promiscuous sex that can result in transmission/spread of STDs. Or really just any choices people actively make that result in others being responsible for their poor choices.
So I just don't want to go into this with a complete rubber seal of approval. People make choices, and some of the time those choices have consequences that may affect other people.
No. I did read it, but that doesn't mean what I remember is correct since its not something I even really am concerned with. People are free to do what they want. I was just pointing out that the medical communities consider it an "at risk", or "riskier than average" lifestyle.
FWIW, thats all I am saying. Honestly, anything that does not directly interfere with my pursuit of life, liberty and happiness is fine by me. I am just making a point that there are more STDs out there in gay/lesbian communities.
Jeremy
I am bad for using the term "Marriage". I mean commitment to a partner for life. I have no clue and I don't have time to dig up the stats so take what I am saying at face value. But from what I recall reading the average "lifelong commited" monogamous relationship among gays don't last very long. Its part of why STDs are more prevalent and the medical folks label it an "risky" lifestyle. Not that I agree with any of it, but assuming that is a fact its an interesting one.
Jeremy
Their weddings may be more fun because the average length of a gay/lesbian marriage is much shorter than your typical hetero marriage. Though most hetero marriages are pretty sad affairs too.
:-)
My wedding was at a golf course with lots of beer for those who wanted it. It was quite fun
Jeremy
Its not about rewriting code, its just about getting a compiler that can translate your code to "MISL" that will run seamlessly on both platforms.
;) You also need to have complete compatability with the .NET runtime to make it truly seamless. IE: Compile C++ to MSIL and have it run on the .NET and Mono frameworks. And, some day, you could have all Win32 API calls translated into platform independent calls. That , theoreticall, breaks your coupling to Win32 since the Mono framework can then just take the IL and it should be able to process the IL and render graphics calls and Win32 calls without ever actually knowing they were Win32...... I think we all know how likely it is that having a working Win32 API, or compiler than can generically interpret and translate Win32 is though.
The real deal isn't just having the language standardized. It really is kinda meaningless to have the language standardized in my opinion. It is something, but such a small something that its like a "SMALL NEGATIVE" that approaches 0
As long as MS controls MSIL it controls the platform. It will still be NICE to use Mono as it is nice to use Java and C# when they suit your needs. But the dream of compatability is going to be just that, a dream.
Jeremy
According to Bruce Schneier a machine can be built that can do it in 56 hours.
Jeremy
And yet, you somehow posted this to Slashdot. Lets hope you aren't on your work network right now ;)
Jeremy
I do think you have a good point about some languages being good at different tasks. I think its even better to just look at computation from a more generic perspective to get a true picture. Computer programs are written to solve problems for the most part. Most problems will have a "problem domain" so to speak. The problem, and problem domain, are key factors in choosing a language.
:)
System Automation - For a quick and easy automation task on a server I administer I choose a light-weight and dynamic language that lets me very expressively accomplish some task.
Big Business - Automating a 10,000 person brick and mortar company with virtually no technology. Things like ERP, CRM, and B2B (god forgive me for the buzzwords) are what are on your mind then. You need to integrate payrolls, manage sales, and integrate with other businesses and resources out there. Your problem domain is so large that multiple languages and tools will be used to fill the gap. And.. depending on your business you may even get into domain specific languages *or* invent new domain specific models of computation that are better at solving your particular problem.
So you can look at the most simple thing to a very large task and see that there are enermous differences in scope and scale, yet we are solving a problem, or a series of problems. Each problem and facet of the problem must be addressed and solved (correctly). Some of the time your model and business domain may not fit anything very well . Some of the time it will.
We have learned that OO programming tends to work reasonably well at modelling most problems in the business domain. As a system administrator you know that something like Perl or Python is quite good at solving problems in the system administration domain. As an embedded programmer you know that assembler can be used to solve problems when efficiency is absolutely required... but the point is that if you take things one level hire and look at the domain of the problem you are solving certain types of tools tend to work better than others. By knowing the TYPES of tools out there, not just one languages implementation of a concept, you can remain more flexible. If the problem is large enough and has enough scope, using a domain specific, possibly in house created/project specific, language to model the problem can be acceptable. It all just depends on the problem you are solving
Jeremy
Its on a mainstream news outlet. How bad do you think they can possibly be. The whole point was that they were completely digitally removed to the point where if you just saw them as pictures you would have NO idea.
Jeremy
Many internal, or administrative, web applications target Internet explorer to do fancier DHTML etc. without worrying about writing standard compliant code ;)
;) Its just an internal application, what does it matter?
Get it done
Jeremy
Thats why there is a new movement to design lightweight Java frameworks that are small enough and agile enough to move with the times.
Check out Spring and Hibernate. They are easy enough to learn that its not like a MAJOR time investment to get your work done (as compared to say EJB). At the end of the day you learn what you need to accomplish A B and C. The frameworks are written well, are easy to understand, and attempt to have as loose a relationship to your application as possible.
A winning combination in my opinion. Something Java has sorely lacked. In all of my J2EE developments I found it cumbersome, but tolerable after many man hours of investment in reading documentation and experimenting. Then I found Spring and Hibernate. I spent a few hours playing around and I was doing things that had me thinking, "Well....". But then I moved on to writing more code and it just worked. Maybe its a fad, maybe not, but at least it works for me and it was damn easy to do.
Jeremy
See here:
Encarta
Search for the text "1978" and it takes you right to his 1978 house of representatives run. I have read about this from multiple sources, and this link is encarta. Seems credible enough to me.
Jeremy
Interesting that most business applications that get work done from day to day are basically nothing but database operations. Sure there is a lot of processing and such in some businesses, but it is mostly about tracking and managing data for analysis or improvement of business process.
Jeremy
There is a good quote I read/heard somewhere. "Discipline is what you do when no one is looking">
Jeremy
Its mostly things you would never want as a home user. Things like access to the administrative file shares IE: \\server\c$. Other networking things, no IIS, etc.
Jeremy
Oh yeah? Well we had to create our proteins molecule by molecule!
We used it about a year or two ago to implement a Jabber compliant chat client :)
:)
The actual code dealing with the jabber protocl was event driven and quite modular
Jeremy