I think the "Connections" series attacked this issue the best. Innovation is an itterative proccess and it builds on itself. Different things change at different times and it takes society time to obsorb these changes. The defining factor for the integration of the technology that we see today was World War II. How long would it have taken for the aircraft industry to catch on without the industrial build up of the war?
Now wait a few years and you will see some interesting things. We already see the reactions of businesss to the increased communication abilities given by the internet. Non-Compete and Non-Disclosure agreement are in part a result of workers abilities to shop the world for employment opportunities. "Copyright protection" is the result of content providers inability to accept the fact the instant content access dilutes the value of a single work. How much is one work worth when thousands can be accessed in seconds. Increased competitition lowers prices.
Add DNA research, advance transportation techniques, advances in particle physics...etc into the equation and we haven't seen anything yet!!!
I don't think many "consumers" will be up in arms against this kind of thing. They don't think twice when they are told that they are denied access to data on their own machine. (Try deleting certain cookies files under windows 95)
Now the corporate market is another question. I am not talking about desktops but small-scale (intel) server roleouts. In another life, I worked for a company that had a couple thousand rackmounts. They did the following:
1.They quite often used NFS, rcp, ghost and just about anything else they could to keep these things running.
2.To make matters worse, many of these machines would pull info from RS6000/Sun/HP/SGI machines.
3. Some rackmounts used strifing or mirroring.
I can guarantee that the people that run things at this place won't even pay attention until problems crop up. That's when things will get interesting.
How many lawyers do you know that design electronic coponents? They live in a legal not physical world as do most of the execs that decide what to do with these cases. Lisence and let the other guy spend the legal fees is the usual course of action. Later they can bitch that the competition was not nice enough to fight the claim.
Once 'Whistler' aks Win2000 consumer hits the shelves everyone(but us linux users) will need an eventual upgrade unless they bought theor machine within the last year. They most likely won't get the snappy performance they got with Win9x without at least:
500+ Mhz PIII/Athalon
128+ MB memory
IMHO, I think this has a lot to do with Intels recent troubles. Intel has had production problems and two recalls in recent memory. No matter what you think of AMD, they don't have the capacity to supply the market on their own. However, AMD does represent an alternative source of parts and slow adaption of AMD parts may have cut into many PC makers earnings and time to market. Wait until Intel get their act in order, we may see a quick upswing.
In general, most analysts think that what the big boys are doing now is indicative of the future. Give Wintel a six or seven months and we will see another upgrade cycle.
Take my workplace for example, they typicaly pay an entry level H1B worker 35-42K while the entry level US citzen of the same skill would get paid 45-50K. Ofcourse, there are hidden costs such as legal fees. But the end result for the US workers will be a downward wage pressure since accountants will judge salary ranges on medians and deviations of the current batch of salaries.
Most H1B workers come from countries where it would be difficult to demand commenserate salaries. So they are mostly happy with what they get EVEN though a FREE MARKET would give them 15-30 percent more in real earnings.
So, in the end the employers are the only ones with no downside. The extra legal fees are easily offset by the guaranteed retention.
I am not saying that you should not be working under an H1B. I think you should have been given a green card so that you could freely participate in the job market. Those who can not are prisoners or indentured servents.
I just don't aprove of the hinderance of the free market. Free Market means more than just corprate rights. Captilism and a Market economy are a good thing. They just need to be balanced with human rights. Everyone deserves the right to do the best for themselves and their family.
The basic theories of economics still apply. Nedessary skills are still necessary skills. For worker 'A' to perform job type 'B', he will still need the a minimum set of skills to perform adequitely. For your average software 'engineer' to acquire a skill set he needs either personal capital (ie. a nice home - setup,books etc) or professional experience (ie. those mainframes at work). Now most of the students from India, Nepal and other 'third world' nationals could not afford a nice net connection and tons of recent books. They only picked up skills after being in the US for some time. I on the other hand, have owned many computers, worked for companies that exposed me to a great many different proprietary products, and collect documentation religiously. This has raised my skillset. The H1B workers often face a great deal of difficulty since their skillset is often initially limitted to what they learned at an American university which while good may not be as full as it otherwise could be. In time they will catch up but initially they are not as well situated skill per dollar.
Now, in normal circumstances, a higher skillset may demand higher wages. In the land of Software Engineering this is not entirely true. First we must understand that a Software Engineer is not really an Engineer at all. Engineering in the stricted sense of the word is a requlated and licensed proffession. For example, I may be able to out doctor the best doctor in the world and make 205K+. However, doing so would have legal consquences. This is why hospitals haven't started their own H1B push and most doctors are in private practice and can bill directly. How many unemployed doctors do you know.
The lack of requlation/licensing in the software industry is the reason for the downward wage pressure in the IT industry. Corporations can easily sacrifice quality in the name of time to market with little legal/regulatory consquence. If a REAL engineer consistantly builds bad bridge he may be held liable. Not so true of Software Engineers.
On a personal note, I am leaving my current possition for just the same reason. Time to market and poor design are emphasised. They hired seven very bright H1B's have been hired. But their skillset is not yet up to snuff. Yet they are still put on critical projects and taught to emphasize time to market over all else. My last project could have been modular and virtually bug free if I had one more week's development time. The H1B workers that I spoke with would like to do the same but are not permitted to because of their workers.
Now, my opinion on the matter is that H1B workers should be allowed to work in the very least in same manner as green card workers and be put on a US citzenship track. I don't mind competing as long as it is not against indentured servents.
And besides, a great many many of the H1B workers that I know are great guys/gals and would make great additions to a nation that was built on immigrants(at the expense of the natives). They don't deserve to have their bargaining potential deminished by limitations placed on their workplace modility.
My original posting was written just befor I left for work and I was in a rush. But it did get my point accross.
My point was that niether side of the issue is absolute. I did make the point that coders who document and design their project on their own are often more valuable than managers who blindly follow a "religious process".
Some of the biggest messes I have ever seen were made by "midnight coders" who whipped up non-manageable code. They were often valued for their productivity even though their code required constant patching. These were the quys who got things done.
What is needed is a sound thought process with out regard to personal desire or personal ego. Only when crap is called crap are both management and software guys doing themselves service.
I don't hate management. In fact, the manager I admired the most started off as an accountant. When he thought I was doing a crappy job, he walked up to me and told me why and how I should fix it. Unfortunately, the current trend seems to be toward blind productivity:
...."If it works today... fix it later"
This is the wrong attitude and these methodologies prescribe the opposite but they are implemented too often in environments were "fix it later" has already taken its toll. How often have you seen a software shop toss out junk code to get things right?
It is a sociological problem rather than a management problem. One of the reasons Dilbert is such a clever commentary is that it comments on the sociological problems that have saddly grown common.
kafka93 wrote:
While you might conceivably have some valid points, the fact that you're apparently unable to express yourself in a coherent way suggests that, by-and-large, you are wrong.
No, I wrote the article in a rush.
The paranoid, who cannot write decent documentation or do the things that should be done as they should be done will always cry foul: since you're unable to follow the processes of careful engineering, they [ we all make mistakes! ] must be a cynical method by which to protect one's own job.
I have no idea about what you are trying to say.
My point is that (poor work ethic + ass kissing) = the death of any project. Quality review is the essential element to the success of any project. If the organization has the gumption to kill bad work before it becomes entrenched, the code fumblers and ass kissers will fall by the wayside in order to protect business interests.
And for the record.... I have worked with some great consultants and managers. I have also seen a great many that could kill the best shops.
Some braindamaged programmers cause as much or more damage as managers. OH!, some of the crap code I have seen. CMM in particular adresses some of these issues forcing a "design" phase. However, the f'ed up programmer will just screw this phase up.
This makes things worse for those of us that do the work. We have to fix flawed documentation and code.
So the correct chain is:
1. someone who is well intentioned creates a methodology.
2. MBA or consultant screws it up.
3. Code monkey screws up implementation.
4. One of the few guys that does the work fixes it until he/she get pissed off and go somewhere else to do the same thing.
Also, I think we might work together. I've read Dilbert and I know I work with these guys. I think I will get all buff and become a male exotic dancer. At least that field is honest and pays well.
Management methodologies can be good when omplemented with computer science taken into mind. I think the primary reason these methodologies where developed was to combat incompetence and job security. I think all of us have seen these:
* obfuscating code in an attempt for job security
* software design on the fly that kills any chance of future enhancements and optimizations.
* any combination of the top two with worthless documentation to enhance job security.
* C++ code that uses classes in a way that is utterly insane. There are no damn interfaces and every thing is publicly accessable.
Management methodologies can help to combat this kind of thing but I have very often seen them create there own problems. For example, I have seen a "cmm 4 certified" shop rewrite/wrapper OS services to fix problems at the application level. Just throw the POS away and write it correctly. All to often, emphasis is placed on process and method rather than algorithm design, use of datastructures and eficiency. These are the things that aren't given decent metrics and are only discoverable by a well trained computer scientist or mathematician.
Personaly, I would hire a bunch of people that new there stuff CS wise and tried to document and design before I would take a "process" junky. A good size subgroup of MBA types care more about power (Job Security) than well thought out design. Time to market matters more than design. So a lot of extra people that have nothing to do with design are hired to document, the documents become worthless and the designers/programmers are still the only people who really know the code. But the manager doesn't care sinse more people equals more power.
OR TO SOME IT UP READ THE DILBERT BOOKS AND SEE WHAT THEY HAVE TO SAY ON MANAGEMENT AND BUSINESS PLANS. IT IS THE SAME DEAL!!!!
Slashdot Mirror
Now wait a few years and you will see some interesting things. We already see the reactions of businesss to the increased communication abilities given by the internet. Non-Compete and Non-Disclosure agreement are in part a result of workers abilities to shop the world for employment opportunities. "Copyright protection" is the result of content providers inability to accept the fact the instant content access dilutes the value of a single work. How much is one work worth when thousands can be accessed in seconds. Increased competitition lowers prices.
Add DNA research, advance transportation techniques, advances in particle physics ...etc into the equation and we haven't seen anything yet!!!
Now the corporate market is another question. I am not talking about desktops but small-scale (intel) server roleouts. In another life, I worked for a company that had a couple thousand rackmounts. They did the following:
1.They quite often used NFS, rcp, ghost and just about anything else they could to keep these things running.
2.To make matters worse, many of these machines would pull info from RS6000/Sun/HP/SGI machines.
3. Some rackmounts used strifing or mirroring.
I can guarantee that the people that run things at this place won't even pay attention until problems crop up. That's when things will get interesting.
Merry Christmass!!!
How many lawyers do you know that design electronic coponents? They live in a legal not physical world as do most of the execs that decide what to do with these cases. Lisence and let the other guy spend the legal fees is the usual course of action. Later they can bitch that the competition was not nice enough to fight the claim.
500+ Mhz PIII/Athalon 128+ MB memory
IMHO, I think this has a lot to do with Intels recent troubles. Intel has had production problems and two recalls in recent memory. No matter what you think of AMD, they don't have the capacity to supply the market on their own. However, AMD does represent an alternative source of parts and slow adaption of AMD parts may have cut into many PC makers earnings and time to market. Wait until Intel get their act in order, we may see a quick upswing.
In general, most analysts think that what the big boys are doing now is indicative of the future. Give Wintel a six or seven months and we will see another upgrade cycle.
Most H1B workers come from countries where it would be difficult to demand commenserate salaries. So they are mostly happy with what they get EVEN though a FREE MARKET would give them 15-30 percent more in real earnings.
So, in the end the employers are the only ones with no downside. The extra legal fees are easily offset by the guaranteed retention.
I am not saying that you should not be working under an H1B. I think you should have been given a green card so that you could freely participate in the job market. Those who can not are prisoners or indentured servents.
I just don't aprove of the hinderance of the free market. Free Market means more than just corprate rights. Captilism and a Market economy are a good thing. They just need to be balanced with human rights. Everyone deserves the right to do the best for themselves and their family.
The basic theories of economics still apply. Nedessary skills are still necessary skills. For worker 'A' to perform job type 'B', he will still need the a minimum set of skills to perform adequitely. For your average software 'engineer' to acquire a skill set he needs either personal capital (ie. a nice home - setup,books etc) or professional experience (ie. those mainframes at work). Now most of the students from India, Nepal and other 'third world' nationals could not afford a nice net connection and tons of recent books. They only picked up skills after being in the US for some time. I on the other hand, have owned many computers, worked for companies that exposed me to a great many different proprietary products, and collect documentation religiously. This has raised my skillset. The H1B workers often face a great deal of difficulty since their skillset is often initially limitted to what they learned at an American university which while good may not be as full as it otherwise could be. In time they will catch up but initially they are not as well situated skill per dollar.
Now, in normal circumstances, a higher skillset may demand higher wages. In the land of Software Engineering this is not entirely true. First we must understand that a Software Engineer is not really an Engineer at all. Engineering in the stricted sense of the word is a requlated and licensed proffession. For example, I may be able to out doctor the best doctor in the world and make 205K+. However, doing so would have legal consquences. This is why hospitals haven't started their own H1B push and most doctors are in private practice and can bill directly. How many unemployed doctors do you know.
The lack of requlation/licensing in the software industry is the reason for the downward wage pressure in the IT industry. Corporations can easily sacrifice quality in the name of time to market with little legal/regulatory consquence. If a REAL engineer consistantly builds bad bridge he may be held liable. Not so true of Software Engineers.
On a personal note, I am leaving my current possition for just the same reason. Time to market and poor design are emphasised. They hired seven very bright H1B's have been hired. But their skillset is not yet up to snuff. Yet they are still put on critical projects and taught to emphasize time to market over all else. My last project could have been modular and virtually bug free if I had one more week's development time. The H1B workers that I spoke with would like to do the same but are not permitted to because of their workers.
Now, my opinion on the matter is that H1B workers should be allowed to work in the very least in same manner as green card workers and be put on a US citzenship track. I don't mind competing as long as it is not against indentured servents.
And besides, a great many many of the H1B workers that I know are great guys/gals and would make great additions to a nation that was built on immigrants(at the expense of the natives). They don't deserve to have their bargaining potential deminished by limitations placed on their workplace modility.
My point was that niether side of the issue is absolute. I did make the point that coders who document and design their project on their own are often more valuable than managers who blindly follow a "religious process".
Some of the biggest messes I have ever seen were made by "midnight coders" who whipped up non-manageable code. They were often valued for their productivity even though their code required constant patching. These were the quys who got things done.
What is needed is a sound thought process with out regard to personal desire or personal ego. Only when crap is called crap are both management and software guys doing themselves service.
I don't hate management. In fact, the manager I admired the most started off as an accountant. When he thought I was doing a crappy job, he walked up to me and told me why and how I should fix it. Unfortunately, the current trend seems to be toward blind productivity:
This is the wrong attitude and these methodologies prescribe the opposite but they are implemented too often in environments were "fix it later" has already taken its toll. How often have you seen a software shop toss out junk code to get things right?
It is a sociological problem rather than a management problem. One of the reasons Dilbert is such a clever commentary is that it comments on the sociological problems that have saddly grown common.
Peace.. and Happy New Years....
No, I wrote the article in a rush.
The paranoid, who cannot write decent documentation or do the things that should be done as they should be done will always cry foul: since you're unable to follow the processes of careful engineering, they [ we all make mistakes! ] must be a cynical method by which to protect one's own job.
I have no idea about what you are trying to say.
My point is that (poor work ethic + ass kissing) = the death of any project. Quality review is the essential element to the success of any project. If the organization has the gumption to kill bad work before it becomes entrenched, the code fumblers and ass kissers will fall by the wayside in order to protect business interests.
And for the record.... I have worked with some great consultants and managers. I have also seen a great many that could kill the best shops.
This makes things worse for those of us that do the work. We have to fix flawed documentation and code.
So the correct chain is:
1. someone who is well intentioned creates a methodology.
2. MBA or consultant screws it up.
3. Code monkey screws up implementation.
4. One of the few guys that does the work fixes it until he/she get pissed off and go somewhere else to do the same thing.
Also, I think we might work together. I've read Dilbert and I know I work with these guys. I think I will get all buff and become a male exotic dancer. At least that field is honest and pays well.
* obfuscating code in an attempt for job security
* software design on the fly that kills any chance of future enhancements and optimizations.
* any combination of the top two with worthless documentation to enhance job security.
* C++ code that uses classes in a way that is utterly insane. There are no damn interfaces and every thing is publicly accessable.
Management methodologies can help to combat this kind of thing but I have very often seen them create there own problems. For example, I have seen a "cmm 4 certified" shop rewrite/wrapper OS services to fix problems at the application level. Just throw the POS away and write it correctly. All to often, emphasis is placed on process and method rather than algorithm design, use of datastructures and eficiency. These are the things that aren't given decent metrics and are only discoverable by a well trained computer scientist or mathematician.
Personaly, I would hire a bunch of people that new there stuff CS wise and tried to document and design before I would take a "process" junky. A good size subgroup of MBA types care more about power (Job Security) than well thought out design. Time to market matters more than design. So a lot of extra people that have nothing to do with design are hired to document, the documents become worthless and the designers/programmers are still the only people who really know the code. But the manager doesn't care sinse more people equals more power.
OR TO SOME IT UP READ THE DILBERT BOOKS AND SEE WHAT THEY HAVE TO SAY ON MANAGEMENT AND BUSINESS PLANS. IT IS THE SAME DEAL!!!!