Slashdot Mirror

Abstract: A proposal to create a GPL'd curriculum for training new Linux users and system administrators. Discussion on creating local "campuses" of such a university.

Introduction: There has been a ground-swell of publicity for Linux in the press, starting with the Netscape Open Source move earlier this year. The technical merits of Linux are most universally praised, but there is criticism over the user interface, and ease of use. In many corporate environments, the 'Total cost of ownership' for Microsoft products is thought to be lower, due to the difficulty and expense in finding qualified Linux/Unix system administrators.

While on one hand, this helps the Linux community by keeping salaries high and job opportunities plentiful, we must be aware that if the situation continues this way, Linux will not be successful in the corporate environment. Therefore, it is in our best interest to train new users and administrators.

One of the things that the corporate world finds important in training programs is a uniform rule by which to measure success. Tests such as the Novell CNE, Microsoft MCSE, etc. are well regarded because of the certain uniformity of knowledge expected of those who past the test. While few employers will hire based solely on a test score, such a certification can help make the final decision between two candidates, and is easily justifiable to more senior management.

It is easy enough for a reasonably well versed Linux user to teach a class of windows users a little bit about how to deal with Linux, I feel that the advantages in developing a standard curriculum and standard test are great enough to put forth this proposal and solicit comments. Below are my thoughts, and I welcome comments and criticism.

1) users are coming from different backgrounds, and separate tracks are needed (separate courses) to address this, not boring someone who already knows about TCP/IP, while not leaving a newbie in the dust. How is this done? I would suggest an entrance placement test. From this point, a the composition of the class can be developed. It may turn out mostly MCSE's who want to add Linux to their resume, it could be recycled VMS admins, or it could be a bunch of know-it-all high school kids who want to have some paper proving that they are knowledgeable. Based upon the results of the test, certain areas that all students are familiar with can be safely skipped, and at other times students can be broken into focus groups to work on that in which they are weak.

2) I believe that GPL'd training materials are useful. I professor would not be required to use all, or any specific work, but rather be encouraged to choose the best out of what is available. However, for the sake of uniformity, a standardized pool of questions, from which the certification test will be made, needs to be agreed upon so that certification is uniform.

3) Actually programming and operation should be encouraged, even required. However, the final certification should not rest on the grading of assignments, for the sake of standardization. Another grade, that of classroom participation, may be given in addition to the outcome of the test, but in no way should such a grade enter into the final certification score.

4) While some of us may be tempted, out of our enthusiasm for Linux and love of teaching, to not charge any money to students or to charge too little, I would encourage anyone interested in teaching to stick to a standard and reasonable range of fees which would be formed by a rough consensus of teachers. First, many corporate types will pay for their employees without hesitation, but would question the value of education received at no charge. Second, this may end up being a lot of work, more and of a different type than some teachers will be familiar. Third, often the facilities will need to be rented or kept up. Fourth, we want to attract and maintain professors who are not only good technically, but are good and patient teachers. This requires money. Right Now, I suggest $400 for a 20 hr. class.

5) As much as I hate to start yet another, ad-hoc standards bureau, someone needs to keep track of the tests, administer them, and keep a database of who has passed and who has not. Such a central bureau should collect a small fee, maybe $20 per student, per exam.

6) Let's get some idea of who is willing to teach, has experience, and resources. In each geographical area, put all willing participants in touch. Often, such classes might be administered under the aegis of local Linux Users Group.

I live in Cambridge, Mass. (a city built on education). I am willing to talk and meet with others who are interested in forming Linux University, Cambridge Campus.

7) Let's start talking about what people need to know, and then work on how best to teach it.

Abstract: A proposal to create a GPL'd curriculum for training new Linux users and system administrators. Discussion on creating local "campuses" of such a university.

Introduction: There has been a ground-swell of publicity for Linux in the press, starting with the Netscape Open Source move earlier this year. The technical merits of Linux are most universally praised, but there is criticism over the user interface, and ease of use. In many corporate environments, the 'Total cost of ownership' for Microsoft products is thought to be lower, due to the difficulty and expense in finding qualified Linux/Unix system administrators.

While on one hand, this helps the Linux community by keeping salaries high and job opportunities plentiful, we must be aware that if the situation continues this way, Linux will not be successful in the corporate environment. Therefore, it is in our best interest to train new users and administrators.

One of the things that the corporate world finds important in training programs is a uniform rule by which to measure success. Tests such as the Novell CNE, Microsoft MCSE, etc. are well regarded because of the certain uniformity of knowledge expected of those who past the test. While few employers will hire based solely on a test score, such a certification can help make the final decision between two candidates, and is easily justifiable to more senior management.

It is easy enough for a reasonably well versed Linux user to teach a class of windows users a little bit about how to deal with Linux, I feel that the advantages in developing a standard curriculum and standard test are great enough to put forth this proposal and solicit comments. Below are my thoughts, and I welcome comments and criticism.

1) users are coming from different backgrounds, and separate tracks are needed (separate courses) to address this, not boring someone who already knows about TCP/IP, while not leaving a newbie in the dust. How is this done? I would suggest an entrance placement test. From this point, a the composition of the class can be developed. It may turn out mostly MCSE's who want to add Linux to their resume, it could be recycled VMS admins, or it could be a bunch of know-it-all high school kids who want to have some paper proving that they are knowledgeable. Based upon the results of the test, certain areas that all students are familiar with can be safely skipped, and at other times students can be broken into focus groups to work on that in which they are weak.

2) I believe that GPL'd training materials are useful. I professor would not be required to use all, or any specific work, but rather be encouraged to choose the best out of what is available. However, for the sake of uniformity, a standardized pool of questions, from which the certification test will be made, needs to be agreed upon so that certification is uniform.

3) Actually programming and operation should be encouraged, even required. However, the final certification should not rest on the grading of assignments, for the sake of standardization. Another grade, that of classroom participation, may be given in addition to the outcome of the test, but in no way should such a grade enter into the final certification score.

4) While some of us may be tempted, out of our enthusiasm for Linux and love of teaching, to not charge any money to students or to charge too little, I would encourage anyone interested in teaching to stick to a standard and reasonable range of fees which would be formed by a rough consensus of teachers. First, many corporate types will pay for their employees without hesitation, but would question the value of education received at no charge. Second, this may end up being a lot of work, more and of a different type than some teachers will be familiar. Third, often the facilities will need to be rented or kept up. Fourth, we want to attract and maintain professors who are not only good technically, but are good and patient teachers. This requires money. Right Now, I suggest $400 for a 20 hr. class.

5) As much as I hate to start yet another, ad-hoc standards bureau, someone needs to keep track of the tests, administer them, and keep a database of who has passed and who has not. Such a central bureau should collect a small fee, maybe $20 per student, per exam.

6) Let's get some idea of who is willing to teach, has experience, and resources. In each geographical area, put all willing participants in touch. Often, such classes might be administered under the aegis of local Linux Users Group.

I live in Cambridge, Mass. (a city built on education). I am willing to talk and meet with others who are interested in forming Linux University, Cambridge Campus.

7) Let's start talking about what people need to know, and then work on how best to teach it.