TO: Libertarian National Committee
From: John Famularo Chairman, Computer oversight Committee
Date: 22 April 1989
Subj: Initial Review of the LNHQ Computer system
The system is currently operating under an emergency interim procedure until a minimum set of procedures can be documented and installed to the satisfaction of the National Director and the LNC Chairman. During the interim procedure Dean Ahmad is analyzing and documenting the entire membership system and simplifying and/or condensing some of the many programs that comprise the membership system. John Famularo is assisting in 'he handling of day-to-day operational problems, while Dean Ahmad concentrates on the clean-up.
The networking hardware and software has been physically installed but will not be completely implemented until the clean-up is completed. The scheduled completion date for the clean-up is May 10th.
After the clean-up we will update the office procedure manual developed by Don Ernsberger to reflect any operational changes created through clean-up programming changes. We will then expand the system documentation by cross-referencing the program documentation with the office procedure manual. We will identify all the data files and their backup and restore procedures.
When we have an acceptable minimum system we will call that system release 1.0 of the Membership system. Any modification to the system thereafter will have to follow the following general procedure:
- Written System Request for a System Change
- Review by National Director set priorities
- Review by Computer Committee document changes
- Cost Benefit analysis - estimate man-hours/elapsed time
- Develop - unit test
- Systems Test
- Acceptance test
In order to follow the above procedure, a new computer environment must be established that has a test environment separate from the production environment. We will call this version 1.0 of the LNHQ computer environment. It will take a few weeks to a month to set this up and test it. In order to properly test the test environment we must construct a set of test data transactions for all system functions.
The next version of the Membership system will be version 1.1, which will have to be specifically defined after version 1.0 is fully documented, will generally consist of adding concurrent network access to certain transactions, and program controlled printer spooling of various reports.
In my opinion the sequence of events from approximately January 1988, which leads to, the current systems crisis is not due to any lack of programming skill or dedication to the project on the part of any individual involved in the LNHQ system. Rapid turnover of personnel, exacerbated by the lack of systems or program documentation, further exacerbated by an over optimistic estimation of the man hour requirements of system enhancements that were not properly cost justified, combined to make the recovery task disproportionate in comparison to the original requirements. If Alan Lindsey, Kirk McKee, or Dean Ahmad were not the exceptional programmers that they are, there probably would have been a complete systems collapse many months ago. We are very lucky that these people were available when they were. However, in the future we must adjust the system so that it is not dependent on one or L small number of individuals of exceptional capabilities. The underlying true cost of maintaining such a system is extremely high. This high cost is masked by individuals volunteering their time for free or at reduced cost. If the pool of these volunteers is small, we run the risk of not having any one being available in an emergency. It is therefore necessary to have the largest pool of zero or low cost sources of systems and programming talent available. The most simple system that performs the required functions that is also fully documented is the best system for our current environment.
Until other arrangements are made, there is no budget or staff at the LNHQ to perform systems and programming development. In the interim, any member of the computer oversight committee who desires to assist in programming systems changes or testing other members changes will be given a copy of the test environment that they can install on their own equipment. This procedure should greatly expand the number of people who could be of emergency assistance to the LNHQ staff in the case of production emergencies. They would also be in a better position to contribute to the analysis and evaluation of the current documentation and any proposed system changes.
A more detailed report on the adequacy of the current computer environment will not be possible until after the completion of phase I which consists of the following:
- Version 1.0 of the membership system.
- Version 1.0 of the LNHQ systems environment
- Documentation of Current pending systems requests
- General System requirements medium - Long range
Finally, the transition from an unstructured to highly structured business environment will take the cooperation of everyone involved in interfacing with the LNHQ. No modifications of any kind should be made to any LNHQ procedure without the written approval from the National Director. This includes even touching either of the computer terminals. I am also suggesting that all changes of recorded data on the computer system be performed by a member of the LNHQ staff or if that is not possible or practical that a member of the staff be present whenever any changes are made. Many of us will get the urge to bypass these procedures because we will feel that "just thi-s 'one time can't hurt", but any situation that could potentially cause the LNHQ staff to lose time due a lack of knowledge of the true operating environment will be multiplied in its effect in time and money as an ever widening circle of individuals are affected.
If you know of anyone who would be interested in serving an the computer oversight committee, please contact me. Specific programming or hardware skills are not required.
John D. Famularo
1420 Locust St., Apt 36-D
Philadelphia, PA 19102-4221