At 04:39 PM 6/15/98 -0400, Lou Proyect wrote: >The main thing that prevents this is that the portion of the project life >cycle that precedes the actual programming and testing requires heavy >involvement with end-users and knowledge of their business. Outsourcing >works best for highly technical applications whose requirements are fairly >straightforward. For example, build software that will serve as a bridge >between mainframe and Unix applications. Where things get more complicated >is, for example, an application for online student registration. At >Columbia, this involves a lot of give-and-take, which works best in on-site >discussions. Thanks for that information. As I understand it, custom software requires on-site specialist. However, with the Braverman's argument about 'degradation of labor" in mind, that brings further questions. How likely is a modular software design that paralles the assembly line. Modules designed to accomplish particular tasks are develeped by an army of relatively low skilled low paid programmers. Then, a single specialist assembles a custom-fit application from a great number of moduels available to him or her. The reason I am asking is that so far, Braverman's argument seems to hold in every field, including health care - where a great number of tasks is routinized and performed by relatively low skilled and paid employees rather than physicians. In some HMOs you may not even see a doctor, if you so chose. So if every kind of work faces de-skilling, as Braverman argued, that would also hold for computer programming, no? And if that is the case, then the restrictions Lou mentions are only temporary, right? One implication is tha if the proletarianization of computer professions is imminent, those highly paid boys and girls now infatuated with high-speed capitalism may soon wake up in a gutter. And that is an interesting prospect for left organizing, isn't it? Regards, Wojtek
