Linux/open software (was Bill Gates writes...)

Les Schaffer godzilla at netmeg.net
Fri Aug 7 08:50:09 PDT 1998



>>>>> ">" == Max Sawicky <sawicky at epinet.org> writes:

>> More seriously, I've been mulling over for some time now the

>> idea of trying to start a campaign for left-leaning non-profits

>> to convert to Linux.

nice idea.

>> I'm sitting here looking at my two Red Hat CD-ROMS, trying to

>> get up the gumption to *try* to install them on my work machine

>> without immobilizing myself for a week.

"just do it". we're here for you...

>> The migration of important institutions (e.g., the labor

>> movement) would contribute to a critical mass of users which

>> could cause significant leaps in the development of Linux-world

>> software.

do you mean here just the sheer number of additional users would cause significant development, or that the labor movement would have specific software needs that would lead to further development?

>> I'm strictly an observer in the computer biz, so I'm interested

>> in what others more closely involved think.

Personally, i have been using Linux for 3.5 years now for all my scientific and engineering simulations and reports. Most all of the science and engineering departments at universities that i am familiar with have switched or are switching to Linux based workstations. Ikea, i am told, is all Linux, at least for its world-wide communications system. There are lots of __good__ places to go on the web for reading about all this kind of propaganda, if you're interested.

Having used HP Unix, SunOS and Solaris for over 10 years in research and coroprate settings (ie, where there was always a dedicated system administrator around) i can say easily that the quality of the work experience in Linux is much higher for several reasons:

1.) bugs get fixed pretty damn fast. there is little or no face saving (we'll fix it quietly and sell 'em a windows 98 within a year)

2.) friendly help is near at hand. Sys admins at first were reluctant about Linux (memo to Chuck Grimes: okay, freeBSD too) because they thought support was an issue. But they were just too used to calling HP or Sun on the phone (and waiting) for their problems. Now, everyone using Linux|BSD know that all they have to do is post a message on USENET group, and generally, 95% of your problems get a solution within one day (max). In fact, its quite amazing. when i run into a problem, i dont even post for a day, cause i just know someone else out there in the world has likely run into the same thing. and sure enough, withinin a day i find a post which is relevant. if not, i post.

OR, you find out that a problem you are having is a deep issue that will take time to resolve, and you can follow the open discussions on its solution over time.

3.) Source code availability. Sometimes just a little tweak here and there gives your system that well-tailored look-and-feel, adaptable to your circumstances, under the care and guidance of a world-wide group of programmers interested in tweaks and twists.

3a.) Control of future releases: if you want to see code move in a certain direction, say in the kernel, you post your arguments to the linux-kernel development list, and organize people around your ideas. If you convince the gatekeeper (Linux Torvalds) that the code is good and belongs in the kernel, you're ideas are in, and get built upon by others.

Under certain circustances this leads to loud fights: but in ful public view so that a reasonably intelligent sys admin can follow the debates and learn what the issues are, which all in all leads to an incredibly educated mass of people working together.

4.) non-kernel applications: okay, you all know the arguments that a large part of coimmercial software has not been ported to Linux|BSD. but that is changing, at first slowly, then more quickly. At my office in CT, i have gotten da bosses to seriously consider putting Linux on their next machine if, by the time they make their next purchase, circuit analysis, finite element, and CAD programs are available at reasonable prices. I have trained these guys like obedient dogs to curse Bill Gate's every time they have to call me in to put out a fire on one of their Windows systems. They recognize now the difference between a pretty face and phone number (Windows GUI, MicroSoft support) and what it takes to do day-to-day pounding on the keyboard. I have taught them well, because they watch me, on the phone, winding up telling the apps tech support person what is what, and they realize that "Support" is more window dressing than anything substantial.

enough for now. back to work.

les

p.s.

Okay, one last thing re/ computers, and this connects with some of Chuck Grime's agit-prop, but is only tangentially associated iwth this discussion on Linux:

One of the things i see happening in the engineering community over the last, oh, 5 years, is the developing notion that one can just buy a pretty piece of software and magically have a new capability in your firm. I am talking mainly in small to medium sizes companies. What is happening is bizarre: as an engineering consultant, i am finding more and more businesses that THINK they are solving problems with expensive (mainly Windows) software, but have not invested in the __labour__ (people) and __training__ to use these tools as they were designed to be used either maximally (eg. CAD) or within the limits and understanding of their inner workings (finite element software for stress analysis, fluid flow simulation, etc). What in fact is happening is, these companies think they are saving money in personell (either in the engineering staff or in the computer admin staff) because the software looks self-contained and manageable by an (relatively) untrained person ..... but this is an illusion which causes more harm than good for the productive efforts of the company. yes, another of the contradictions.

(I could tell ya stories.... )

-- ____ Les Schaffer godzilla at netmeg.net ___| --->> Engineering R&D <<--- Theoretical & Applied Mechanics | Designspring, Inc. Center for Radiophysics & Space Research | Westport, CT USA Cornell Univ. schaffer at tam.cornell.edu | les at designspring.com



More information about the lbo-talk mailing list