trb (06/07/82)
I decided to respond to eagle!cw's mild tirade against the honor of hackers in this space. He responded here, and I think my comments might be of some interest to more than a few, so here I go... (The alternative is to solicit responses for the creation of a net.floyd!trb-flame newsgroup, but I'm too lazy; all hackers are lazy, that's why we only eat ring-dings and never shower.) I found some items of extreme interest in eagle!cw's comments about hackers; here's the first: I mentioned that some self-described hackers are probably too modest for their own good. I think that the observer here is at fault. Hackers as a group (in my experience - that goes for this whole note) tend to be pretty rational types. Modesty, or lack thereof, is what we hackers like to call syntactic sugar (you non-hackers can read that as "insignificant, vestigial stuff, not worth dealing with"). In our quest for the good and beautiful we eschew syntactic sugar. The concept of modesty (or lack thereof) goes the way of all syntactic sugar, that is, it gets gc'd (that's garbage collected, "thrown away" for the rest of you) because it performs no useful function, it just obfuscates (shouldn't obfuscate have gone with eschew?). You seem to think that hackers don't think systems through cleanly, that they program with axes. On the contrary, I have found that the folks that I respect enough to call hackers are quite good at design; they've used enough bad system to know what not to do. Here's another comment that struck my fancy: I was also struck by the definition of hacker as one who patches begin ... end; until it works. This technique reminds me very much of a technique called "top-down programming," one which is highly touted by the most conservative of programming managers, and a technique which (in proper doses) definitely has its place in our craft. You wanted to hear about my working methods, hmmm. I think what makes me a hacker is the skill I have derived as a result of my devout interest in timesharing systems (TOPS-10 and UNIX for me). As an example, a friend of mine was noodling around reading his UNIX manual and he asked me "What is the key used in the example on the crypt(1) man page?" I immediately replied "znorkle" (this was the PWB 1.0 manual, "znorkle" was changed to "key" for 2.0). He was only semi-surprised. The point here isn't that I had recalled such a bit of trivia, but rather that I had a pretty thorough familiarity of the manual because I had a pretty thorough interest in my tools. Not different from a guy who has a thorough interest in hacking cars, and I think that both of us should be respected and valued for our dedication and knowledge, even though our hands might be dirty and our hair might be long. Hackers make a pretty interesting topic of coversation, better, at least, than a hundred people posting to net.nlang to see if anyone wants to start up net.nlang.urdu and then realizing that they really have nothing to say. I have lots to say about this; I post this as a response to a request. If any of you are really interested in continuing this discussion, it might get long(er) winded, and I would think that a separate newsgroup or private communication would be more fitting. Andy Tannenbaum Bell Labs Whippany, NJ (201) 386-6491