Slashdot Mirror
Can The eXperimental Computing Club Survive?
Logic Bomb writes: "Salon has an article about the famous Berkely eXperimental Computing Club, from whence came GTK, the GIMP, early web browsers, and many other goodies. The article has a nice overview of the club's history, as well as questions about its continued existence. Apparently the rise of collaboration over the Internet has made it much harder to find recruits."
I don't know anything about XCC, but maybe I can shed a little light on the issue by telling you what's going on at my own university, up here in Canada.
Currently, at Carleton University, there exists a club with a similar purpose called Nexus. It was intended as a student-run research group, whereby students could gain access to the equipment they needed for certain projects, and have a better way of finding other people to join teams to work on said projects.
It's in danger of dying out this year.
What's killing it? I would suggest that it's the computer science department itself. In the last few years, the faculty of the computer science school at Carleton University has increased the workload of students exponentially. At the current state, we have the following "killer" courses, to weed out people who can't program:
204 - C++, the profs require you to use Visual C++, with the intent of forcing you to recognize all of the little things that are wrong with Microsoft's compiler. (Their words, not mine.) One of the killer assignments includes mixing operator overloading, templates, and exceptions. This is turning into a really frustrating course for a lot of people, because it's supposed to be an intro to C++ course. It's turning into a fight-the-compiler course.
304 - Object Oriented Software Engineering. You are forced to use ObjecTime, which is the most crash-prone, unuseable, finicky, flaky CASE tool I have ever seen. This is also a group work class, which only compounds the problems for a lot of people. Generally, the class is very beneficial, but with the amount of work and spare time it takes, and the required use of ObjecTime it definitely qualifies as killer.
384 - Algorithms. I agree that this should be a "killer" course. But there is a lot that isn't taught in this course that should be, simply because we're TESTED on it afterwards. They use the CLR book, too, which means that a lot of your own questions are never answered by the book, as there is no solution guide available to students. (And the authors don't want this to change.)
484 - Algorithms II. This is the course that's supposed to be THE killer. And from what I've heard, it is. (I'm only in third year right now.)
In addition to these required, core courses, we're also required to take several other courses in third year and above. Our options can be considered killer, too, as some of them (307, which is a Scheme/Prolog course, and requires as one assignment that you reimplement scheme within scheme; 302, a compiler course that I've also heard was a real time consumer) are just plain insane. Especially whenever you consider that each of these students is potentially taking a full course load, involving at least four other courses. (I've heard many a story of a student dropping all of their other courses just so they could get one of the above-mentioned courses enough of their time.)
It really doesn't help that we've got some really really crappy TAs at times. I've had friends in 204, Intro to C++, show me assignments they've gotten back with "WOULD NOT COMPILE" written on it. The reasoning of the TAs was as follows; I put the disk in the drive, I opened the project, and I hit compile. It stopped compiling, so it mustn't work. (In this case, it was because the TAs were compiling it on the disk. Visual C++ requires a good 10 megs swap space to compile anything, after all.) I really pity the people handing in this last assignment, which is the above-mentioned killer assignment for that course. The only TA is flying back to China as soon as he's finished marking them, so students will have absolutely no recourse if it doesn't run. (The new C++ prof absolutely refuses to remark assignments.)
(Part of the problem, I think, is that a lot of the staff is just plain bad at teaching. The currently reigning C++ prof has a fetish for overloading every single operator. Last year's operating systems prof didn't know how to use fork() or IPC. Students have to make up for their profs' bad teaching by learning the material on their own time, which only makes things worse.)
This is what's killing the computer science societies at Carleton. There are no longer people with free time available to volunteer to run the offices of Nexus. There are rumors circulating already about more general services like CCSS shutting down, and the problem's only going to get worse. With CCSS and Nexus, it's becoming the case where the only people who actually have the time to run the services are first-year people. And now we're starting to find that we can't even attract their help. (Recent graduates are not allowed to help, thanks to some rather bizarre restrictions on clubs and societies, so this is what we're stuck with.)
If this keeps up, we really are going to lose all of our societies.
It would be a shame to say the club is dying. In the traditional sence yes, but as was said it is due to collaboration over the internet. So instead of saying the club and it's purpose is dying, let's say anyone can be a member now, if not of the club of it's purpose.
As x approaches total apathy I couldn't care less.
Now though, I find that there are not nearly as many computing clubs to be found, at least where I live. Has the Internet sounded the death knell for computer clubs?
It would seem that old-fashioned face-to-face contact is somehow becoming unfashionable. The Internet does not provide an adequate replacement for physical socialising. I also suspect that the quality of work that gets done over the internet is of an inferior quality. This is to be expected though, given the lack of acceptable criticism that the Internet engenders. Criticism is usually dismissed as 'flaming' and is ignored. In real life such behaviour would be more muted, and would be responded to.
Hopefully, one day, people will realise that the Internet is not a panacea for all of societys Ills, and is not a replacement for society. I hope to see the rise of computer clubs again one day.
KTB:Lover, Poet, Artiste, Aesthete, Programmer.
KTB:Lover, Poet, Artiste, Aesthete, Programmer.
There is no
I believe that the problems with recruitment are not due to collaboration between people over the Internet, or even to freshman coming in with a hacker support community already formed on the net. The bigger issue is that being a geek is not such a socially unacceptable flaw anymore, especially around MIT.
Two things used to drive people to the SIPB: Access to good hardware and people with whom to geek. Both of these encouragements are starting to slip away. And the social barrier to entering SIPB has always been high, due in large part to elitism among its members and the perception of elitism by the community.
10 years ago undergraduates couldn't run Unix in their dorm room, or even have networking. If they wanted to hack they had to come to SIPB. The machines in the office used to be a big draw. That is not so anymore. Unix workstations that we can afford are comparable to freshman's new Intel machine. Our fastest workstations are rivalled by my $800 Linux box. And providing even faster machines would not really help. I don't need an Ultra60 to hack on.
The draw of people to geek with is still strong though. That is what keeps me and others coming back to the SIPB office. Between current members and the cruft (members who are no longer students) who hang around there is an alarmingly high density of real world computer experience. Its a great place to go to have people tell you your new design is crap or your idea for a company will fail. But more and more students never feel that attraction, never feel the need to hang out with other geeks.
In some way this is because of the net, and the fact that these students come in with support networks. But even more than that it is the omniprecense of geeks at MIT. People who need their fix of geeky social interaction in real space can usually get it in their living group. The odds are there are a couple of people who run Linux and understand when you talk about your projects. They aren't very useful when it comes to critique, but you can get expertise online. People who would have become SIPB addicts^Wmembers in the past now get by on geek methadone, amply available in their current social groups.
Its a difficult problem to solve. I don't have a solution, or I would have deployed it at MIT. Sometimes we think that people don't know about us. We have talked about Slashdot (a common brand of methadone) banner ads targetted to net-18. But really the problem is harder than telling people we exist. It is to convince people that our membering process is worth the time and effort, and that we are a social group they want to join.
We are not dying. We do not have significantly fewer members then we have had in the past. But people no longer seek us out. Many of our members finally get over their inneria and join when their existing social scene flies apart. Others (like me) get dragged in by people in their living group and eventually get members by virtue of having sat around the office so much. But this year we didn't get any freshman. They don't realize that they need us, yet.