Would A Youth-Run Computer Lab Last? 30
Dragon218 writes "Recently, I have been chosen to be part of a team of computer savvy youths who have been tasked to build and maintain a computer lab. We are probably going to get around $100,000 in grant money. The idea is to have it be a public lab, the trick being the users would help in commercial Web design or other projects to raise money to keep it active. I'd like a reality check on this. Also, if any of you have ideas to help this thing along, I (don't like to use the pronoun 'we' without the group's consent_ would appreciate it. If you would like more information about the group sponsoring it, you can check out http://brycchouse.org." Assuming the people in question know what they are doing, I don't see a problem with this at all, do you? Good luck, Dragon218.
Check Conflicting aditudes (Score:1)
Go for it! (Score:1)
Limit who has Root/Admin access (Score:2)
If you limit the lab to running either Linux or Windows NT, you can limit who has access to those who can be trusted (you'll learn in a hurry). Windows 2000 is more secure by default than NT 4.0, but if you do use 4.0, grab these command scripts [umt.edu] I created in my last job. They were designed for NT 4.0 running IE 3.0, but it gives you a good idea of how to set directory and file permissions. They'll require some updating to work with NT 4.0 and IE 5.
When the lab I used to run (on a campus) was running Windows 95, I'd spend at least 8 hours a week rebuilding and reinstalling. Once we went to NT and I got secure directory permissions (secure permissions being the key), I was in that lab 1-2 hours per month.
(An no flames about using Linux. This was in pre-1.0 kernel days, and the software we had to run still isn't available on Linux. The real challenge is not to make something work, but to make what you've been stuck with work perfectly.)
You can't make money if the resource isn't available.
Public access is the key (Score:3)
I was the systems manager for a large student organization last year (no link, the website's down *sigh*). Most of my headaches came from users, and the odd things they thought of to do to my computers in their spare time. And that was with a relatively restricted user base of about 50. If you have more users which change frequently over the space of a year, I think they'll be even more trouble.
I'd say the difficult thing is not getting teenagers to properly administer and staff a computer lab, but getting teenagers to be good users. Here is where your OS can be your best friend or your worst nightmare. Maintaining a multi-user cluster of computers is much easier with a multi-user OS. Which one you pick depends on the knowledge and needs of your users, but I would highly suggest at least getting Windows NT, if not something beefier. Having a clearly defined policy on recreational use helps, too. Is web surfing okay? Would game-playing be tolerated? What about pornography (yes, it comes up!)? When users know what's expected of them, a large majority of them will try to fulfill those expectations.
Another thing which will make maintaining your computers easier is separating the duties of those involved. It sounds as if you in the sort of environment where everyone knows everyone else and they all get along. Trusting one guy to be completely in charge of purchases, for example, and another to handle installations means that your team won't be working at cross-purposes. Nothing is worse than having to spend hours of your time to fix someone else's 'fix'. Find some way of identifying what each person is responsible for and make certain they stick to it. The former because people get very unhappy when they perceive someone else is doing their job and the latter because people get very unhappy when they perceive they're doing someone else's job.
Jonathan David Pearce
RE: youth-run computer lab (Score:2)
http://www.harambee.org/ [harambee.org]
http://www.harambee.org/news/ps n_lab_earthlink.html [harambee.org]
Why not? (Score:1)
At the school I went to, the computer labs were staffed and managed all by students. The Student managers for the 11 labs reported to one of 3 staff people who were in the Information Management department. Outside of that, there were 3 or 4 student technicians who floated around to all the labs to fix various problems.
I managed for a year, and ran as a floating tech for a year, and we had very little 'interference' from our staff supervisors. THe managers even had almost-complete (read as: everything except things dealing with administration rules and such) control over things like hiring new staff people. As far as the techs, we worked out of the office with the staff people, and ocassionally had tasks prioritized by them (you _have_ to get these upgrades done before the rest of that stuff), but for all intents and purposes, all 11 of our labs (comprising something like 400 computers and workstations) were student-run.
Re:Limit who has Root/Admin access (Score:1)
Yah, weve got somthing like that (Score:1)
*Not a Sermon, Just a Thought
*/
Mouse Balls (Score:2)
I make the kid who messed things up fix it with me. They all know the policy and know that they are obligated to spend a fewe hours if they mess up. I've yet to have a major problem (that can't be fixed with a little work) and I've seen some of these kids helping each other out when encountering similar problems. Last week I had one kid install policies on his friend's machine to keep him from messing it up. LOL, exactly what I was avoiding, but it was nice to see him take the initiative and figure out how to do it. I hope that eventually I'll be able to let them run it on their own.
Re:Limit who has Root/Admin access (Score:1)
If you've got the perms set correctly when you Ghost the first one out, that's great. If you just daily re-Ghost the workstations (I know some places who have resorted to this), it's dealing with the symptoms, and not the problem.
Oh, and there's the classic problems like the one you describe.
The one thing I meant to add is that, make sure that no "user" account has admin access. Create a seperate admin account for each admin, and a seperate user account for each admin. The admin MUST use the user account for normal stuff (and problems show up more quickly), and then must change identities, or logout and back in to switch to the admin account. Helps you create an audit trail, and insulate things. Slower, a nuisance, but less easy to ruin something (like with a virus).
Ghost/tar & dd/dump & restore (Score:1)
If you make all of your machines identical the amount of problems you will have decreases dramatically. Since we've been creating machines from images rather than by hand the amount of problems we've had has decreased dramatically.
You can use either ghost for NT/95 systems or you can use tar and dd for *NIX. If you have one UNIX box you can use it to make images of HDDs, just plug them into a spare IDE port and dump from disk to image. You can also make a boot floppy and mount the images via NFS. (My personal favorite UNIX is FreeBSD which has its own stripped down version called picobsd [freebsd.org] although the best place for information is the mailing list (small@freebsd.org) or list archives. Check it out!)
--
Eric is chisled like a Greek Godess
Re:Limit who has Root/Admin access (Score:1)
At my school, I'm a student tech in the lab - unfortunately, the admins won't log in with ordinary accounts because they're doing admin stuff so often (of course, if Windows had an equivalent to "su", logging in normally as an admin wouldn't be neccessary).
Re:Go for it! (Score:1)
Be wary of turnover (Score:1)
While you are designing the lab, take into account the turnover rate that you will experience, and plan for it in advance.
Many things will cause turnover:
- Changing Schools
- Better Job Offers
- Personality conflicts
- Terminations
and if you do not have a ready pool of replacement employees, or worse, if existing staff has belittled possible replacements, your lab will dry up and wither.Frankly, if you develop a reputation as having good people, local companies will snap your best people up. Who can resist a job offer of $21.50 an hour when they are working for $6.25 an hour?
Youth run labs work... when... (Score:1)
I helped start a youth computer lab and technology training program called Youth Tech. This was a community computer lab completely constructed and maintained by myself, and about 20 other local youth.
We were also funded by a healthy grant from Weed & Seed -- no, not fertalizer :). Anyway, the whole experience was a huge success for about two years. What happened? The greedy Son of a -----s on the board of directors of the YMCA (whom hosted the program) ganked everyone's job and ran off with the newly installed computer lab (two of them) and all of the donated hardware and software. Wow! Now there's a great display of Christian Ethic (tm)
So, I don't think you are going to have great difficulty getting youth to participate in the program and stick with it. Make it worth their while. If they can get paid well (at the time I worked for Youth Tech, and was the youth president toward the end, I was making $6.25/hr, at the time it was about $2.50 above minimum wage for a 15 year old).
Anyway, I think that programs like this are a great thing. It introduced me to the hardware end of computers (I had already been programming for 4 years prior to this). If you plan well not only will you have a lot of committed youth, you will have a lot of support from the community as well! Just watch out for those greedy scums who bilk the loop-holes and take advantage of community oriented grant money. Try to maintain this lab independently of a parent organization, or become your own organization. Otherwise, the availability to be taken advantage of is ever-so-present.
Pardon the choppy grammar, but things are a bit rushed! :). Good luck and best wishes!
Getting People to Do Things (Score:3)
So far all the comments have been about the technical side of things, which looks, from here, to be pretty cut-n-dried. The real issue is the human engineering one.
This project -- any volunteer project involving more than two people -- has serious People Issues. I speak from experience: I often organize hoardes of volunteers in my spare time, sometimes even to do pro work to raise money for our organization.
First, you're talking about getting users to contribute work. Since you're talking about selling these services, they'd better be pretty professional. You're talking some pretty heavy management burden here. Getting amateurs/volunteers to put out high quality work consistently can be hard, because volunteers take (all too often) the attitude "you should be grateful I did the work at all".
Second, you're going to have to convince people that it's worth while. Note that the OS movement is powered by the "scratch my itch" dynamo. Is it anyone's itch to make corporate web pages, to keep a public lab open? That strikes me as a very hard sell. You might get someone to do it once, but will they keep doing it?
Third, speaking as someone who does web production professionally, it's a bitch. Not the technical side -- the business side. The businesses are constantly coming back with piddling copy revisions, trying to get you to do free work ("it's just a little change"). And, BTW, my agency bills ~$50/hr for my time doing DHTML, and I am considered a jr. level DHTMLer. Boston. And I don't even do design, just implementation. Just in case you were planning on charging $20/hr.
Fourth, you're talking about drawing volunteers from the public. You don't even have any filters between you and them (such as all belong to the same club or some such). You're going to have a hell of a job making character judgements of the strangers that walk in your doors, and you're going to have to make hundreds of them, constantly. Not merely "Do I trust this person not to loose the next Morris worm from my machines?" but "Do I trust this person not to flake out in the middle of doing a customer's project?"
These are the areas in which *real* head-aches arise. Basically, you need to be -- or to have to hand -- a real "people person", or more properly a "leader of (wo)men" to bear the brunt of managing volutneers.
Is something like this possible? Sure: but it completely requires some highly charismatic person to marry their souls to this project. Someone who builds group esprit, who defuses conflicts, who has a natural air of authority, who personally exemplifies the values of the organization (e.g. brilliant hacker, hard worker, tolerant with newbie users, etc.), whom people trust and follow.
These people don't grow on trees. These people have their pick of volunteer organizations to work for, not to mention their pick of 80hr/week jobs. If you are or have such a person, then your organization is unspeakably lucky.
If you don't have such a person, and you aren't prepared to try to be that person, you are going to drink the dregs of a very bitter brew. Volunteers who manage volunteers often have the worst and the most embittering cases of disillusionment and burn-out. They rail about the stupid, lazy, and vicious nature of their fellow humans, and swear never to bother trying to help someone again. They feel taken-advantage of, rattled and hurt -- and unconfident of their own abilities.
Think about the human dimensions very seriously indeed.
----------------------------------------------
Problems of student labs (Score:1)
I have worked with several High Schools and public computing facilities (mostly college level), and have learned a few tricks to making student labs work. If it is a public use lab, and Jr. High/High School students populate it, expect *AT LEAST* the following problems (it seems they are applying to college labs more and more now, too):
1) Missing mouse balls. Solution? Superglue. Sorry, can't clean the mice once you do this, but at least you don't have to buy 50,000 mouse balls.
2) Games/Porn/Surfing. Define your policy clearly. You should mention something about noise while you are at it (is it okay?). My recommendation on games/surfing is that it only be allowed during "off peak" hours; that is, no games or surfing when someone is waiting on a computer!
3) Viruses, trojans for Windoze. Have an Internet connection? Someone will install Back Oriface. Viruses will be common, too. Install good AV software.
4) Distruptive behavior. One common stage of youth learning computers involves trying to sabotage (sp?) each other's work. Read about a new Windoze denial of service? Someone is going to try it in your lab. If you run Unix, expect fork bombs. (If you run Unix, I recommend turning OFF remote logins except for Sysadmins on the workstations)
5) Piracy. Do you really think 15 high school students purchased that networked game they are playing? YOU COULD BE LIABLE, though! You have to monitor this, sadly, to reduce your liability. I would make an official policy of "no software installs" -- except possibly in each user's personal space. You can get around the requirement to audit your computers for unauthorized software by simply re-installing them every week or so (very easy to do with Ghost; simply make a special boot disk, put it in 5 computers at a time, load the image off the network and reboot them).
6) Printing! If you have free printing, you'll do a lot of printing.
7) Theft. This is the one I hate the most. Lock your hardware down. I recommend you also lock down your keyboard and mouse, all your power cables, etc. This is even more important if you can't monitor the lab 24 hrs/day.
8) Servers. Lock them up! Don't let them be used directly except under supervision. This is because you want to encourage students learning new environments to have someone "look over their shoulders." A second set of eyes avoids mistakes.
Oh yes, expect lots of grumbling. People won't like all your policies, and some people "know" they can do a better job at running a network. It's important that you somehow communicate that ego isn't the most important thing to have when working with computers. If you can turn out ego-less, qualified administrators, you will be doing us all a service!
--
Joel Maslak
I'm going to suggest a radical idea here... (Score:3)
Amount of Funding = $100,000
25 iMac DVs = $32,475
1 G3 Server running OS X Server = $6,000
5 DV Camcorders = $5,000
100BT Ethernet setup = $100-200
Total = $43,675
Don't know what else you need, but that looks pretty good to me. Don't need a *lot* of software, as each iMac comes with iMovie (DV editing), AppleWorks (works program) and PageMill (HTML editing software), along with some other software that I'm not familiar with. Of course, if you've never used Mac OS, or are somehow required to use Windows, then this won't work for you... Mac OS is pretty easy to learn though.Anyway, was just an idea. Feel free to disregard if neccessary.
What kind of lab? (Score:2)
This, to me, leads me to ask a question: What is the goal of your lab? Is it just a place where kids can check e-mail, browse the web, and maybe do homework with office applications? Or is this a lab aimed at introducing computers to youth who may not have much experience with them?
The way I see it, it's a question that can't be answered with a blanket solution. If the lab is more of a 'web-cafe' for the community center, so to speak, you'd just need to have some cheap Celeron's run Netscape off the network, and maybe have some office suit installed.
Looking at the what the youth center was designed to do, however, doesn't make me believe that this lab will be just a way for users to browse the internet or type a paper. If you want to attract the interests of talented youth in your area (graphic artists, programmers, web designers, the 'people-skilled'), you'll have to give them access to tools they may be interested in, but haven't yet used.
Think about it. If you're a semi-savvy 17-yr old kid, where would you rather work? A computer lab where you get paid to sit and make sure no one steals the mice while you chat with your friends on Instant Messenger? Or a lab where the youth community is involved in computer projects they are interested in (programming software, graphic arts, web design, etc.)? I would see the lab really succeeding if it uses a mix of the different OSs. Windows for those who just need access to the Internet or want to relax with a Quake DM, Linux for those who want to fiddle with kernels, etc, and Macs with scanners and video-editing software for the digital artists. That way, you don't need to rely solely on one aspect of the lab (iMacs with DV, learning Linux) to draw in both users and motivated volunteers.
Also, I tend to agree with the poster that argued for giving users free reign of the system. Again, here. What's going to attract more users and volunteers? A place where you can't fiddle with the systems, or a place where you can try and figure out (if you want) how a system works, both software and hardware wise. Taking that a step further, why not have a couple of boxen in permanent 'in construction mode'). Not expensive systems, of course, just a place for computer users can see what's 'under the hood' so to speak. "Here's the video card... it's on this PCI slot, and you can remove it like this *POP* ... let's install an AGP card." If you have a well-stocked library of computer books (can you ask for book donations?), a lot of users and volunteers will have a place to start, and maybe something will catch their eye as well ("Perl, what's this?").
The way I see it, the more interactive a lab is, the more likely are users to flock to it, and if you can spark enough interest in the users, you can have some of them become devoted volunteers.
Just have a clear mission plan, set out before any of the equipment is shipped to you or you hire more than a skeleton crew. This sounds like an awesome project, but not one to be taken lightly. With sound planning, this is sure to be a success. Best of luck!
Student Run UNIX Lab (Score:2)
It would take some special youths. (Score:2)
LK
Re:Limit who has Root/Admin access (Score:1)
This is from "building a computer network 101," but the first thing you need to do is make a list of what you want people to be able to do on your network. If they are going to only be surfing, emailing, and maybe a little writing then your cheapest and best bet might be Linux/StarOffice/Netscape. Notice that I said might. I know that StarOffice is big and clunky, but it is adequate for typing up nice papers and presentations. It will also read from/to most MSWord, PowerPoint, and Excel files. I'm sure you know the rest of the pro-linux rant (Security, File-sharing, free-updates,
If you need your network primarily for office type applications, then go with WindowsNT/2000. Now you will have the advantage of being able to use MSOffice, and still have a good level of Security, File-sharing,
Finally, to keep with the topic of thread, I think that Ghosting is an okay solution for computing. It is a mistake to assume that, "treating the symptoms and not the problem" is always bad. There's a recent trend in the medical industry towards just that, especially with the soon-to-be boom in old people. If you have to use Windows98 on your network, then ghosting everyday might be a good way to prevent people from installing software. Just make sure that you can re-ghost a box during the day, just in case something get's screwed up. It does seem messy to have to do this, but running Windows 98 in a multi-user environment is already pretty messy. (this isn't a flame, all the pro-MS people were saying, too)
Defining Youths (Score:1)
People of any age could help out though.
Re:I'm going to suggest a radical idea here... (Score:1)
You can netboot the iMacs over a 100BT LAN, and each user would have his own space to do stuff in on the server (so you can use whatever Mac you want), as well as shared applications that everyone can have access to.
I'm not sure how fast our LAN at college is, but I know it's based on Cat5 cable. Anyway, we have a bunch of Power Macs in the physics lab, and the applications are hosted on a "phys-apps" server. And they are slow. So slow, it takes almost as long to net-load Word as to boot locally. That's probably Word's fault, since only Word and Excel are like that, but I thought I'd bring it up anyway.
Another thing to consider is the point of execution. At college, everything runs locally, so it's reasonably fast once loaded. If you put N people on a single server, they get 1/N of the computing time.
Oh, and since you have $56,325 left over, you might want to consider G4 instead for the server. I'm not sure it's strictly necessary, but it sounds cool :) Also, get a backup system!
Other than that, it sounds like a good idea.
-- LoonXTall
Re:Limit who has Root/Admin access (Score:1)
But if you're stuck using Win9x, permissions don't exist and users are free to trash computers and install all sorts of crap...
I could even set something to run as a service, even though the System Policy Editor (tm) was used to disable registry editing. Here's to Notepad!
-- LoonXTall
Roberts rules of order. (Score:1)
People are almos always more of a headache to manage than computers. Expectations will be high, and unless the group can work in a concerted effort, enthusiasm will dwindle when nothing seems to be happening.
I was part of a school Computer lab that was admined for the most part by students. It was a great place to hang out and most of us got to explore what we wanted to as well as providing very basic tech help for others who came in with projects.
Two things that were very useful to us in this endeavor were
1) An authority figure nearby who would and could back us up if things got beyond our control. She was very non-tech, but she knew how to deal with teens.(poor Mrs Kilpatrik)
2)A really cool group of geeks who wanted a space that they could play in. We all had very different skill sets and levels, and we were always asking each other how to do this or that. Having a friend teach me how to write macros and install software made it much more interesting than taking a course.
So try to have a good core group to work with. Have regular (bi-weekly/monthly) meetings with an adult about to deal with ideas progress and or problems on an ongoing basis. And try to offer one on one tutor like sessions to those who show an interest in what's going on and who show up regularly.
Just my $US.01-1/2 eh.
Re:I'm going to suggest a radical idea here... (Score:1)
That is slow... however, I think that NetBooting over 100BT is supposed to be pretty fast. In fact, Apple demoed 50 iMacs running 50 different full-screen (I think, not 100% sure on the size) QT movies without skipping a beat off 1 G3/400 server. So I figure 25 should be just fine. Plus, if it's really all that slow I suppose one could try to limit access to the internal HD, and make it read-only or something and run all the apps off it.
Plus if it's Word 6 (or an old Mac trying to run Office 98), then no wonder really... Word 6 has a bad reputation in the Mac realm... slower then molasses even on a G3/233 (which at the time was fast.)
Another thing to consider is the point of execution. At college, everything runs locally, so it's reasonably fast once loaded. If you put N people on a single server, they get 1/N of the computing time.
Pardon my ignorance, but I'm not 100% sure what you mean by this... I'm assuming you think that the server runs the apps, and sends the display to the iMacs (ala X Windows)? The apps use the RAM/CPU/DVD on the iMac, they just go across the network to get the data. (Correct me if I assumed wrong...)
Oh, and since you have $56,325 left over, you might want to consider G4 instead for the server. I'm not sure it's strictly necessary, but it sounds cool :)
Yup, that was a typo in the cost breakdown that I missed... I meant to say G4, if you look earlier in the post you'll notice that. :-)
Also, get a backup system!
Do you mean backup as in "if the G4 server breaks down" or backup as in "a machine to backup the server's HD every night at 4 AM"? Either would be good - in fact, you could combine the two... buy another G4 server and have it copy the contents of the main server onto itself, then all you'd have to do is switch things around a little bit and you'd be back up. OS X Server has a lot of UNIX in it, so if you wanted to do something like that and had any UNIX experience, it wouldn't be too hard.
Other than that, it sounds like a good idea.
Thanks, it's appreciated. :-)
Re:I'm going to suggest a radical idea here... (Score:1)
Y'know, I wouldn't be surprised if they were running 10BT at college. The network's been around quite a while... the computers in one lab were networked LCIII's! (They were replaced this year by PC's, and Win95 won't let the data collection software work.)
Word 6 has a bad reputation in the Mac realm...
Yup, Word 6. It runs fine, it just loads slowly. As does Excel 5.x. (Can't remember the exact version.)
I'm not 100% sure what you mean by this...
The server math (which isn't all that accurate, because users don't put equal loads on the server) was based on the X Window concept. I was trying to point out that it would be better to have the apps run on the iMac than the G4 Server. I think everyone knows it's more efficient in disk space terms to put the apps on the server.
Do you mean backup as in "if the G4 server breaks down" or backup as in "a machine to backup the server's HD every night at 4 AM"?"Tape drive" was what I had in mind at time of posting. But a redundant server + tape drive would work too :)
Another thing to add to the lab may be a few Zip or Superdisks. I favor Superdisks because they're a logical upgrade to floppies; but you may confuse newbies when their Superdisk doesn't work in a 1.44MB drive, or irritate Zip owners. Oh well, can't please everyone...
-- LoonXTall
Questions you need to ask!!! (Score:1)
Second... Grant mony abounds, and you can find more where that came from! Appily for every grant you can!
Third... Look for local resources. User groups, in the linxu and mac comunity abound, and they are always friendly and willing to lend a hand if they have it to give!
Fourth... Buy your hardware from a major vendor, the big name companies will cut you a discount because your a non profit (are you?)!!! Also uniform hard ware and setups are easier to maintain.
Fith... You should have some basic rules, and a time sceduele set up! Qualify your youth techs, and have them cross train each other... Kids learn from kids better than from adult! This will aslo create a more open atmosphere, and this is what you want. A comunity in and arround your lab will weed out any bad eggs!
Sixth... Dont forget those hidden costs... you might be tempted to spend the full 100k on computers, but you need spare parts, and some resereve if you run into any unexpected problems. Dont forget the little things like the electric bill... you could clobber your agency if you arent prepaerd for it!
Seventh... You need voluntiers... local buisness is the way to go. Tell them what you are looking for and they will go to their pool of employies and ask. They had this at my last job and I was out and about at schools doing consulting for free, and ended up teaching a lot of people a lot of things.
sorry for my spelling, feel free to email me if you want!
Re:Limit who has Root/Admin access (Score:1)
But even when regedit is disable with poledit, you can run poledit and remove the regedit restriction - even though poledit is a registry editing tool - it's not restricted. Very foolish on MS' part - anyone who wants to get around the restriction could just put poledit on an ftp server and then download and run it on the restricted workstation (or bring it in on floppy, or CD..)