Computing desk
< January 28	<< Dec \| January \| Feb >>	January 30 >

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

January 29

Unexplained internet activity

My windows icon for network activity lights up frequently when I'm not using the LAN/internet. When I pull up the 'Local Area Connection Status' there will typically be more packets sent than received. I have a router, asdl modem, 'Windows Live One Care' virus and spyware monitoring, and my firewall is on. I realize that some of my programs auto update, but it can't be as frequently as my light flashes. The unexplained network activity is really driving me nuts.

How can I easily find out what's autonomously pumping through my LAN/internet connection when I'm not using it? Is there a simple little monitor or shell program that will show where the data packets are going to/from, like a scrolling shell where I can watch it in real time without having to open a log file. I'd really like to know what Windows is doing with my computer. thanks. TungstenCarbide (talk) 00:46, 29 January 2009 (UTC)[reply]

Get a packet analyzer, or a software firewall like ZoneAlarm. --98.217.14.211 (talk) 01:21, 29 January 2009 (UTC)[reply]

A packet analyzer would help tell you what the info is. It could be something as nasty as spyware calling home, or something benign as your AV software checking for updates or IM proggie checking for activity. Ched (talk) 01:42, 29 January 2009 (UTC)[reply]

I wouldn't be too concerned if this is the only symptom. Installing a packet analyzer is a good idea, and you will be amazed how much ambient traffic exists even on your own personal network. The internet is worse. My recommendation would be Wireshark, which is both open-source and free. Shadowjams (talk) 01:56, 29 January 2009 (UTC)[reply]

If you don't want to install anything, you could also try the built in command "netstat" (in windows, mac, and linux, it's the same). This will show you what connections are in use. Generally it only works for TCP connections since UDP is connectionless. In windows you will probably want to use the following command netstat -o 15.

The command netstat -h will tell you more options. Shadowjams (talk) 02:04, 29 January 2009 (UTC)[reply]

You might also consider installing a personal firewall (compared here) on your computer. Typically they can be configured to only allow connections to and from specifically authorized applications, such as your web browser. If any other program attempts to connect, you will be alerted to the program's identity and given the option to allow of deny the current (and future) connections for that program. This should help you track down what connections are being made, although not what information is being passed. -- Tcncv (talk) 05:24, 29 January 2009 (UTC)[reply]

Do also read this thought-provoking page. -- Hoary (talk) 11:52, 29 January 2009 (UTC)[reply]

I believe what you are looking for is TCPView from www.sysinternals.com Sandman30s (talk) 12:02, 29 January 2009 (UTC)[reply]

Thanks everyone. Those programs were just what I was looking for. So far I've deleted btdna.exe. This really pisses me off. Once a year I have to download a torrent. Afterwords I uninstall bittorrent, yet it still leaves this program lying around, always on, firing off barrages of UDP events to 800 IPs around the world whenever it feels like it. TungstenCarbide (talk) 17:10, 29 January 2009 (UTC)[reply]

Try uTorrent. --grawity 09:35, 2 February 2009 (UTC)[reply]

No program-writing program

One of the few things computers still can't seem to do adequately is write programs. Is this because it's not actually possible to create a satisficing general-purpose program-writing program, or just because no programmers want to render themselves obsolete by creating one? Neon Merlin 05:36, 29 January 2009 (UTC)[reply]

The essential thing the programmer does is provide information. For example, "when this happens, we should do this". I can imagine a program for writing programs, but it would then need to ask a person a great deal of questions to figure out exactly how the program should work. Let's take a very simple program to "ask a person their name, then print it backwards". The program-writing program would have to ask "how long can the name be ?" and "by backwards, do you mean 'John Doe' becomes 'nhoJ eoD' or 'eoD nhoJ' ?". StuRat (talk) 06:05, 29 January 2009 (UTC)[reply]

Right, for general purpose program-writing program we have to wait till Artificial intelligence software becomes reality. For limited use, Wizard (software) are useful. Many Integrated development environment software offer wizard that take input from end user and create (sample) programs (to demonstrate concepts of programming language). manya (talk) 07:19, 29 January 2009 (UTC)[reply]

I'm pretty sure the programmer who was able to accomplish such a task (assuming such a task is possible) would be both obsolete and obscenely wealthy. Yeah, I'd say if it were possible today it would definitely have been done already. Although, in a way we do have programs capable of making other programs: we call 'em compilers. Unfortunately, the use of these compilers requires significant domain knowledge and training. There have been several efforts over the years to provide so-called natural language programming, but Dijkstra argued such was foolishness. The problem is that any sufficiently robust system exhibits the inner-platform effect anti-pattern: to provide as many features as the original platform, the inner platform ends up at least as complex. Other efforts have focused on visual programming with slightly more success, notably Simulink, but that program is focused primarily on system simulation using discrete (individually separable) components, which is a tiny fraction of a generic programming language's solution space, and still requires significant domain knowledge to operate. In essence, this question is like asking "why isn't there an automated assembly line that anybody can use to make automated assembly lines?" – 74 07:32, 29 January 2009 (UTC)[reply]

To expand on a point that 74 briefly made: There are program-writing programs, they're just not magical or intelligent. Compilers are certainly program-writing programs, and so are more specific programs like parser generators. Also, many dynamic Web pages need to write JavaScript programs on the fly. The reason that program-writing programs don't make programming obsolete is that you still have to be able to precisely tell them what to do, which is programming. A program-writing program that did not itself have to be programmed would have to be able to take imprecise requirements and turn them into a precise sequence of steps, which would require an AI equivalent to an educated computer programmer. rspεεr (talk) 08:16, 29 January 2009 (UTC)[reply]

It's because it's not possible to create anything remotely close to app that writes apps. Think about it this way. Before you can even begin to write an app, you have gather business requirements. So first, the program would have to be able to have actual discussions with human beings and have it actually understand what people are talking about. Those don't even exist right now and might never be possible.

People sometimes get the impression that computers are intelligent because they can perform complex calculations very quickly. The reality is that computers are extremely stupid. According to our AI article, no program can even pass the Turing test which is an very poor indicator of intelligence. The Turing test doesn't even measure intelligence. It measures the ability to trick a human into thinking it's human. A better indicator would be an actual IQ test which I severely doubt that any of these programs could score higher than a 0. A Quest For Knowledge (talk) 14:19, 29 January 2009 (UTC)[reply]

In addition to the information gathering portion, there also has to be an estimation phase. That is, what resources would be required to complete the task. Often, a programmer has to say "we can't do that, but how about if we do this, instead, which should be just as good or even better". A programmer often suggests additions, too, like "since we're printing out all these error messages, wouldn't it make sense to store them in an error log, too ?". So, this program-writing program would not only need to follow instructions, but make suggestions, too. Also, programmers must deal with faulty info. For example, I've had documentation that says "you provide this input to this call and it returns this output", but I've found the documentation is wrong in some way, so updated it myself. You can also get specs which are illogical, and must figure out where the error is, rather than implement them. StuRat (talk) 14:26, 29 January 2009 (UTC)[reply]

The act of computer programming, is to precisely describe what you want the computer to do. That process is being simplified all the time, and there are many simplified, or graphical programming languages. I don't understand how that process could be removed entirely.

I suppose you're imagining a programming language that is the same as conversational English. If that's the case you've got a long time to wait. Natural language processing is one of the most difficult problems in computer science, we are not anywhere near a general solution. APL (talk) 14:36, 29 January 2009 (UTC)[reply]

We have to be rather careful about how we answer this. First and foremost, in order for some 'program writing program' to know what program to write - we have to tell it. The form in which we tell it is...what? ...It's another program! Even if you somehow managed to have it accept standard English, you'd still have to say "Look up every one of my employees in the company database, look at how much they earn and write out a check for their monthly salary putting their name and address into the appropriate box on the envelope.". But I'm not sure you really want natural language English here because it's too ambiguous. I remember trying to write a natural language 'adventure' game - it's really not easy: "Hit the Wizard in the head using the sword"...OK, no problem..."Hit the Wizard in the cave using the sword"...ah...not so easy is it? So we tend to want to use a form of language that's a little more formal and is defined unambiguously. That's just a programming language. Your first reaction will to avoid things like C++ :

template <class T> const void *Get ( T &x ) { x = 6 ; return &x ; }

... and instead use

"ADD A TO B AND MULTIPLY BY C GIVING D"

...but sadly, that's a language called 'COBOL' which is probably the most hated language on the planet!

So where all of this leads us is to a program writer that uses a formal language which is not horribly verbose. That program-writer is what we call a 'compiler' and we use them all the time!

The issue is how high-level the language is. Very high level languages do exist - but they tend to be limited to small problem-domains. Very general-purpose languages tend to be quite low level.

Attempts have been made to make 'full-scale' program writers - one that comes to mind was arrogantly named "The Last One"[1] (meaning 'the last program that'll ever be written') - but it was a total disaster.

Many people have claimed to make 'program writers' using graphical input such as flow charts. If you buy a Lego Mindstorms system - it' comes with a flow-chart-like program generator. But in truth that's no different from conventional programming - except that you are doing it graphically instead of with a text editor - which may be a little less intimidating for the neophyte - but for people who do this for a living, it's frustrating as all hell! But it has it's place. I've just finished writing a graphical programming language to let artists write 'shader programs' for computer graphics cards - and many computer games make use of a system called Kynapse for programming AI and other game 'logic' without directly writing lines of code.

More directly to the answer the OP wants - we have genetic algorithms and neural networks that (in effect) "write programs" - but without a formal description of what the program has to do - they "learn by example". These things are becoming relatively mainstream. My car (a MINI Cooper) has an engine management computer that learns how to optimally tweak the engine parameters to suit the climate, the kind of gasoline and your driving style. When you take it in for a service and they have to download a new version of the software into the car, your miles-per-gallon drop by about 20% until it re-learns over then next few weeks.

SteveBaker (talk) 16:07, 29 January 2009 (UTC)[reply]

Another point to consider is that even if such a program could be built, could you afford to pay for it? We had a potential client where we recommended buying a utilty that would help upgrade their code. The utility cost 17 grand and the client couldn't afford that. 216.239.234.196 (talk) 16:45, 29 January 2009 (UTC)[reply]

Oh, there are plenty of things computers can't do well. F'rinstance, I've heard that computers can't cuddle. I dunno whether that's true or not, but every time I tell mine to wash the dishes or wash the windows, it plays dumb and tells me it can't find wash (I suspect it just likes being dirty; its always full of dust bunnies, and sometimes I can barely move the mouse for all the grime). On the other hand, the plethora of compilers available suggests that computers can write programs almost as well as they write sonnets. -- Fullstop (talk) 18:14, 29 January 2009 (UTC)[reply]

My computer will vacuum my floor. A Quest For Knowledge (talk) 18:29, 29 January 2009 (UTC)[reply]

Removable USB Disk Drive for Ubuntu

I recently received a new USB disk drive (a Western Digital "My Passport Essential" if that makes a difference). I plugged it into my laptop under Ubuntu 8.04 (Hardy Heron), and using the Partition Editor (GParted 0.3.5) partitioned the drive into a NTFS partition and a ext3 partition. The partitioning seemed to work O.K.; when I plug in the drive, the NTFS partition shows up as "/media/disk" and the ext3 partition as "/media/disk-1". Unfortunately, I can't copy anything into the ext3 partition (the NTFS partition seems to work). When I right-click copy something on the Desktop and then attempt to right-click paste it into the /media/disk-1 window, I can't even get the Paste option. Using the command line, I noticed that there were no write permissions for group and other on /media/disk-1 (unlike /media/disk , how'd that happen?), so I sudo chmod go+w /media/disk-1 to get that in, but it still doesn't work. Newsflash! I just discovered that I can use the command line to copy files into the ext3 partition (or create them with vi for example). I still can't do the right-click Copy/Paste though. What gives???

Also, when I get the properties on the partition (by right-clicking on the disk drive symbol for it and choosing Properties), it says under Contents "(some contents unreadable)". Wha...??? (It doesn't say this for the NTFS partition.) And what could those unreadable contents be, since there was nothing there except the "lost+found" directory when I looked at Properties.

A few more questions: where did Ubuntu come up with the names /media/disk and /media/disk-1 (as opposed to "/media/MyNiftyNewDisk" ? Is this something I can change?

Also, I thought about the /etc/fstab config file as being something I should look into regarding this problem, but there are no entries for these partitions in it, presumably because they are on a removable drive. Is there somewhere else I should examine? TresÁrboles (talk) 06:49, 29 January 2009 (UTC)[reply]

See [2]. Relating to the second question: my old iPod's FAT32 partition had a label, and when Ubuntu auto-mounted it, it used the partition's label. Have you set a label for your NTFS drive? Surely, /media/disk and "160GB Media" can't be worse than letters (F:, G:, etc.)? --wj32 ^t/c 09:52, 29 January 2009 (UTC)[reply]

The consensus in the thread you linked is just to chmod the mount point, but isn't that what I did with "sudo chmod go+w /media/disk-1" as I mentioned? TresÁrboles (talk) 00:34, 31 January 2009 (UTC)[reply]

And have you tried Properties... for your disk? Can you change the label there? --wj32 ^t/c 09:54, 29 January 2009 (UTC)[reply]

Well, on the Properties window, there's a Volume tab with a Settings dropdown. On the Settings is a space called "mount point". It's blank but looks like an editable field. This may be where you can do it... but I decided I'm not going to bother at this point. It looks like the Ubuntu system automatically creates the /media/disk and /media/disk-1 mount points when it automounts the partitions (and then also removes them when you unmount them). TresÁrboles (talk) 00:34, 31 January 2009 (UTC)[reply]

You have a broken file system on your disk. When Linux spots the file system is broken, it remounts as read-only (see /var/log/syslog for this happening). It does thiss silently (bug 28622). You can fsck it to fix it. --h2g2bob (talk) 23:08, 29 January 2009 (UTC)[reply]

No, I don't think that was it. I didn't see any such messages in the /var/log/syslog file. Also, I had done nothing with this disk besides partition it. In addition, there was obviously something weirder going on since, like I said earlier, I could use the command line to write files into the partition. TresÁrboles (talk) 00:34, 31 January 2009 (UTC)[reply]

Thanks anyway for the help. The problem has resolved itself, and it looks like I can now copy and paste into both NTFS and ext3 partitions. Unfortunately, I don't know what happened to make it work. But there was some useful (mostly Windows) stuff I learned along the way, which I'll write below for reference (for myself, if no one else).

What I did: Plugged the drive into a Windows XP desktop system, saw the only the NTFS partition as expected; didn't see an "Eject" option like I usually use for a USB flash drive, so I just yanked it out (I hadn't noticed the "Safely Remove Hardware" icon on the Taskbar); plugging the drive into my Ubuntu laptop gave me a message that it couldn't mount the NTFS partition -- obviously I should not have just yanked the drive out; however, copy/paste now worked for the ext3 partition!?! Unmounted the partition, removed the drive, and plugged it back into the Windows computer in order to re-remove it properly; however, when I tried the Taskbar "Safely Remove Hardware" it kept giving me the "device ... cannot be stopped right now" message. I wasn't aware of any program using anything on the drive, and restarting the computer didn't help. Luckily, googling led me to this which led me to the Process Explorer program; while looking for \device\harddisk stuff didn't lead me anywhere, I learned from this other post to do a search on "G:\" (where G is the particular drive letter assigned to the drive in question). That in turn showed me that some process called mim.exe was using the drive, and this handy site said that it was part of Musicmatch Jukebox. Whaaa?? I don't think I've ever used that program in my life, but armed with that knowledge, I decided it was safe to end the mim.exe process through Windows Task Manager. After I did that, I was able to use the "Safely Remove Hardware" taskbar thing, unplug the drive, plug it back into the Ubuntu machine, and voila! both NTFS and ext3 partitions pop up and I can now copy/paste into both.

The only thing remaining (I hope) is that in the Properties window for the ext3 partition, it still says "some contents unreadable" which I don't know is normal or not. TresÁrboles (talk) 00:34, 31 January 2009 (UTC)[reply]

That's because lost+found can't be read by you - it's only accessible to root. --wj32 ^t/c 05:24, 31 January 2009 (UTC)[reply]

videos in powerpoint

is it possible to embed videos from youtube etc, onto a powerpoint presentation? —Preceding unsigned comment added by 88.106.79.249 (talk) 10:37, 29 January 2009 (UTC)[reply]

Depending on your operating system, you can usually embed Windows media files or Quicktime files without too much difficulty. You can't embed straight FLV (the native YouTube format) into Powerpoint. You need to download the FLV from the YouTube video and then convert it to another format, first, and then you can import it to Powerpoint. --98.217.14.211 (talk) 12:34, 29 January 2009 (UTC)[reply]

Or, to be more specific: you can download the video with something like vixy.net (which only works for me about 50% of the time), and have it be converted into either an AVI file (Windows) or a MOV file (Mac, or Windows if you have Quicktime player installed), and you should be able to import it. Powerpoint can be a little finicky about such things, and you have to remember to include the movie file when you copy the Powerpoint somewhere else (if you do), since I think it usually just references it and does not really "embed" it in the file itself. --98.217.14.211 (talk) 12:43, 29 January 2009 (UTC)[reply]

Use KeepVid. It works the best (for me at least) and offers to let you download in iPod format, which is more easily converted into AVI. Buffered Input Output 16:09, 29 January 2009 (UTC)[reply]

In addtion to all these, you may also try the Safari Browser, which can download almost any video without any plug-in. Simply go to Window--->Activity

and check out for a file with 0.5 MB to 5<< MB size. Double click on it, and you will start downloading. For Firefox, I think Sothink Video Downloader is the best. Along with this, you may also try the Sothing Video Converter, which converts the flv. video to desirable formats. Both of these are free and you may download them at this link. Thank you, for such an interesting question. Anirban16chatterjee (talk) 18:05, 29 January 2009 (UTC)[reply]

I hate my computer...

Alright...here's yet another problem for someone to solve.

Yesterday, the fan in my case started making grinding noises. I turned off the computer, and carefully removed the fan and checked it (it was broken, btw). So then i turned the computer back on, and the CPU fan began racing at full speed. My BIOS settings say that system fan control is enabled. So, i reconnected the broken fan, thinking it would fix the problem, but instead both fans started going full speed. Did i break something, or do i need to DUMP THIS SHITTY eMACHINES T5010??? ANY help would be appreciated. Buffered Input Output 16:07, 29 January 2009 (UTC)[reply]

Sounds like the fan's temperature sensor isn't being read. Check the leads/connectors and the senor itself. Many bios' will show you what they think the temperature at the fan is. 0° == no reading. -- Fullstop (talk) 18:24, 29 January 2009 (UTC)[reply]

I didn't know fans had a temperature sensor. The fans I'm familiar with have an RPM sensor, while the internal temperature is measured by sensors on the motherboard. Anyway, operating a computer with the case open can cause less than optimal air flow patterns, which may lead to a fan speed increase. Also, if one of your fans is broken/missing it could be confusing the BIOS. If you replace the broken fan (which should be quite inexpensive) and put everything back together again, you might still have to modify your BIOS (try disabling fan speed control then re-enabling it) to get everything back in order. – 74 20:17, 29 January 2009 (UTC)[reply]

If someone could direct me to a 90mm case fan, i could actually fix it, but all the fans i find are 120mm or 80mm; 90mm seems to be a weird size. Buffered Input Output 20:51, 29 January 2009 (UTC)[reply]

Don't know where you're located, but NewEgg, starting as low as US$2.99. -LarryMac | Talk 21:27, 29 January 2009 (UTC)[reply]

Programs for drawing illustrations

What program would you use for drawing things like or the illustrations on this page [here]? --Mr.K. (talk) 18:50, 29 January 2009 (UTC)[reply]

You would want a vector graphics editor, like Inkscape or Adobe Illustrator. --Sean 19:45, 29 January 2009 (UTC)[reply]

my guess would have been a 3D Graphics Program Ched (talk) 22:48, 29 January 2009 (UTC)[reply]

You can do some kinds of diagrams with some kinds of programs - but sooner or later you'll need at least one of each of the three major categories:

A 'raster' paint program - Photoshop ($650) or GIMP (which is free).
A 'vector' graphics editor - Adobe Illustrator ($250) or Inkscape (which is free).
A 3D graphics editor - Lightwave, Maya or 3DStudio (~$500 to $7000) or blender (which is free).

In all cases, I strongly recommend the free version. In most projects, I find myself using at least two out of three of those programs in every piece of graphics I do! SteveBaker (talk) 00:48, 30 January 2009 (UTC)[reply]

Cool, do you have an example you can show? TungstenCarbide (talk) 01:26, 30 January 2009 (UTC)[reply]

I don't make diagrams much - I do interactive 3D graphics for games and simulation - but the same three basic tools are involved. Here is an ancient example: http://www.sjbaker.org/steve/personal/sg.png - the 3D objects are modelled with a 3D graphics editor - the textures that give them their appearance were painted using a raster paint program. In this case I don't recall using a vector editor for any part of the process - but I might have done used one for the lettering and various decals on ship and helicopter...I don't remember. SteveBaker (talk) 13:13, 30 January 2009 (UTC)[reply]

Override IP address in Firefox

Is there any way to (either for testing purposes, or to deal with problems accessing the DNS) instruct Mozilla Firefox to request a page from a particular IP address, even though the URL has a domain name that may correspond to a different address? Neon Merlin 19:18, 29 January 2009 (UTC)[reply]

Perhaps edit the hosts file on your development system? -- Coneslayer (talk) 19:25, 29 January 2009 (UTC)[reply]

You could always just put the IP address in the URL--which works for at least the individual page, and might work for subsequent pages (depending on how the server is set up). It is also possible to dynamically rewrite links from a domain to an IP address using something like greasemonkey, even on sites that use fully-qualified links. The easiest solution, though, is probably the hosts file modification Coneslayer suggested. – 74 20:30, 29 January 2009 (UTC)[reply]

Most smaller sites use name-based virtual hosting, so entering the IP address directly would not work for them. --Sean 22:27, 29 January 2009 (UTC)[reply]

Ah, thanks for the correction. Technically, even name-based virtual-hosted sites' URLs can be redirected to an IP address (a local development machine, for instance), but no URLs (virtual hosted or otherwise) can be successfully redirected to a site with name-based virtual hosting (unless the site supports an alternative virtual host access method). – 74 23:01, 29 January 2009 (UTC)[reply]

ntoskrnl.exe

An error with this file is stopping me from logging into windows (xp). I have no windows cd and no recovery disc but i do have a vista cd and a fujitsu siemens drivers and utilities cd. Can i repair the problem or should i give up and just install vista? 90.221.219.238 (talk) 21:57, 29 January 2009 (UTC)[reply]

Computer Hope carries an article on this. I've also seen several downloadable "repair ntoskrnl" proggies that will rewrite the file. If you have access to a bootCD (LiveCD, ERD, BartPE, Ubuntu, etc.) you may be able to find a copy in a backup folder (I don't remember which one) you could copy to the root. It's a long shot, my experience has been that when you get this error, it's just the first in a long line of system file errors to come. Ched (talk) 22:53, 29 January 2009 (UTC)[reply]

Edit video image size

I want to make a video from several clips. These clips have different image sizes. Is their a way for me to standardize the final video so that the image doesn't change proportions trough out the length? Currently I am using "Movie Maker". I don't know if it's at all possible to do anything like this with this program. Thank you in advance. Maziotis (talk) 22:43, 29 January 2009 (UTC)[reply]

Departmental Wiki

Keep in mind my computer literacy is that of an end user. My question revolves around what it would take to develop a wiki for an academic chemistry department website. The intention would be that the general internet audience would see only a regular web page. However once signed in an authorized user (one of a few hundred members of the department) could easily modify the entire site to keep it up to date. In addition the method by which wiki maintains histories would allow administrators to actively track changes. Any help in pointing me in the right direction would be greatly appreciated.--OMCV (talk) 23:55, 29 January 2009 (UTC)[reply]

Installing MediaWiki (the Wiki that Wikipedia uses) is fairly easy - you need first to install a package called "MySQL" on a suitable internet-connected server. You need to have an HTTP server that handles PHP (which is the programming language that MediaWiki is written in...if you're running web pages from that computer then you probably already have that) and you have to download and unpack MediaWiki itself. If everything went well, you should have a working Wiki that looks and feels a LOT like Wikipedia. You'll then need to create yourself an account and a system's admin account. Then you can choose what privilages anonymous users and users with accounts have. But (for example) preventing anon-users from editing articles, you force everyone to have accounts in order to edit. If you disable the ability for anon-users to create accounts, then you have a 'by invitation only' system where any signed-in user can invite more users to 'join' by creating an account for them. If you disable account creation for signed in users too - then only you (as sysadmin) will be able to do it. There are many, MANY more things you can do. Adding plugins for things like math rendering and such is also useful - but get it all working first. There are lots of tutorials out there - you shouldn't find it too difficult. SteveBaker (talk) 00:30, 30 January 2009 (UTC)[reply]

Frankly I would suggest that you get a tech person (an eager undergraduate, perhaps) to look into other content management systems other than MediaWiki, which is notoriously difficult to get up to speed and is really overkill for your needs. Tell them that you want something that supports user accounts and versioning (histories). There are a lot out there that do this that are not as complicated as MediaWiki. --98.217.14.211 (talk) 14:06, 31 January 2009 (UTC)[reply]

What?!!? That's a terrible answer! (well, apart from the "eager undergraduate" part) MediaWiki is really quite easy to install and hundreds of thousands of people have gotten 'up to speed with it' because they use Wikipedia or other MediaWiki sites. It's by far the most well known such system on the planet! Having used various other Wiki's (Twiki, clearspace, etc) I can tell you that MediaWiki is a million times better than any of it's competitors. I use MediaWiki for my local car club (my members have no problem getting to grips with it) - and I use my personal Wiki for keeping in touch with other members of my family - who also have no problems using it. 72.183.123.248 (talk) 17:46, 31 January 2009 (UTC)[reply]

Thanks for the replies and double thanks to Steve for his comments on "Programs for drawing illustrations". I was wondering if 98.217.14.211 could suggest what the "tech person" would install and how much would it cost?--OMCV (talk) 03:55, 1 February 2009 (UTC)[reply]

It really may be overkill to hire a full-time administrator. After following the installation tutorial and following the configuration manual, you don't really need a whole lot of "administration." I've operated numerous multi-user wikis that I install-and-forget. Depending on how fast you can follow instructions, setup and configuration can take anywhere from 5 minutes to one or two hours (I think I've even beat 5-minutes on a ready-to-roll Ubuntu system). Nimur (talk) 05:11, 2 February 2009 (UTC)[reply]

In case it was not clear, all of the necessary software is free and open source software and is available without charge. Nimur (talk) 14:32, 2 February 2009 (UTC)[reply]

Wikipedia:Reference desk/Archives/Computing/2009 January 29