Wikipedia:Reference desk/Archives/Computing/2009 January 24

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 24

PPPoE Password

For an unknown reason, my DSL modem (2Wire 2071) stopped working this morning. The error message stated that PPP authentication had failed, and that the username and password were incorrect. It was quite strange, considering I've never had to enter any username or password before, but here it was, all of a sudden this morning. Now the problem is, I have no idea what the username or password might be. I've accessed the modem via browser and found the username, but the password shown was simply "XXXXXXXX", which did not work when I entered the login credentials. I've contacted my ISP to inquire about this problem and if I could reset or retrieve the PPP password, but no luck. So I'm stuck to using dial-up for the time being until I can get this resolved. Any suggestions on how to find out the correct username and/or password? Running Windows Vista if that matters. I appreciate it. Vic93 (t/c) 02:52, 24 January 2009 (UTC)

Your particular modem is also a router. Let me get this straight: You can still log into the modem, but you can't change the PPP password? Or are you locked out of the modem entirely? What did your ISP say, by the way? They should have your PPP password. The password for the router/modem is another story.--K;;m5m k;;m5m (talk) 03:11, 24 January 2009 (UTC)

Usually when I start up my modem, I wait a while until I get a solid green DSL light and a green Internet light and from there, everything's a go. No Internet light or a red light indicates a problem. Open up a browser when there's a red light, there's usually an error message. This time it was that PPP Authentication had failed, as in I couldn't connect to the internet because of the wrong credentials (at least, that's what I think is happening). I can access the modem via web browser, by going to 192.168.0.254. I had tried to change the PPP password, but it seems it still doesn't accept it (still returns the failed PPP authentication error). My ISP pretty much told me to go through the textbook routines (i.e. restart modem, reset modem, check DSL filters) and just ignored the problem. I asked them if I could retrieve the password, but they said that was a separate issue altogether, and they're elevating the case. Sounds like they didn't know what I was talking about. =\ Vic93 (t/c) 17:50, 24 January 2009 (UTC)

Your ISP's tech support is the best place to call (again) for help, but here are some ideas. The PPPoE username and password are possibly the same as the username and password for your ISP e-mail account. (Did you or someone with access to your account recently change or reset the e-mail password?) Since you can access the internet another way, you might try looking on your ISP's website for a way to reset the e-mail password yourself. --Bavi H (talk) 23:05, 24 January 2009 (UTC)

It also, can depend on your ISP, some DSL providers have switch to a Bridge1384 network, and other to a Ethernet network. It is possible, that your ISP has either switched their configuration in there Redback Redback_Networks. Or the config in the modem is set wrong. Nick910 (talk) 00:29, 25 January 2009 (UTC)

Don't know why, don't know how, but it started working again today. Weird. It was probably a problem on their side, I'm guessing. Thanks to everyone who helped though! Vic93 (t/c) 21:00, 25 January 2009 (UTC)

Second natural language of open source

English is clearly the main language used by the FSF, OSI, Linux distro vendors and developers, etc..., but what is the movement's second language? German and Spanish are clearly the primary secondary languages, but Japanese and Italian also seem to be prevelant. --hello, i'm a member | talk to me! 03:42, 24 January 2009 (UTC)

I'd say German, due mainly to the Knoppix distros. StuRat (talk) 06:26, 24 January 2009 (UTC)

You'd think that the Finns should deserve it, but yeah, I agree, it's probably German. Belisarius (talk) 14:51, 25 January 2009 (UTC)

Outlook Web Access security

Hiya. I'm thinking about running OWA, just a couple of quick questions:

1. It always says that when logging out, you should close all open browser windows. Would not doing that (or forcing users to do that) compromise security?
2. Is it possible to monitor access (accounts, times, IP addresses etc.)?
3. Does it have access to the "Outlook" options-menu, for altering signatures etc., or must that be done from the actual computer?

Thanks! ╟─TreasuryTag►contribs─╢ 08:42, 24 January 2009 (UTC)

Regarding 1, I think this is mostly (justified) paranoia about cross-site scripting and cookie-theft.

Regarding 2, this info is available in the server logs, but I don't know of any way for regular users to see that

Regarding 3, OWA has its own settings (which means you need to set the sig again in OWA).

Note, incidentally, that there are virtually two different OWA clients. If you're using IE you get a very full featured (really very nice) client that's a lot like the Outlook client. If you're using another browser like Firefox you get a pretty basic webmail (entirely inferior to Gmail). They're so different that one might as well consider them unrealted programs. Mimetic Polyalloy (talk) 10:15, 24 January 2009 (UTC)

I'd assume 1 actually refers to session cookies, which may hold sensitive data, or even keep you logged in. Whilst logging out should have dealt with these, it seems a reasonable idea to close all windows and hence manually wipe them all. - Jarry1250 ^{(t, c)} 10:45, 24 January 2009 (UTC)

It could also be an attempt to prevent anyone with access to the computer from using the back button to view private communications. Even when properly logged out of the site (cookies and all), back button functionality may display secure content without performing a request to the server (which would be denied due to the logged-out status). Closing all the windows theoretically removes temporary information stored on previous pages (but not cached). -- 74.137.108.115 (talk) 14:02, 24 January 2009 (UTC)

Diode Differences

First off I'm no electrician! I would like to make a very simple box that takes seven 3.5mm stereo connectors and outputs the whole lot to one 3.5mm stereo socket. I tried a quick mock up just using headphone splitters but sound 'leaks' from one device into another causing a drop in volume. I'm assuming that if i use a diode on each of the input terminals I can stop this from happening. A quick look at diodes and there are rectifier, signal, zener, schottky and bidirectional transient voltage ones. Can anyone advise, or even inform if this crackpot scheme will work? thanks Kirk UK —Preceding unsigned comment added by 87.82.79.175 (talk) 09:15, 24 January 2009 (UTC)

First, straight-wire combining multiple outputs can lead to multiple sources attempting to drive the line, and might damage sources that can't handle the conflict. A simple mixer circuit can combine multiple inputs using resistors. A Google search for "audio mixer" will provide any number of projects and products, ranging from the simple resistive combination above to professional audio equipment, allowing you to select according to your application and budget. -- 74.137.108.115 (talk) 14:37, 24 January 2009 (UTC)

The previous advice is good if you really want to combine the 7 signals into one. However, if you just want any one of the 7 signals to pass through at a time, then a switch box would be better, so you can select which of the 7 signals are passed through at any time. StuRat (talk) 15:06, 24 January 2009 (UTC)

I'm more of a digital guy - analog stuff is a bit of a black art for me...but I don't think diodes help because you have A/C signals. Personally - I'd stick a 741 op-amp on each input and use a simple resistive mixer to mix the outputs of the amplifiers together. The amplifier chips should stop the signal from one input being pushed back into the other - and the amps can be adjusted to give you volume controls for each input. You can pick up a quad 741 chip in Radio Shack for a couple of bucks...circuit diagrams for hooking them up are everywhere on the web. SteveBaker (talk) 02:30, 25 January 2009 (UTC)

You need a virtual earth mixer using an op amp. This will add all your signals together and give you one out put. Of course for stereo, you need one summer (mixer) for each channel. I'm pretty sure there's a circuit for this under Opamp applications. YES there is! Look at the summing amplifier paragraph.--GreenSpigot (talk) 03:56, 25 January 2009 (UTC)

Also see here [1]--GreenSpigot (talk) 04:08, 25 January 2009 (UTC)

OK - but how do those prevent the input from one source being fed back into one of the other sources? SteveBaker (talk) 07:03, 25 January 2009 (UTC)

The passive mixer in the lk doesnt. but the opamp mixer does prevent coupling because all the current from the inputs go to a virtual earth at the opamp inverting input. —Preceding unsigned comment added by GreenSpigot (talk • contribs) 13:07, 25 January 2009 (UTC)

Ummm... No. Consider the situation where input one supplies 10v and input two supplies -10v. Since the R's are equal, the voltage of the "virtual earth" will already be 0v, so the op amp (with an output of 0v) will not draw the current. The current will travel from input 1 to input 2. Any excess voltage at the input junction will be drawn by the opamp (a 10v and 0v input, for example, would not feed a significant current back the 0v input), but if two inputs differ in sign then there will be a feedback current (depending on your definition; see below). The key, in this circuit and the resistive join above, is that the resistances are chosen such that any feedback currents will be acceptable to the inputs. -- 74.137.108.115 (talk) 19:42, 25 January 2009 (UTC)

Upon further thought, if you consider the -10v input as defined to accept an incoming current (similar to how a 10v input is defined to provide an outgoing current) then your definition of "feedback current" might exclude the example given above. (In that case, a "feedback current" for a -5v input would be to source current to a -10v input.) But, regardless of your definition of "feedback current", the key is still to choose resistances such that no input is driven outside of its current limitations. If the resistances are too small, this circuit can damage inputs just like a straight-wire join. -- 74.137.108.115 (talk) 21:06, 25 January 2009 (UTC)

This really is a non problem. Obviously the input resistors on each channel must be sufficiently large such that they do not load the source too much. Sources may be typically around 1kohm. So if you use 10kohm input resistors you aint going to affect the source voltage too much.GreenSpigot (talk) 01:26, 29 January 2009 (UTC)

Drawing trees with OpenOffice Draw (or some alternative)

Here you see a syntax tree. I want to create something rather similar, although with unlabeled nodes. Donald Derrick's "TreeForm Syntax Tree Drawing Software" (explanatory video here) doesn't seem designed for this. Consider this diagram and imagine that "AgrP", "VP" and the rest were removed to form an unbroken line from the top to the bottom right with lines branching off it to the left: that's the kind of thing I want to do. Having all the lines at a default to form an equilateral triangle would be a welcome bonus.

Surely this kind of thing ought to be possible with OpenOffice Draw but (maybe because it's a program I'd never tried till today) I found that if I had two points that were (pretty much) horizontal I couldn't even get Draw to draw two lines from them to form an equilateral triangle. (After much trial and error I could get two lines that met at a point and stopped there. But there was no obvious way to move the point to make the result equilateral.)

Does OpenOffice suck, or am I approaching it in the wrong way? (I even have Haugland's OpenOffice.org 2 Guidebook at hand and that doesn't help -- it's full of laborious explanations of how to produce visual gimmickry.)

If there are suggestions for other software, I can use a Linux, Mac OS X or Windows computer. Thank you. Morenoodles (talk) 10:49, 24 January 2009 (UTC)

I would use Dia for this. It is called Dia because it is specifically designed for drawing diagrams. I'm sure other programs could do it, but I prefer programs that are designed to do it. -- kainaw™ 15:09, 24 January 2009 (UTC)

I'm no linguist, but I've read enough Language log to know that a great tool for diagramming sentences is phpSyntaxTree. It'll diagram the heck out of your sentences! Go to the site and enter

[TP [Spec] [T' [T] [AgrP [Spec] [Agr' [Agr] [VP [Spec] [V' [V] [...]]]]]]]

It's not exactly what you want (the vertical line isn't completely vertical), but it's pretty darn good, IMHO. If you want to have much more control over how it looks, I'd recommend Inkscape, not having any in-depth knowledge of software specifically made to do diagrams. But I know that you could easily do it in inkscape. Belisarius (talk) 14:48, 25 January 2009 (UTC)

Many thanks to both of you. Dia seems to do the job, but I've made a note of phpSyntaxTree and Inkscape and will investigate both soon. Morenoodles (talk) 05:36, 26 January 2009 (UTC)

Error Message on Startup

I use Windows XP SP2, and for a few days now, I've been getting this error message everytime I switch on my PC.

Can't run 16-bit Windows Program.

Insufficient memory to run this application. Quit one or more Windows applications and then try again.

Please help me get rid of it. Thanks in advance! La Alquimista 16:11, 24 January 2009 (UTC)

I'm assuming that the memory in your computer hasn't changed recently, and that POST displays the correct amount of RAM. Either the amount of available RAM has changed (perhaps due to a near-full hard drive providing limited virtual memory), or additional programs are being run at startup. This page provides instructions on how to check startup programs and conserve memory. Finally, you might want to consider purchasing more memory (it's relatively cheap and one of the best computer upgrades available if not already maxed out). -- 74.137.108.115 (talk) 19:16, 24 January 2009 (UTC)

You may want to check your "msconfig" and see if something has been added there. If so, just uncheck it - and it won't try to run at startup. Ched (talk) 19:41, 24 January 2009 (UTC)

I did check msconfig, and found an executable file by the name of remind.exe, which is probably something to do with my Turbo C++ version 2.4. I deleted the file from its directory, but now the message has changed to:

Cannot load or run C:/TCWIN45/PIPELINE/remind.exe. Make sure you've typed the name correctly.

La Alquimista 05:14, 25 January 2009 (UTC)

It's not the file you want to delete, but the setting that's telling Windows to load that file. In msconfig, you can untick the box next to that entry, and next time it boots it will do a "selective startup" without that item.

If that solves the problem, then you can remove it from the startup list permanently using regedit [insert Microsoft-style disclaimer about the dangers of using regedit here!]: Next to the item in msconfig, it will tell you the path to the value in the registry which is loading that program ("HKLM" is short for "HKEY_LOCAL_MACHINE", "HKCU" is "HKEY_CURRENT_USER"). Run regedit, find and delete that entry, and Windows will no longer try to run it on startup. - IMSoP (talk) 19:30, 25 January 2009 (UTC)

how to get a long-living cellphone

I'm looking for a long cellphone with a long life expectancy. But all models I've looked at are described in reviews as having a fair chance of breaking within one or two years, which I find unacceptable (for environmental reasons). So how would you go about looking for something more long-living? Particular brands, models, etc. that you would recommend/avoid... and why? Or which forums/websites/... would you look at?

(If you're thinking of particular models: I'm looking for nothing gimicky, just plain calling and text-messaging to be used with a SIM card. Reliable (minimum of lost calls, etc.)!!! Reasonable volume and decent batdullllasssssssstery life. Vibrate if possible. Sturdy (given that I want a long-living phone). And I need tri-band or quad-band (with the European frequency bands and at least one--probably 1900--US band).)

Thanks for answering - Thanks for answering (talk) 16:26, 24 January 2009 (UTC)

All the phones I've had (mostly Sony Ericsson, but also a Nokia or two) have served me 2 or 3 years before I have passed them on in good working condition to various younger relatives. They have then gone on to serve my relative for a further 2 years or so. I suppose it depends on how well you look after your stuff. Of course, if you throw it around, mix it with your loose change and keys, and send it through the washing machine, it's lifetime will be seriously reduced. Astronaut (talk) 20:23, 24 January 2009 (UTC)

I recently bought a cell phone, and bought a couple of extra lithium ion batteries for it. This is because, in the past, I've had the battery stop working, only to be told that they don't sell that battery any more, so I can't get a replacement. The kicker, though, is being told that any minutes left on the phone will be lost. I've also noticed that it seems harder to get phones with separate buttons any more, instead they want to sell those phones with the pressure sensitive spots behind a solid sheet of plastic. That type is probably better at avoiding spilled liquids getting inside, but I've had bad luck with those type of "non-buttons" before, notably on a 1983 Pontiac Trans Am, which had them for the door locks, which took to locking and unlocking spontaneously. So, to me, that technology is just total junk, and I avoid it. Give me separate buttons every time. StuRat (talk) 21:04, 24 January 2009 (UTC)

Avoid the iPhone unless you're going to buy a case that covers the screen (it breaks easily). RAZRs are pretty good for longevity, as long as you don't drop them in anything at all (coffee, swimming pools, you name it: the RAZR dies instantly when tainted by liquid). The RAZR can handle anything; dropping, throwing, you name it, it'll survive it. Mine has even gotten run over by a truck... twice. And I still use it. Slider phones tend to be able to survive some intense dropping. Some people say that flippy phones can break if you flip them too much, but don't believe them, it's a myth. flaminglawyer 21:26, 24 January 2009 (UTC)

Hmmm... that's interesting; my last "flippy" phone did in fact break from excessive flippage. I then ordered three "parts" phones online, and all three had at least partially broken hinges. That being said, my old phone served a good while before becoming unhinged, and my new phone is another "flippy" model. The hinge *is* a point of failure, but it'll last a good long while if you aren't too rough with it. -- 74.137.108.115 (talk) 01:03, 25 January 2009 (UTC)

Sorry, let me rephrase that - Flippy phones rarely break from over-flippage, but that's just about the only thing they break from. So any broken flippy phones you find will be from over-flippage. flaminglawyer 03:57, 25 January 2009 (UTC)

I have a Sony Ericsson that's about five years old and works fine. That's for one nation. For another, I have a Casio that looks as if it might withstand drops and so on. I dropped it about 50 cm once onto a not-terribly-hard surface and its mike stopped working, or got disconnected. That was irritating, but the phone company fixed it without charging me anything. -- Hoary (talk) 15:34, 25 January 2009 (UTC)