Wikipedia:Reference desk/Archives/Computing/2013 March 20

= March 20 =

Networks project
I have to do a network project for college and I'm at a loss with it. I've included what I have done below. Despite it's short length I've been working at if for a while. The brief is intentionally vague, It's basically just design a network for some place under the headings, General Analysis, Network Topology, Manual and Special Notes. I think I'll be OK with the Network Topology. I was just hoping people could read over it quickly and make suggestions on what I could add, or if I have anything wrong. Links to online resources would be helpful also. It's a FETAC 6 Network course.

General Analysis of Requirements

This is a project to design a new network for The River Lee Hotel. The Hotel has 182 bedrooms, features an executive floor and a conference center. The hotel measures 70 × 20 meters and is approx. 5 floors tall. The entire hotel including all the rooms will need to have access to WiFi internet. Staff will have 3 computers at the front desk. These will be connected to an onsite server which houses the hotels database. The conference hall is equipped with several Ethernet ports to provide a high quality network access for business people in addition the WiFi. The internet is provided by Eircom, through a DSL cable.

Network Topology

Manual
 * The network is divided into 3 subnets.
 * 192.168.0.0/24 Is the subnet for all the staff computers and servers.
 * 192.168.1.0/24 Is the subnet for the conference room.
 * 192.168.2.0/24 Is the subnet for WI-FI devices to use.
 * All the staff computers (including the server) are connected to a switch. This switch is connected to a wireless router.
 * Because the range of the WiFi router is insufficient to cover the entire hotel, a Netgear Universal WN3000RP Wi-Fi Range Extender is used to extend the range of the network easily.

IP Address


 * The routers inbuilt DHCP server hands out IP address to any wireless computers that want to connect. The range of IP addresses is 192.168.2.1 to 192.168.2.250 allowing 250 wireless devices to connect at the same time. The wireless connection is not secured to allow guests to connect to the network easily.
 * Devices on the staff network are all given static IP addresses.
 * The router is given the IP address 192.168.X.251/24 where X is the appropriate subnet number. This router connects to the Eircom DSL router.

Cabling


 * All the networks are wired together using standard CAT 5E cables.

Thanks, 83.70.217.140 (talk) 01:31, 20 March 2013 (UTC)


 * Putting an entire hotel on one DSL connection to the wider internet would give disappointing results. Find out if your provider offers any faster connections. EdJohnston (talk) 05:24, 20 March 2013 (UTC)


 * 1) The word "hotel's" needs an apostrophe, since it's possessive, not plural.


 * 2) I'd be wary of range claims for WiFi. They probably say "up to X" range, which means with a clear line-of-sight and ideal conditions.  What you need to know is the minimum range, in the actual hotel.  I'd do testing in the actual hotel to determine the average range, then provide a fair margin of error beyond that.


 * 3) Can't you estimate the length of the cabling ?


 * 4) If this is an existing hotel, versus a new construction, then the issue also comes up as to how to get the cables into the walls. You could tear out the walls, but this is time-consuming, disruptive, and expensive.  Some type of surface-mount wiring in attractive conduit might be acceptable, at least in the conference center and office areas. StuRat (talk) 09:16, 20 March 2013 (UTC)
 * Got to second the bit about speed. It is a real problem as hotels can be in rather inacessible places as far as broadband is concerned but you just have to cater for people bringing along ipads and smartphones. The notebook use is inconsequential compared to that. Dmcq (talk) 22:54, 23 March 2013 (UTC)

An assembly example that seemed to be doing something the long way
I was looking at a Microsoft assembly example that did the following: 1)It declared in the uninitialized data .data? section a dd variable retvalue 2) At the end of the program, it moved retvalue into eax 3) it XORed eax with itself so that eax had all 0s 4) It invoked ExitProcess with eax as its one parameter (invoke ExitProcess,eax). My question: why do all that when you could just have at the last line 'invoke ExitProcess,0' ? — Preceding unsigned comment added by 67.163.109.173 (talk) 02:20, 20 March 2013 (UTC)
 * It would help to see the example. Do you have a link?  RudolfRed (talk) 02:32, 20 March 2013 (UTC)
 * http://securityxploded.com/assembly-programming-beginners-guide.php 67.163.109.173 (talk) 02:37, 20 March 2013 (UTC)
 * You have the mov backwards; in that syntax, the destination is on the left. The return value for stdcall is stored in eax, so it's saving the return value before exiting the process.  (This is still pointless, of course, without more to the program!)  As for XOR,   &mdash;   is only 3 bytes, whereas   is 5.  For a trivial program, of course, this doesn't matter, but shortening the program text is important for caching in general.  --Tardis (talk) 05:13, 20 March 2013 (UTC)
 * only takes 2 bytes (6A 00). -- BenRG (talk) 17:50, 23 March 2013 (UTC)
 * That's "push imm8"; I could be wrong, but I think it'll get promoted only to 2 bytes of stack space (in the 32-bit mode we're presumably using). --Tardis (talk) 23:58, 23 March 2013 (UTC)
 * No, it pushes 16/32/64 bits in 16/32/64 bit mode unless an operand size override (66) is present. -- BenRG (talk) 22:54, 26 March 2013 (UTC)


 * That's not an official Microsoft code sample, it's just code by some guy. The rest of the tutorial seems pretty badly written as well. -- BenRG (talk) 05:37, 20 March 2013 (UTC)
 * That's not Microsoft code sample, but it is Microsoft assembly sample, with syntax parentheses on (Microsoft assembly), i.e. MASM, not (Microsoft sample). --CiaPan (talk) 07:56, 20 March 2013 (UTC)


 * Yes, I meant it was for use on a Microsoft system, not made by Microsoft company. 67.163.109.173 (talk) 13:51, 20 March 2013 (UTC)


 * I can answer two bits of that: 1) "XOR AX, AX" is the fastest way of getting a 0 in the AX register on an 8086 CPU (on newer CPUs, there are methods that are as fast, but nothing faster, so the tradition of using XOR to zero has stuck around). 2) "invoke" is an assembler-specific directive that turns into a rather amazing amount of machine code.  I don't remember the details of the STDCALL calling convention, but I'm fairly sure that one of the instructions "invoke" hides is a "push", and as I recall, a "push REG" is faster than a "push IMMED" on (at least some) X86 CPUs. --Carnildo (talk) 02:11, 22 March 2013 (UTC)
 * I think people still use XOR or SUB to clear a register because it's shorter (at 2 bytes) than any other approach. -- BenRG (talk) 17:50, 23 March 2013 (UTC)
 * It used to be that you needed to clear a whole register using XOR or else accessing bytes via it went quite slowly. The wonders of optimisation. Dmcq (talk) 22:57, 23 March 2013 (UTC)
 * That's still true, but it's writing part of the register and then reading a larger part that causes the problem (a partial register stall), not the other way around. -- BenRG (talk) 22:57, 26 March 2013 (UTC)

Reduce background proccess priority/improve user experience in Windows XP
My mum's computer has the following specification:

Windows XP Home OS MSI-7100 motherboard AMD Athlon 3200+, 2 GHz CPU 2 GB RAM NVIDEA GeForce 6000 GT graphics card

When copying files from Android MicoSD card to system disk, the user experience degrades to an unnacceptable level. Specifically, the mouse cursor movements are delayed and videos played on YouTube stutter. CPU usage is below 100% most of the time.

How to resolve so that user experience is not impacted during this and similar background operations? How to prrioritise user experience at all times? --92.28.82.133 (talk) 09:45, 20 March 2013 (UTC)


 * You can try bringing up the Task Manager and lowering the priority on those processes, but, honestly, I haven't noticed much difference when I've tried that. My suggestion: Just wait until she's done with the computer to do mass copies. StuRat (talk) 09:56, 20 March 2013 (UTC)


 * Has the computer been on for a long time? In my experience with my old computer running XP, if I kept it running for over one week then performance would begin to degrade exactly as you describe. -- 143.85.199.242 (talk) 15:44, 20 March 2013 (UTC)


 * That's never been my experience. I think you must have been running some background task with a memory leak. -- BenRG (talk) 18:13, 21 March 2013 (UTC)

Windows copy is very bad. Try using FastCopy 92.233.64.26 (talk) 10:44, 20 March 2013 (UTC)


 * While I agree the native windows copy is not the best (I personally use Teracopy) I doubt it's the problem. It is highly unlikely that copying from a USB device would degrade performance as described. The system drive (probably SATA) has a much higher data throughput than a simple SD card. The bottleneck would be at the card - not the system drive. Secondly, at a hardware level, copying is not CPU intensive. CPU usage should be around 10% - if that. I would first do a "chkdsk /f" of both the system drive and the SD card. File system errors could cause high CPU usage. If the system drive is very full (85% +) there is likely to be a lot of file fragmentation. I would use MyDefrag to sort that out. If your drive is very full you can quickly free up some space by turning System Restore off - this will delete all your old restore points. Be sure to turn it back on. Only do this if you are sure your system is virus/malware free and you won't need to roll back.


 * I have encountered PCs with 2 different kinds of antivirus and Windoze Defender running all at the same time. Not a good idea. I have also encountered scheduled antivirus scans kicking in in the background and "killing" performance.


 * You may get a slight performance increase by going to the Control Panel -> System -> Advanced -> Performance Settings -> Adjust for Best Performance. (It will display XP in the Classic Win98 look). 196.214.78.114 (talk) 11:38, 20 March 2013 (UTC)


 * Sounds to me like the usual system performance slowdown experienced by older Windows machines. Have a look around for unnecessary background apps (do you really nead to check for updates for Windows, and your anti-virus, anti-malware, flash, adobe acrobat, realplayer, skype, etc... every time you log in?), resource hogs (Flash video is one of these so perhaps best to avoid YouTube), things that use a lot of CPU and/or memory, etc.  Empty the temporary file locations and defrag the hard drive. Make sure there is no malware using up the resources to send out spam mails.  Reinstalling Windows will almost certainly speed it up again, but that is a pretty drastic choice and perhaps the last thing you should try (rather then the first thing you should try).  Astronaut (talk) 17:42, 21 March 2013 (UTC)


 * I think antivirus software is the most likely culprit. Try disabling it while copying files and see if it helps. Unless you're not running any, in which case ignore me. -- BenRG (talk) 18:10, 21 March 2013 (UTC)

Thanks everyone! I optimised for best performance via Control Panel > System > Advanced > Performance/settings, stopped certain programs running at start-up (including Ad-Aware Antivirus since I already have MS Security Essentials running) including some accessed via Start > Run > msconfig > Startup. I also defragmented the System drive with MyDefrag's "Monthly" option (which will probably be run annually!) 78.150.234.51 (talk) 22:54, 23 March 2013 (UTC)

Codecademy for IE8
Hi,

I am looking for a website similar to codecademy to ideally learn Python or a similar language. However my PC at work has IE8, which is as you may know a dreadful browser and not supported by codecademy. I am unable to install a better browser due to my company's insistance on using the most backward software available, so I am looking for a similar website compatible with IE8, does anyone know of any good ones? Thanks! 80.254.147.164 (talk) 10:11, 20 March 2013 (UTC)
 * Weird, my question has only just become visible to me today.... (OP here) 80.254.147.164 (talk) 13:06, 21 March 2013 (UTC)

Possibility of Recovering a Canceled Wikipedia Edit
Hello, I was wondering if there was any way that I might be able to extract the large amount of content that I had entered into the text box used by the mathematics reference desk to store pending section additions from session data stored either on my computer or on Wikipedia's servers. I accidentally lost it in some clipboard hi-jinx that intervened with the usual copy-and-paste operation that I use to save this content in a local TextEdit (.txt) file. Apparently, my '⌘-C' keystroke did not register with the computer which I was using and therefore did not refresh the clipboard as I was expecting it would.

Please help me,

RandomDSdevel (talk) 15:03, 20 March 2013 (UTC)
 * Are both of the programs you used to type still open from the time you used them to type the content? Try using the Undo command (⌘-Z/Ctrl-Z) and see if you can recover any content. When I copy text from one area to another, I like to Cut it from the originating program, then immediately undo it in the same program (⌘-X/Ctrl-X, then ⌘-Z/Ctrl-Z). That way, I know that, if it was "removed" from the text editor, it was successfully copied to the Clipboard, while the text is still there because I undid it if I happen to need it later (at least, that's how it works with Windows -- I don't use Macs). Let me know if I'm misinterpreting your problem. -- 143.85.199.242 (talk) 15:31, 20 March 2013 (UTC)
 * I do the same thing. Copy just doesn't cut it, as far as telling me it worked successfully. :-) StuRat (talk) 17:31, 20 March 2013 (UTC)
 * No, I closed both programs (Safari and TextEdit) before logging out last night and only realized that my text didn't copy successfully this morning when I looked back at my dump file today. Instead of the text that I had expected to find upon opening the file, I found that the text file now contained what had previously been copied to the clipboard.  Although I doubt that even the most remote possibility exists that I might be able to recover my data through some sort of Safari or Wikipedia server history or cache file, could you help me figure out if I could?
 * RandomDSdevel (talk) 16:00, 20 March 2013 (UTC)
 * Was the text ever saved ? If so, there might be a backup copy, if that application keeps them.  If not, I'm sorry, but I think it's gone for good by now. StuRat (talk) 17:31, 20 March 2013 (UTC)


 * I usually edit in a Safari browser, and unless I either close it actively or back over an edit and start forward again, overwriting it, the edit is still there even if I have opened twenty new pages before returning to it. I am not quite sure what you mean by cancelling an edit. as long as the browser is still open you should be able to return to the edit unless you have already gone to an earlier state before the edit and started moving forward by opening new pages, rather than using the previous and next page arrows. μηδείς (talk) 13:00, 21 March 2013 (UTC)
 * I know it's a little late now, but in the lost horse/barn door security paradigm, you might investigate Lazarus browser extension, which exists to solve the problem of restoring vast quantities of text box data that vanished when the browser hiccups (we've all been there). Just installed it myself and haven't tested it, though. Gzuckier (talk) 17:05, 21 March 2013 (UTC)

Thanks for all of your help, everybody. I have now confirmed my suspicions that my text data is unrecoverable, but have, per Gzuckier's instructions, installed the Lazarus extension in my user instance of Safari as a fall-back mechanism for any future situations in which the event that prompted this post happens again.

–&thinsp;RandomDSdevel (talk) 20:44, 21 March 2013 (UTC)

Traceroute
For many years the first hop on my traceroutes has always been "10.224.208.2", which is a private ip address and basically seemed to be synonymous with localhost. Example;

Tracing route to example.com [192.0.43.10] over a maximum of 30 hops:

1    7 ms     9 ms    10 ms  10.224.208.2

However, today the first hop displays as "ComputerName [0.0.0.0]" (ComputerName being the name of the computer). Example;

Tracing route to example.com [192.0.43.10] over a maximum of 30 hops:

1    7 ms     9 ms    10 ms  ComputerName [0.0.0.0]

I don't understand why this has changed, since I have not knowingly changed any settings or hardware. Nor do I understand why the computers apparent local ip is now "0.0.0.0".

The computer connects via ethernet cable directly to a cable modem, there are no routers or anything else.

I want to learn why this change happened, and if possible return to having 10.224.208.2 as the first hop on the traceroutes

Thank you. 92.233.64.26 (talk) 16:45, 20 March 2013 (UTC)
 * I can't answer the question, but I would have expected the first hop to be your default gateway rather than localhost (check with ipconfig if you're running Windows). Is this your own private network or at work? AndrewWTaylor (talk) 17:14, 20 March 2013 (UTC)


 * Sure we can answer the question! 10.224.208.2 is an address on your private network, not on the public internet.  It's usually in your LAN, but it doesn't strictly have to be.  This "Class A" address space is a very old convention, specified by IANA, "following the standards set by RFC 1918 and RFC 4193" (meaning, it's a convention amongst routers to help cooperate when they're all on the internet; but is not actually a technical requirement of The Internet Protocol).  Still, you'd be hard-pressed to find any router in the world - unless you're a specialist in network engineering - that uses 10.x.x.x for anything other than private or local routes.  Lastly, just because you don't see a separate box doesn't mean there isn't a separate host in your cable modem.  It's very possible that your "modem" is actually many devices in one box: a router, a residential gateway, and a modem, and an RF transceiver.  It's non-trivial to say where the 10.224.208.2 host is, geographically - it could be the first router on your link to the Internet Service Provider; or it could be a miniscule little IP-enabled microcontroller inside your modem box.  Nimur (talk) 03:33, 21 March 2013 (UTC)
 * Thank you for the reply but I am not sure that it answers the question. I understand that 10.224.208.2 is private ip address, as I said that in my original question, but why has it suddenly changed to 0.0.0.0? And how can I change it back? The cable modem does not have any router functions at all, it has only 1 ethernet port which 1 computer is directly connected to via an ethernet cable 92.233.64.26 (talk) 12:36, 21 March 2013 (UTC)


 * The way tracert / traceroute works is that it sends out packets with variable time-to-live settings, and watches for ICMP "time to live exceeded" packets scattering back from routers in the network. It then reports the source IP address that those packets contain; there's no further interpretation to be done (bar the option of a reverse DNS to show a hostname). In your case, the first hop in the IP connection used to report 10.224.208.2, now it's reporting 0.0.0.0. So that first hop device (not your computer) has changed somehow.  I've no direct experience of DOCSIS connections, so I don't know if the first device is the cable modem or the CMTS. If it's the former, then the cable company may have upgraded the modem's firmware remotely (which is often something they do with modems they supply); if it's the latter case then they may have changed the configuration of the CMTS. -- Finlay McWalterჷTalk 13:59, 21 March 2013 (UTC)


 * I see you're on VirginMedia in the UK. Looking at Virgin's forums, I see the change is apparent for many people (perhaps everyone). Looking at this 2010 post, the first IP address reported is in the 10.x.x.x private network. The posting by swt4ajp in that forum page briefly shows some clues as to the structure of Virgin's IP network; it looks like they're doing intra-Virgin routing (for iptv and iptelephony) on 10.x.x.x, and mapping you into a public 92.x.x.x address when you go out to the internet proper.  But at some point they've changed - this 2013 post the person gets 0.0.0.0 like you.  That doesn't mean the 10.x.x.x addressing has necessarily gone away (if you do a netstat when you're watching iptv, you might still see connections to Virgin's media servers on another 10.x.x.x address), but that the first hop (which, if you don't have a "Superhub", seems to be the CMTS) is responding differently than it used to.  If you don't have a performance issue, then you can put this down as a nebbish curio; if you do then you should probably yell at Virgin's support people. -- Finlay McWalterჷTalk 14:46, 21 March 2013 (UTC)


 * Okay I understand now, they've probably changed something remotely at the cable headend that has caused this, that explains how it happened without me doing anything at my end. I checked the firmware version and it is the same as before. I'm still a little confused as to why the name of my computer is being arbitrated to this first hop though, if the first hop is actually the modem or CMTS 92.233.64.26 (talk) 17:03, 21 March 2013 (UTC)


 * When traceroute receives the IP packet of the remote host, it's running it back through the name client (DNS, usually) to try to resolve that into a printable name. It's probably calling getnameinfo. I guess if getnameinfo is fed the bogus address 0.0.0.0 it's returning the name of your computer. -- Finlay McWalterჷTalk 17:10, 21 March 2013 (UTC)


 * Yes, I just checked (with gethostbyaddr) and Windows returns the localhost name when fed 0.0.0.0  ; Linux, given the same info, errors with "unknown host"). -- Finlay McWalterჷTalk 17:16, 21 March 2013 (UTC)
 * Thank you for explaining, I marked this question as resolved now 92.233.64.26 (talk) 17:28, 21 March 2013 (UTC)

I have a question (about Wikidoc)
I found a E-mail address to someone on wikidoc. and sent the message. In case the person is not currently useing wikidoc. How by knowing there E-mail and full name. On wikidoc can I try to get more information on them to contact them. Like can I get a hold of a another E-mail address or phone number. Does anyone know — Preceding unsigned comment added by 76.191.18.223 (talk) 20:29, 20 March 2013 (UTC)


 * Wikidoc is not connected to Wikipedia, despide of having a common wiki in their names. Ask directly there how to contact any member. Wikipedia doesn't even have a page about Wikidoc, although it has a page about everything else. OsmanRF34 (talk) 23:20, 20 March 2013 (UTC)
 * Or maybe you could ask User:Wiki Doc about it.


 * I added to your title to make it actually useful. StuRat (talk) 06:59, 21 March 2013 (UTC)