Wikipedia:Reference desk/Archives/Computing/2009 August 17

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.

August 17

Does using Memory Card ruggedly as compared with Pendrives reduce it's life?

Hello friends!, I've been using my Micro-SD (8GB) Transcend memory card for all purpose data copying or for backups instead of pen drive Mass Storage device...But some of my friends keep telling me that using memory card as a substitute of Pen drive is not advisable. They say it will worn out soon,because it's not meant for such rugged data storage. But what I believe is that we use MEMORY CARD READER to read\write data to memory cards.Then how does that reduce the life of the card. Is this true?.Does anybody know if this is true or not?.Thanks —Preceding unsigned comment added by Balan rajan (talk • contribs) 10:15, 17 August 2009 (UTC)

All Flash memory technology has a limited life, measured in the number of times that a given page is written. It doesn't matter much whether it's in the form of a USB stick "pen drive" or an SD card. Probably a bigger issue is whether you will lose it, drop it in the sink, etc. --FOo (talk) 16:15, 17 August 2009 (UTC)

The amount of erase-write cycles before it becomes an issue is generally quite high (usually around 100,000+); you're more likely to physically break it first. HalfShadow 23:49, 17 August 2009 (UTC)

No...The issue is,for small form factor memory cards,the read\write cycles could be considerably less!...That led to this question...Because Pen-drives are rugged because of it's size and safety form factor... —Preceding unsigned comment added by 122.164.62.13 (talk) 04:43, 18 August 2009 (UTC)

It sounds like you are asking whether the two devices use MLC Multi-level cell or SLC Flash Single-level cell - as far as I know both use MLC in general (though some do use SLC) - but you could take them apart to check. Thus the memory would be identical. Given the very low price of USB memory I wouldn't expect it to contain a superior product. Though this information isn't final, it depends on what you bought. 83.100.250.79 (talk) 21:17, 18 August 2009 (UTC)

Clarify - this press release [1] states that "Transcend insists on only using original SLC (Single-Level-Cell) NAND Flash chips in its microSD™ cards". That was in 2006 and refered to 256MB cards, a web search shows that SLC chips are available in microSD format at least up to 8GB. Any more would be speculation.

However features and specifications may change from country to country, and over time. An USB memory stick may also use SLC too. Other factors may be an issue. I can't say for certain - though it definately doesn't sound like the microSD card you got is at any disadvantage - this does not apply to all, obviously. 83.100.250.79 (talk) 21:31, 18 August 2009 (UTC)

Let's be honest here. Whether one technology happens to have a longer life than the other – and that is likely to depend on the manufacturer far more than the form factor – either one will last far longer in the type of use you describe than you'd ever want to use it. For 'data copying' and 'backups', that sounds like fairly intermittent use. Let's assume that you completely overwrite the device with new data three times every day (which ought to be a serious overestimate). That's about a thousand writes per year. Further, let's assume that you only get one-tenth the life out of your device as you should: ten thousand erase-write cycles, instead of one hundred thousand plus. In this worst-case scenario, your memory card should be expected to fail in ten years, sometime in 2019. You'll be able to replace your current device using change you find in the couch — if you could even find one for sale. TenOfAllTrades(talk) 21:53, 18 August 2009 (UTC)

I first interpreted your question as asking whether memory cards were physically more fragile (breakable) than USB pen drives, and it seems to me that maybe they are. As for write wear, I don't think you'll come anywhere near having it affect you as long as you're just copying files around between computers, rather than using the memory card as a substitute for a hard drive in a busy system that is writing and reading a lot (e.g. a database). 70.90.174.101 (talk) 09:06, 19 August 2009 (UTC)

Balan (talk) 11:39, 19 August 2009 (UTC) Thanks for your suggestions and answers guys. I've been using my pendrive for USB-OS previously.Yes!,I use my Mass Storage more for USB-Version of linux such as Puppy,Ubunto. Well!, So in this case, Read or write cycles happen all time. Since recently I have changed to Memory cards rather than pendrives because of it's cost-effective nature and more compact so that I can have as many handy cards for different OS or purpose....I have started using it already...Since Transcend's latest Memory card format like this "http://www.cdfreaks.com/hardware/product/75381-Transcend-microSDHC-Card-8GB--MicroSD-Adapter.html" offers bulk transfer speed. Practically I have seen speeds over 34 to 40 MB\sec,which is simply amazing. So is the reason for opting this. Now along with this came this troublesome doubt!....

Windows Vista Explorer

Is there any alternatives to the Windows Explorer shell in Vista? Windows Explorer keeps crapping out on me, not responding and being generally slow. It's just Windows Explorer that's got this problem, everything else works perfectly. I remember years ago being able to use the Windows 95 shell on 98 and ME, but Vista doesn't work with that. Any free alternatives? Thanks —Preceding unsigned comment added by 82.44.55.97 (talk) 12:16, 17 August 2009 (UTC)

When you say "free alternatives" Linux springs to mind, but I don't think you quite meant that... Kotiwalo (talk) 18:38, 17 August 2009 (UTC)

It's probable that you are talking about Windows Aero (and not Windows Explorer, properly). Windows Aero requires a fast computer with good 3D video acceleration; if your computer is old (or not top-of-the-line) it is common to have trouble running Aeoro smoothly. There are many tutorials on disabling Aero and using the "Classic" Windows theme - the process is very easy. Here are some tutorials: How-To-Geek - Disable Aero, and Microsoft Help Troubleshooting Aero. Nimur (talk) 20:11, 17 August 2009 (UTC)

No, I have everything set to the absolute minimal, including disabling all fancy look and feel features. This is actually explorer.exe in the C:/windows directory which is causing me trouble. I often have to end task it from task manager and start a new instance, which fixes the problem for a few hours. I'm looking for a shell replacement —Preceding unsigned comment added by 82.44.55.97 (talk) 20:38, 17 August 2009 (UTC)

I remember looking for an Explorer replacement in WinXP a couple years ago and stumbling upon this list. Perhaps one of the programs listed there would work for you (none worked for me, so I stayed with my beloved FAR Manager).—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); 20:52, August 17, 2009 (UTC)

(ec) I think your options are then fairly limited. Most likely, you have accidentally (or intentionally) installed some shell extensions (which I have found to be consistently the most common source of trouble in Windows, period). Windows does not let you replace Explorer; it's really embedded into the entire graphic user experience. You can use other programs for file-browsing; but it's not recommended to run without explorer.exe (although it is certainly possible - by killing the process manually and not restarting it. If you're willing to deal with the consequential compatibility problems this can cause, then you might be okay). Norton Commander and similar software may be suitable for replacing your file-browsing needs. Nimur (talk) 20:52, 17 August 2009 (UTC)

There are actually quite a few Explorer replacements for XP, and some of them work for Vista as well. Check out this list & review of 10 Windows Explorer replacements. Out of the ones listed on there, Cubic Explorer, Free Commander, and File Matrix look the most promising for you.

edit - looks like Ëzhiki already posted that link. Well, whatever. Indeterminate (talk) 02:14, 18 August 2009 (UTC)

Disable Windows Media Player video conversion

Windows Media Player keeps trying to convert videos when synchronising, despite me knowing that the videos work at their current state (I can manually copy them to device and they play fine). I wanted to use WMP's auto-playlist function for Syncing videos, and all help pages on the net indicate how to turn it off, but it doesn't look like the option is there for my device. By the way, I am trying to sync to a Creative Zen X-Fi 16gb player. Jwoodger (talk) 14:13, 17 August 2009 (UTC)

Digital Electronics

This question has been cross-posted on the Science Desk. You can read responses there. Nimur (talk) 20:47, 17 August 2009 (UTC)

Nokia E63 caller id Display

Hello everyone,

This is regarding my phone I brought from Asia to use in the United States, if some caller is calling me my phone displays for example : +12346789023, how do i make it to display it like +1(234)-678-9023 or even 234-678-9023 there has to be some way, frankly i have been digging into it for quite sometime now, and finally decided to hand it over to you guys! Its really annoying to be using the number in that way, frankly i loose track of them! Hope to receive help!

Thanks everyone! —Preceding unsigned comment added by 123.252.229.186 (talk) 20:07, 17 August 2009 (UTC)

As far as I know, you cannot influence the format the caller's number is displayed in. --Ouro (blah blah) 06:37, 18 August 2009 (UTC)

Thanks for your reply!! Agreed but is it possible to have add ons or other third party software that can influence it to do that, I've been so restless ever since i have got this phone, only on this matter, otherwise the phone is doing great! Please let me know if there is any software or add ons for this matter. Thanks —Preceding unsigned comment added by 123.252.230.93 (talk) 10:44, 18 August 2009 (UTC)

Hard disk problem: Hitch, hitch, hitch

I have a PC which until 2 days ago had one SATA hard disk and one SATA DVD-RW drive. I've now installed a second SATA hard disk that's nearly identical to the first SATA hard disk. The BIOS reports that they are, respectively, the master on bus 1, the master on bus 5, and the master on bus 0. Upon starting up the computer, I noticed that upon displaying the Windows desktop for the first time, Vista would "hitch" continuously for about ten seconds, by which I mean that the mouse cursor and all other screen updates would lock in place; then about 1 second later the mouse cursor's position would update, then another 1-second hitch where everything's locked up, then another mouse cursor position update, then another 1-second hitch, etc. After around 10 seconds the system returned to operating completely smoothly. I then formatted the new hard disk and have mounted it as the E: drive. I get this hitching behavior for 10 or 20 seconds every time the system seems to survey what devices are connected: as in, when I double-click "My Computer", or use Vista's backup utility and it looks for target hard disks. I have not yet tried disconnecting any devices or swapping the cables around to different buses; I wanted to ask first whether this is known behavior of a bad hard disk, and if I should return the new drive. I seem to be able to write to it OK, and I ran the "Check Disk" function from the "Properties" window and it didn't report errors. Tempshill (talk) 20:31, 17 August 2009 (UTC)

Do you have the same file system type on both drives? (e.g. NTFS)? I've had trouble when mixing file system types. Nimur (talk) 21:31, 17 August 2009 (UTC)

Yes, both NTFS — my assumption has been that the file system could not be the issue, as the hitching was also seen to occur while the drive was installed but unformatted. Tempshill (talk) 21:36, 17 August 2009 (UTC)

AutoItv3 Script Adjustment

I was wondering if anyone could help me with a script i am trying to write in AutoItv3. It involves moving a GUI with WM_NCHITTEST. My script is here:

http://pastebin.com/m8c4afe7

Basically the script works and all, but i need it to update the variables $x and $y when the user drags the window. any ideas? Thank you for your time!

4.252.131.153 (talk) 21:28, 17 August 2009 (UTC)

strange packet delays

I have a program split across two linux (Ubuntu 8.10, I think) servers in the same rack, communicating by http via 1Gb ethernet through a switch in the rack. The http server end has to do some database operations and usually responds within 1-3 seconds but occasionally the response time is 12+ seconds. It is rarely higher than 3 seconds without being higher than 12 seconds, i.e. something is occasionally inserting a delay of 10 seconds or so. Any idea what this could be? I'm wondering if there is (for example) some tcp retransmission timeout in that range, i.e. the delay could be caused by some kind of tcp congestion within the rack. Grepping through the linux kernel constants doesn't find that (just the 3 second default RTO). I haven't yet found out whether there are congestion stats available from the switch (have to ask the ops guys about that). Does this sound like anything familiar to any of you? Thanks for any advice. 70.90.174.101 (talk) 21:37, 17 August 2009 (UTC)

Are you sure it's the network latency, and not (for example) swapping to disk or some other algorithmic slowdown that creates such a quantized jump? Nimur (talk) 21:49, 17 August 2009 (UTC)

Not sure at all. It could be anything. I'm casting around for ideas. 70.90.174.101 (talk) 21:50, 17 August 2009 (UTC)

You should consider instrumenting the server code with some profiling tools - at the very least, printing out timestamps and function calls to a log file for analysis. You can use a variety of free software performance analysis tools like gprof to generate graphical representations and find out which program elements are taking up the time. Nimur (talk) 23:34, 17 August 2009 (UTC)

TCP retransmission timeouts last milliseconds - not seconds. That's not it. More likely, you have an incipient hard drive failure and some disk activity is retrying over and over again. That's the only thing I could imagine that would cause random delays of that magnitude. Are there any disk retries in your kernel log? SteveBaker (talk) 01:53, 18 August 2009 (UTC)

You can detect TCP retransmits within your application with the getsockopt(2) system call -- see tcp(7) in the manual for details -- though this may be tricky if your app is not written in C. But yeah, profiling is more likely the right approach. If you're in C/C++, you might also look at using a linked-in profiler such as this one. --FOo (talk) 01:53, 18 August 2009 (UTC)

Thanks, I may try a profiler, but I don't think this is cpu activity (it happens randomly including with what should be lightweight queries). Steve's suspicion of the hard drive matches up with mine. I don't see any disk errors in the syslog, but we're using Seagate disks notorious for freeze-ups. I will look into upgrading the disk firmware. 70.90.174.101 (talk) 02:12, 18 August 2009 (UTC)

Need more info on what your doing. By database operations are you doing SQL insertions or queries etc? 1-3 seconds response seems rather terrible... you might have some underlying problems. —Preceding unsigned comment added by 155.144.40.31 (talk) 02:57, 18 August 2009 (UTC)

They are mostly SQL read queries but the server is heavily loaded and disk bound, so the 1-3 second response is par for the course. 70.90.174.101 (talk) 03:59, 18 August 2009 (UTC)

Run your program with "strace -o log.txt -T myapp" on both sides. It will output lines like:

08:53:28.483783 read(3, ""..., 1024)    = 0 <0.000011>
08:53:28.483844 close(3)                = 0 <0.000013>
08:53:28.483900 munmap(0xb7f88000, 4096) = 0 <0.000019>
08:53:28.484019 ioctl(1, SNDCTL_TMR_TIMEBASE or TCGETS, {B38400 opost isig icanon echo ...}) = 0 <0.000013>

where the bracketed numbers are the time spent in each system call. If it's a network issue, you should see a write() or send() that takes a long time and you can run ethereal/wireshark at the same time and correlate timestamps to see what the holdup is. If it's not a network issue, you can run strace with "-r" to get delta times on each successive call. If you see a large gap in the same place every time, either move on to ltrace, a debugger or your source code to understand what's going on there. If the gap happens at different places every time, suspect an external issue like a bad disk, system load, gremlins, etc. --Sean 13:04, 18 August 2009 (UTC)

Thanks!!! I have thought of that approach as being messy and complicated but I guess it's really not so bad. I will look into it. 70.90.174.101 (talk) 09:08, 19 August 2009 (UTC)

Transmit video signal 10m

I need to transmit a video signal (1024x786) from a PC to a projector over a distance of 10m. I can use either a VGA cable or a DVI (technically, a DVI cable with a DVI -> HDMI adapter at one end) cable. Would either of these work, and what would be better? How do meetings rooms and suchlike run the cabling to projectors (often mounted in celings, which can be quite a distance)? Thank you in advance! — QuantumEleven 22:01, 17 August 2009 (UTC)

I've never seen a 10 meter vga cable but maybe they exist. There are certainly hdmi cables that long from the usual vendors. Similarly for cable routing, there's all the usual conduits, cable ties, and such gizmos. See cable management. Another approach is to put a small computer (e.g. Mac Mini) close to the projector and run the projector from it. You'd communicate with the small computer by wifi. 70.90.174.101 (talk) 22:24, 17 August 2009 (UTC)

I haven't ever used a VGA cable that long, either. There are powered VGA amplifiers that can boost the signal for long stretches. I don't know if there's such a thing as a DVI amplifier. Tempshill (talk) 22:32, 17 August 2009 (UTC)

I have personally used VGA cables and DVI cables of 50 foot length (about 15 meters). The DVI shows zero signal degradation; the VGA shows virtually no signal degradation. The DVI cables were about $50 or $60 on the web. Nimur (talk) 22:36, 17 August 2009 (UTC)

DVI-D - being digital - is an all-or-nothing thing. Either it works and you get perfect picture quality - or it doesn't and it fails utterly...or at least is disasterously intermittant. Both DVI-A and VGA are analog systems - and they are a much poorer choice - the longer the cable, the worse the picture quality - period. SteveBaker (talk) 01:50, 18 August 2009 (UTC)