Computing desk
< September 23	<< Aug \| September \| Oct >>	September 25 >

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.

September 24

Why there are no android app with all 105 keys of IBM keyboard?

really feeling the lack of pageup, pagedown and delete. — Preceding unsigned comment added by 2A03:2880:3010:BFFA:FACE:B00C:0:1 (talk) 04:05, 24 September 2015 (UTC)[reply]

Have you tried Hacker's Keyboard? (link to Google Play Store). —Noiratsi (talk) 16:17, 24 September 2015 (UTC)[reply]

COPY VCD TO PEN DRIVE

My friend has an old video CD and he wants to copy the video to a pen drive to preserve the video. It has the following folders and files. CDI(CDI_IMAG, CDI_TEXT.FNT, CDI_VCD.APP, CDI_VCD.CFG), EXT(LOT_X.VCD, PSD_X.VCD, SCANDATA.DAT), MPEGAV(AVSEQ01.DAT), SEGMENT, VCD(ENTRIES.VCD, INFO.VCD, LOT.VCD, PSD.VCD). How can he copy the video to pen drive? Thank you.175.157.41.176 (talk) 05:12, 24 September 2015 (UTC)[reply]

From memory, and it has been a while, there should be one or two files that are clearly larger than all the rest (possibly the DAT file?), that's all a media player like VLC needs to play the video. Just put that (those) file(s) on the pen drive and try to play it, if it works you are done. Vespine (talk) 06:09, 24 September 2015 (UTC)[reply]

According to this, the video data is in AVSEQnn.DAT file(s). Ssscienccce (talk) 11:42, 24 September 2015 (UTC)[reply]

Also from memory, but yes, that's correct. The AVSEQ01.DAT is simply an MPG file, so you can even rename the extension to MPG to make it easier to classify if you'd like. 64.235.97.146 (talk) 15:52, 24 September 2015 (UTC)[reply]

This likely won't work, because Video CD uses Mode 2 Form 2 sectors with more than the usual 2048 bytes of user data. As far as I know, the CD-ROM and filesystem drivers of major OSes don't support this; they will copy just 2048 bytes per sector and the result will be unplayable. You probably need a dedicated Video CD ripping/transcoding program, but I don't know what to recommend. -- BenRG (talk) 02:37, 27 September 2015 (UTC)[reply]

Thank you all for the advice.175.157.12.52 (talk) 02:46, 25 September 2015 (UTC)[reply]

What is the most powerful lossless compression setting for video created upto now?

OP was tinkering with various compression settings, codecs, color depths so on, suddenly this thought. — Preceding unsigned comment added by 2A03:2880:3010:BFF8:FACE:B00C:0:1 (talk) 06:49, 24 September 2015 (UTC)[reply]

FFV or https://en.wikipedia.org/wiki/Huffyuv 217.158.236.14 (talk) 15:02, 24 September 2015 (UTC)[reply]

I said SETTING, NOT CODEC ONLY. that means encoder+color depth+audio sample rate+color scheme+additional tweaks. I want the whole setting. — Preceding unsigned comment added by 2A03:2880:3010:BFF5:FACE:B00C:0:1 (talk) 16:33, 24 September 2015 (UTC)[reply]

Well, if you change colour depth, or audio sample rate, or most of the rest, it's not lossless, is it? Reduced the colour depth -> lose colour information.--Phil Holmes (talk) 17:17, 24 September 2015 (UTC)[reply]

For humans, audio and video is analog. Any digitized form of real audio/video will be lossy. By lossless, we mean that you compress the DIGITAL information and don't lose anything. The content remains the same. For audio/video, the content is the audio and video. If you change the audio by changing sampling rate or you change the video by changing the color depth, you no longer have the same content. There is no way that can be lossless. The term lossless will only apply to the compression of the codec. 209.149.113.66 (talk) 14:08, 25 September 2015 (UTC)[reply]

There is no possible lossless compression algorithm that works for every possible input stream - that's a mathematical impossibility. But for almost all real world video there is considerable redundancy that can be exploited and you could come up with algorithms that would reduce the size of any real-world movie...but such an algorithm would definitely make the data larger when presented with random noise or something weird that was designed specifically to trip it up. As a 'thought experiment', you could (for example) split your uncompressed image file into individual frames, then convert each one into a PNG file. PNG is a lossless compression scheme - and while there are certainly images out there that it can't squeeze any data out of - those are not likely to occur in a 'real' movie. In practice, there is often a lot of similarity between consecutive frame that could be exploited to do much better than that - but I don't know of any lossless compression for video that you can just pick up and use. 19:15, 25 September 2015 (UTC)

Replacement wireless card for old notebook?

So I think the wireless adapter is gone. Are these parts standardized or are they specific to individual models or makes. This is the part I'd like to try and replace...

https://www.google.co.uk/search?q=BCM94312MCG&es_sm=122&source=lnms&tbm=isch&sa=X&ved=0CAcQ_AUoAWoVChMIpf6nw9KPyAIVBS4aCh0pgAZn&biw=1344&bih=724 — Preceding unsigned comment added by 80.195.27.47 (talk) 12:17, 24 September 2015 (UTC)[reply]

Does it have a USB port? There are plenty of USB wifi cards that are barely bigger than the USB port, see here for example. RegistryKey^(RegEdit) 16:23, 24 September 2015 (UTC)[reply]

It's using a mini PCI express connector, which a standard in laptops. If the device wasn't a WiFI card, you could safely replace it with a new one. The only reservation I'd have about replacing a WiFi card is if the antennae needs to be impedance-matched to the WiFi card. LongHairedFop (talk) 11:12, 26 September 2015 (UTC)[reply]

Some manufacturers have restricted the use of 3rd party mini-pci-e cards by reading the hardware/vendor-ID from BIOS/UEFI before bootup. Not a nice move for customers by the manufacturer. --Hans Haase (有问题吗) 13:19, 26 September 2015 (UTC)[reply]

Can a user patch a security flaw in proprietary software?

Can a user patch a security flaw in proprietary software? Or, is his only alternative to wait until the software developer patches it? --Bickeyboard (talk) 23:16, 24 September 2015 (UTC)[reply]

Do you have access to the source code? Do you know how to fix it? A Quest For Knowledge (talk) 23:18, 24 September 2015 (UTC)[reply]

No, I am thinking about something like Windows XP. Can any company (that's not MS) offer updates on it? I suppose MS won't release an open-source version of it? Can an add-on be implemented to correct a malfunctioning closed-source proprietary software? --Bickeyboard (talk) 23:22, 24 September 2015 (UTC)[reply]

It is theoretically possible to modify a program in binary form; but in normal cases, this type of modification is prohibitively difficult.

However, some security problems can be resolved without modifying the flawed program: for example, a user can make a configuration change, isolate sensitive systems or data, and so on. Such corrective action could be provided as a service by an external vendor other than the original system software vendor. Nimur (talk) 23:54, 24 September 2015 (UTC)[reply]

Asking "Can a user patch a security flaw in proprietary software?" is like asking "Can a man climb Mount Everest?". The answer is "Technically yes but practically no!" 175.45.116.61 (talk) 00:04, 25 September 2015 (UTC)[reply]

Weird example since many men have climbed Everest. StuRat (talk) 04:39, 25 September 2015 (UTC) [reply]

This answer is simply wrong. Why would it be practically impossible to climb the everest? Just say that you don't believe that binary programs can be modified. --Scicurious (talk) 04:54, 25 September 2015 (UTC)[reply]

I'd bet that about as many men have climbed Everest as have patched Windows at the binary level without access to the source code. It's a great analogy! SteveBaker (talk) 19:05, 25 September 2015 (UTC)[reply]

I presume the OP was thinking of averages. However intentionally or not, perhaps the OP's example does have some parallels. In particular, it's sometimes suggested that nowadays anyone can climb Everest if they have enough money [1] [2]. Similarly, if you're willing to pay enough, you can probably get someone to patch a security flaw in proprietary software without the source code.

However if the source code does exist, it may be more effective to pay whoever has the source code to either do it themselves or let someone else do it. Banks and some other major customers are doing or did do this with XP for example [3] [4] [5] [6]. ATMs for example are often still using XP, and while some XP Embedded support is until 2016 and some until 2019 [7] [8], not everyone is using it.

Where the OPs parallel may fail is that you generally aren't going to die from trying to patch your propietary software (whether mostly by your own efforts or someone you paid). Also, although there may be controversy over some people possibly treating Everest as an expensive tourist trip, as the earlier sources also say plenty of people do put in a lot of effort themselves when you can expect this won't be the case for a lot of proprietary software. Finally if you did want to pay money, you're always going to be paying someone else, you can't pay Everest to help you climb it.

Nil Einne (talk) 07:12, 25 September 2015 (UTC)[reply]

I was once given the ROM of an embedded device and asked to fix a bug in it (the device wasn't 'sleeping' when not in use, so it ate batteries at an alarming rate). I had a PROM-programmer that could extract the data from the ROM and give it to me as a file on disk. The embedded machine had only about 8k bytes of program - so I used a "disassembler" to turn the binary into somewhat-human-readable machine code mnemonics...no variable names, no labels, no comments. By pure luck, figuring out how the code worked to the point where I could find and fix the bug turned out to be relatively easy - and I was able to re-assemble the code, program a new ROM and plug it back into the original device.

I don't recommend doing that - and the problem gets exponentially harder as the size of the code increases. If the code has been written by a high level language compiler with optimisation turned on, it gets a LOT tougher.

It's certainly not impossible in theory - for very small programs, it's certainly possible in practice - but for something the size of MS Windows - it would be a practical impossibility without resources considerably larger than the Microsoft team who originally wrote it. SteveBaker (talk) 19:05, 25 September 2015 (UTC)[reply]