less_retarded_wiki/digital.md
2023-11-14 22:25:01 +01:00

15 lines
5.6 KiB
Markdown

# Digital
Digital [technology](tech.md) is that which works with [whole numbers](integer.md), i.e. discrete values, as opposed to [analog](analog.md) technology which works with [real numbers](real_number.md), i.e. continuous values (note: do not confuse things such as [floating point](float.md) with truly continuous values!). The name *digital* is related to the word *digit* as digital computers store data by digits, e.g. in 1s and 0s if they work in [binary](binary.md). By extension the word *digital* is also used to indicate something works based on digital technology, for example "digital currency", "digital music" etc.
Normies confuse digital with [electronic](electronic.md) or think that digital computers can only be electronic, that digital computers can only work in [binary](binary.md) or have other weird assumptions whatsoever. **This is indeed false!** An [abacus](abacus.md) is a digital device, a book with text is a digital data storage. Fucking normies RIP.
{ Apparently it is "digitisation", not "digitalization". ~drummyfish }
The advantage of digital technology is its resilience to [noise](noise.md) which prevents degradation of data and accumulation of error -- if a digital picture is copied a billion times, it will very likely remain unchanged, whereas performing the same operation with analog picture would probably erase most of the information it bears due to loss of quality in each copy. Digital technology also makes it easy and practically possible to create fully programmable general purpose [computers](computer.md) of great complexity.
**Digital vs analog, simple example:** imagine you draw two pictures with a pencil: one in a normal fashion on a normal paper, the other one on a grid paper, by filling specific squares black. The first picture is analog, i.e. it records continuous curves and position of each point of these curves can be measured down to extremely small fractions of millimeters -- the advantage is that you are not limited by any grid and can draw any shape at any position on the paper, make any wild curves with very fine details, theoretically even microscopic ones. The other picture (on a square grid) is digital, it is composed of separate points whose position is described only by whole numbers (*x* and *y* coordinates of the filled grid squares), the disadvantage is that you are limited by only being able to fill squares on predefined positions so your picture will look blocky and limited in amount of detail it can capture (anything smaller than a single grid square can't be captured properly), the [resolution](resolution.md) of the grid is limited, but as we'll see, imposing this limitations has advantages. Consider e.g. the advantage of the grid paper image with regards to copying: if someone wants to copy your grid paper image, it will be relatively easy and he can copy it exactly, simply by filling the exact same squares you have filled -- small errors and noise such as imperfectly filled squares can be detected and corrected thanks to the fact that we have limited ourselves with the grid, we know that even if some square is not filled perfectly, it was probably meant to be filled and we can eliminate this kind of noise in the copy. This way we can copy the grid paper image a million times and it won't change. On the other hand the normal, non-grid image will become distorted with every copy and in fact even the original image will become distorted by aging; even if that who is copying the image tries to trace it extremely precisely, small errors will appear and these errors will accumulate in further copies, and any noise that appears in the image or in the copies is a problem because we don't know if it really is a noise or something that was meant to be in the image.
Of course, digital data may become distorted too, it is just less likely and it's easier to deal with this. It for example happens that space particles (and similar physics phenomena, e.g. electronic interference) flip bits in computer memory, i.e. there is always a probability of some [bit](bit.md) flipping from 0 to 1 or vice versa. We call this **data [corruption](corruption.md)**. This may also happen due to physical damage to digital media (e.g. scratches on the surface of CDs), imperfections in computer network transmissions (e.g. packet loss over [wifi](wifi.md)) etc. However we can introduce further measures to prevent, detect and correct data corruption, e.g. by keeping [redundant](redundancy.md) copies (2 copies of data allow detecting corruption, 3 copies allow even its correction), keeping [checksums](checksum.md) or [hashes](hash.md) (which allow only detection of corruption but don't take much extra space), employing error correcting codes etc.
Another way in which digital data can degrade similarly to analog data is **reencoding between lossy-[compressed](compression.md) formats** (in the spirit of the famous "needs more jpeg" [meme](meme.md)). A typical example is digital movies: as new standard for video encoding are emerging, old movies are being reconverted from old formats to the new ones, however as video is quite heavily lossy-compressed, losses and distortion of information happens between the reencodings. This is best seen in videos and images circulating on the internet that are constantly being ripped and converted between different formats. This way it may happen that digital movies recorded nowadays may only survive into the future in very low quality, just like old analog movies survived until today in degraded quality. This can be prevented by storing the original data only with lossless compression and with each new emerging format create the release of the data from the original.