ph@miro.Berkeley.EDU (Paul Heckbert) (01/04/89)
I've been troubled lately to see people using the word "bitmap" to mean a grayscale or color image. As long as I can remember, however, "bitmap" has meant a ONE BIT PER PIXEL, black and white image. Where is this new, (and I believe) incorrect usage coming from? Several contributors to this news group have used "bitmap" as a synonym for "image" yet I don't think I've ever seen that usage in a book, journal article, or software documentation. Does anybody know the etymology of this bad new usage? I hope the new usage doesn't catch on, because the old, correct meaning of "bitmap" fills a niche, and the new meaning is redundant (we don't need another synonym for "image" and "picture"). If you see people misusing the word "bitmap", correct them! Paul Heckbert, CS grad student 508-7 Evans Hall, UC Berkeley UUCP: ucbvax!miro.berkeley.edu!ph Berkeley, CA 94720 ARPA: ph@miro.berkeley.edu
ksbooth@watcgl.waterloo.edu (Kelly Booth) (01/05/89)
In article <8568@pasteur.Berkeley.EDU> ph@miro.Berkeley.EDU (Paul Heckbert) writes: >I've been troubled lately to see people using the word "bitmap" to mean >a grayscale or color image. As long as I can remember, however, "bitmap" >has meant a ONE BIT PER PIXEL, black and white image. > >Where is this new, (and I believe) incorrect usage coming from? (a) You are correct that "bitmap" often means single-bit-per-pixel. (b) The more general usage appears in the Guibas-Stolfi TOG 1:3 paper from 1982. "bitmaps: arrays of discrete intensity/color values". In skimming through the article again, I cannot find a place where "bitmap" is defined exactly (the quote above is not definitive because "discrete" could be taken as 0-1, finessing the issue in question, though I'm sure it is intended to mean discrete in precisely the sense used in Paul Heckbert's posting a few days ago). But in section 5.1, it is stated that the descriptor for a bitmap includes the depth (number of bits per pixel). Elsewhere in the article most of the examples seem to assume only 0-1 valued pixels, but there are numerous allusions to pixels having depth other than 1.
elf@dgp.toronto.edu (Eugene Fiume) (01/05/89)
In my paper on the semantics of bit-mapped graphics (and other topics)
(computers&graphics 11, 2 (Apr. 1987)), I define bit-maps as Paul
Heckbert suggests, but just because I define something to be so doesn't
make it so.
Both Foley and van Dam, and Newman and Sproull are fairly cagey about
bit-mapping. It's hard to get a straight answer as to whether or not
they intend bit-maps to be bi-level, since they are discussed in the
context of general raster graphics. The point is that if the standard
texts don't define such concepts carefully, ambiguity is bound to
spread.
Example: The index entry for bit-mapped graphics in Foley and van Dam
reads
Bit map, see Raster display, refresh buffer
and the index entry for refresh buffer is oblivious to bit-maps. Now,
if you chase down the page references for "refresh buffer", you will
see that p21 refers to bit-maps, and does not say that the colour space
must be bi-level. In fact, in subsequent references, it becomes clear
that pixels within bit-maps do not need to be bi-level. But then they
define boolean operations on bit-maps, which only make (proper) sense
when pixels are bi-level. (Yes, yes, I know that one can try to make
sense of some "boolean" operations on n-bit pixels, n>1.)
It is clear that a word is needed for bi-level intensity functions
f: Z sup 2 -> {0,1} in which there is a one-one correspondence between
f and an image capable of displaying bi-level pixels. For me, that word
is bit-map. Note that I have distinguished between bit-maps and images
(e.g., for me a bit-map does not have a pixel aspect ratio or shape--
images do).
--
Eugene Fiume
Dynamic Graphics Project
University of Toronto
elf@dgp.toronto.edujbm@eos.UUCP (Jeffrey Mulligan) (01/05/89)
From article <8568@pasteur.Berkeley.EDU>, by ph@miro.Berkeley.EDU (Paul Heckbert): > I've been troubled lately to see people using the word "bitmap" to mean > a grayscale or color image. As long as I can remember, however, "bitmap" > has meant a ONE BIT PER PIXEL, black and white image. And so a bit-mapped display is a display which displays a bitmap. A possible confusion might arise when distinguishing digital raster frame buffer type displays from vector plotting displays (like the old E&S picture system). Perhaps the former would more properly be called a "byte-mapped display!?" -- Jeff Mulligan (jbm@aurora.arc.nasa.gov) NASA/Ames Research Ctr., Mail Stop 239-3, Moffet Field CA, 94035 (415) 694-6290
myers@hpfcdj.HP.COM (Bob Myers) (01/05/89)
>I hope the new usage doesn't catch on, because the old, correct meaning of >"bitmap" fills a niche, and the new meaning is redundant (we don't need >another synonym for "image" and "picture"). If you see people misusing the >word "bitmap", correct them! Gee, I dunno - I agree with your explanation of the origin of "bitmap", i.e., a one-bit-per-pixel representation of an image, but I'm not at all sure that the "new" context ("Hey, you got any new 24-bit bitmaps?") isn't filling a need. The words "image" and "picture" do not, at least to me, convey any idea of a fixed resolution, whereas we all recognize "bitmap" as meaning, say, an "image" which has been scanned (or generated or whatever) at, say, 1280x1024x16 (your numbers may vary). "Bytemap" or "wordmap" (same disclaimer) may be more technically accurate for this concept, but I think that they're too clumsy to gain much acceptance. Bob Myers KC0EW HP Graphics Tech. Div.| Opinions expressed here are not Ft. Collins, Colorado | those of my employer or any other {the known universe}!hplabs!hpfcla!myers | sentient life-form on this planet.
darin@nova.laic.uucp (Darin Johnson) (01/05/89)
In article <8568@pasteur.Berkeley.EDU> ph@miro.Berkeley.EDU (Paul Heckbert) writes: >I've been troubled lately to see people using the word "bitmap" to mean >a grayscale or color image. As long as I can remember, however, "bitmap" >has meant a ONE BIT PER PIXEL, black and white image. > >Where is this new, (and I believe) incorrect usage coming from? Since many grayscale and color "images" are created from layered bitmaps. If I have 8 bitmaps, and stick them into consecutive chunks of memory, it is easier to call it a bitmap, than "some bitmaps". (instead of a 1-to-1 mapping to real display elements, it is a 1-to-1 mapping to imaginary and layered display elements) However, for some computers than don't use layered bitmaps to form pictures (such as byte-per-pixel), the term bitmap is misleading. On the whole though, I don't think it causes much confusion. Especially considering that "bitmap" has meant things other than images for as long as I can remember - allocation maps for disks, etc. I tend to think of a bitmap as a collection (regular array - any dimension) of bits which regulary map from the bits to something else - a display, disk blocks, etc. The word "bitmap" has probably been around longer than bitmap displays. Darin Johnson (leadsv!laic!darin@pyramid.pyramid.com) "You can't fight in here! This is the war room.."
alain@dgp.toronto.edu (Alain Fournier) (01/05/89)
I agree with Paul Heckbert that "bit-map" should be kept for 1 bit deep arrays (by the way Rob Pike, ToG 2,2, (April 1983), and Eugene Fiume, Computers and Graphics, 11, 2, April 1987, have interesting distinctions between the bit-map as an array in memory vs what is on the screen). That leaves us without a good word for multi-bit-maps. My modest proposal is "swatch". The word means of course a sample of cloth or other material, usually variegated, which is appropriate. It is also short. PS: It is not to be confused with "swath". Actually one of my dictionaries defines "to cut a wide swath" as "to make a pretentious display", so it also has its use in computer graphics. And remember: a neologism a day keeps the spell-checker at bay.
paul@hpldola.HP.COM (Paul Bame) (01/05/89)
I can guess at the reason for the misuse of the word bitmap since I've seen the same thing happen locally. It happened when the Po--- collection of bitmaps were posted and became popular here as X backgrounds. People saw "pictures" on their screen which were in fact *bitmaps* in the classic bi-level sense but the image and bitmap quickly became synonyms - unfortunately. When I get the time, I find a memorable cure is to show a bitmap and an image (pixmap :-)) of the same subject at the same time and explain the reason and terminology behind the drastic difference in appearance. --Paul Bame HP Colorado Springs hplabs!hpldola!paul paul@hpldola.hp.com 719 590 5557
nelson@sun.soe.clarkson.edu (Russ Nelson) (01/05/89)
The X Window System uses the term pixmap. Once you hear that word and know what it means, bitmap is hard to interpret as anything but a "map of bits", i.e. ones and zeroes. -russ -- --russ (nelson@clutx [.bitnet | .clarkson.edu]) "I saved the whales!" - Rebecca L. Nelson, 3.5 years old, on receiving her Christmas present of a whale "adoption" certificate. Bless her liberal heart.
cdshaw@alberta.UUCP (Chris Shaw) (01/06/89)
In article X alain@dgp.toronto.edu (Alain Fournier) writes: >I agree with Paul Heckbert that "bit-map" should be kept for 1 bit deep arrays. Me too. The reason being that "bit-map" implies that each bit in the map is a full citizen of equal significance. In a monochrome raster with n-bit pixels, each pixel has a number of bits of different significance: i.e. inverting the most significant bit in a pixel has a different effect than flipping the least significant bit in a pixel. Therefore, each bit in an n-bit pixel is not a full citizen and one should therefore refrain from calling such a raster a "bit-map". Or, as Mark Green says, a bit-map is a folded sheet of paper you get at the auto club to help you find your bits. >That leaves us without a good word for multi-bit-maps. My modest proposal >is "swatch". The word means of course a sample of cloth or other material, >usually variegated, which is appropriate. It is also short. Yes, but how about "raster"? It overloads the term a little, but not too much. So one could say "n-bit monochrome raster", and be immediately understood. -- Chris Shaw cdshaw@alberta.UUCP (or via watmath or ubc-vision) University of Alberta CatchPhrase: Bogus as HELL !
childers@avsd.UUCP (Richard Childers) (01/06/89)
In article <8568@pasteur.Berkeley.EDU> ph@miro.Berkeley.EDU (Paul Heckbert) writes: >I've been troubled lately to see people using the word "bitmap" to mean >a grayscale or color image. As long as I can remember, however, "bitmap" >has meant a ONE BIT PER PIXEL, black and white image. I would think the word is actually a compound word : bit - the smallest *meaningful* quanta in the data set referred to map - a set of data internally relevant, presumed to be relevant to external reality as well bitmap - a set of data using the 'bit' as the smallest unit of data >Where is this new, (and I believe) incorrect usage coming from? Computer scientists, I guess. (-: >Several contributors to this news group have used "bitmap" as a synonym for >"image" yet I don't think I've ever seen that usage in a book, journal article, >or software documentation. You'd prefer 'bytemap' ? Do we need to confuse everyone all over again ? >I hope the new usage doesn't catch on, because the old, correct meaning of >"bitmap" fills a niche, and the new meaning is redundant (we don't need >another synonym for "image" and "picture"). If you see people misusing the >word "bitmap", correct them! Since the meaning will continue to be extracted by a bit-by-bit analysis of the data set - admittedly, the bits are now interpreted as groups, but that, too, is in flux, as the size of color tables increases - I think 'bitmap' is appropriate. If you're correct, however, then even using it to refer to a grayscale image is incorrect. >Paul Heckbert, CS grad student >508-7 Evans Hall, UC Berkeley UUCP: ucbvax!miro.berkeley.edu!ph -- richard -- * Supernovae are a blast * * * * ..{amdahl|decwrl|octopus|pyramid|ucbvax}!avsd.UUCP!childers@tycho * * AMPEX Corporation - Audio-Visual Systems Division, R & D *
shawn@pnet51.cts.com (Shawn Stanley) (01/07/89)
Oh my. I don't see why so many people have decided that the term "bitmap"
must always refer to a one-to-one correspondence of bit to pixel. The word
"bitmap" is like many other words; if you take it apart, it means a mapping of
some feature to a particular bit representation. I think attempting to
confine "bitmap" to black & white, or on & off pixel representation is asking
everyone tconfine themselves to B&W, while the technology has obviously gone
beyond that.
Oh well. Term wars are the worst. You can find any use you want in a book,
and the only means of supporting it seems to be stating the author's name.
With that in mind, I'll just take a glance here at the glossary of [unnamed
book] by [unnamed author]. Maybe it provides grounds for agreement.
Bit field: When a byte or word is viewed as a bit pattern, several bits taken
together may hold a particular item of information.
Bit plane: On EGA, the video buffer is divided into four sections, referred to
as bit planes 0 - 3. In 16-color modes the four planes are in parallel, so
that four bytes are located at a particular memory address (the latch
registers intermediate movement of data between the CPU and video memory). In
some cases the planes may be CHAINED, that is, they are combined into one or
two larger planes.
So, can we say "bitmap", "bit field", and "bit plane"?
UUCP: {rosevax, crash}!orbit!pnet51!shawn
INET: shawn@pnet51.cts.comkchen@Apple.COM (Kok Chen) (01/07/89)
In article <375@avsd.UUCP> childers@avsd.UUCP (Richard Childers) writes: > > bit - the smallest *meaningful* quanta in the data set referred to The word BIT was first suggested by J. Tukey (yes, same one the Cooley-Tukey FFT refers to) as a contraction for BINARY-digit. For a reference to that, refer to C.E. Shannon, "A mathematical theory of Communication," BSTJ, July and October, 1948. The modern rigorous definition is based on the unit used to denote Entropy. When the logarithm used is base-2, the unit is in BITS, if it is base-e, the unit is NATS (Natural digits) and when it is base-10, it is in DITS (Decimal digits). So, please don't use BITS to denote units of information that a single two-state flip-flop cannot represent. You will only confuse Information Theorists (i.e., add to Entropy :-). If you really want to bastardize the mathematical definitions, you might want to use BYTS :-) :-) Afterall, on a DEC-10 (20), a byte is anything between 1 bit and 36 bits. Then, you can go ahead and coin the next bastardization - a BYTMAP. :-) :-) Kok Chen {decwrl,sun}!apple!kchen Apple Computer, Inc.
myers@hpfcdj.HP.COM (Bob Myers) (01/07/89)
>Yes, but how about "raster"? It overloads the term a little, but not too >much.So one could say "n-bit monochrome raster", and be immediately understood. NO! A "bitmap", "pixmap", "swatch", or whatever the hell you want to call it is a representation of an image, in digital form, in the frame buffer memory. A "raster" is only one thing, and it's something completely different - it's the visible effect of raster-scanning the phosphor of a CRT with an electron beam. I'm picky about this because (a), for the last few years I've been deeply involved in the CRT displays here, but also (and more importantly) (b), there are several ways to "display" a "bitmap" (insert your favorite word there) that do not involve a raster-scanned CRT monitor. For example, would you call the image produced by a color ink-jet printer a "raster"? Bob Myers KC0EW HP Graphics Tech. Div.| Opinions expressed here are not Ft. Collins, Colorado | those of my employer or any other {the known universe}!hplabs!hpfcla!myers | sentient life-form on this planet.
peter@ficc.uu.net (Peter da Silva) (01/07/89)
I don't know. I generally use 'bitplane' and 'image' when talking about raster graphics, and I'm used to 'bitmaps' having something to do with disk allocation strategies and such. I guess I haven't been paying attention. -- Peter da Silva, Xenix Support, Ferranti International Controls Corporation. Work: uunet.uu.net!ficc!peter, peter@ficc.uu.net, +1 713 274 5180. `-_-' Home: bigtex!texbell!sugar!peter, peter@sugar.uu.net. 'U` Opinions may not represent the policies of FICC or the Xenix Support group.
mike@ai.etl.army.mil (Mike McDonnell) (01/07/89)
In article <8901050417.AA16210@explorer.dgp.toronto.edu> alain@dgp.toronto.edu (Alain Fournier) writes: >I agree with Paul Heckbert that "bit-map" should be kept for 1 bit deep arrays >(by the way Rob Pike, ToG 2,2, (April 1983), and Eugene Fiume, Computers and >Graphics, 11, 2, April 1987, have interesting distinctions between the bit-map >as an array in memory vs what is on the screen). >That leaves us without a good word for multi-bit-maps. My modest proposal >is "swatch". [some stuff deleted] I like "pixmap". This is used to mean a multilevel image in the X Window System and is therefore probably on its way to becoming a standard term already since X is becoming big. X also uses "bitmap" for a 1-bit displayable array. Pixmap must, of course, be qualified when it is used, as in "8-bit pixmap". -- Mike McDonnell at the U.S. Army Engineer Topographic Laboratories, Bldg. 2592 Fort Belvoir, VA 22060-5546 TEL:(202)355-2716 NET: mike@ai.etl.army.mil
bouma@cs.purdue.EDU (William J. Bouma) (01/07/89)
bitmap
1 bit map
8 bit map
24 bit map
bytemap
3 byte map
wordmap
Everyone knows that a "bit" is one binary digit.
Most people are sure a "byte" is eight binary digits.
Nobody knows how big a "word" is.
Assume you need a word to mean a list of binary digits which represent a picture
when shoved into the screen memory of some machine, n-bits per memory word. What
is the clearest and most logical way to describe this bunch of data? "n-bitmap".
"Bytemap" is certainly no more correct than "bitmap" and will add a layer of
confusion as to the size for some people. When I say "bitmap", you understand
me, no? When I say "480x480-24bitmap" you understand me even better? So where
is the problem????
To me an "image" consists of a bitmap plus other information like size and/or
a color table. But if someone were to use the word "bitmap" to refer to what
I would call an "image", I would still understand them and not accuse them of
destroying the english language since the most relevant part of an image is the
bitmap after all. So let's get back to talking graphics, not semantics!!!
--
Bill <bouma@cs.purdue.edu> || ...!purdue!bouma julian@riacs.edu (Julian E Gomez) (01/07/89)
Like others have pointed out, "pixmap" seems like the best bet. We're already used to asking how many bits in a pixel, whereas we're not used to asking how many bits in a bit. -- "Have you ever wondered if taxation without representation was cheaper?" Julian "a tribble took it" Gomez julian@riacs.edu
sow@ulmo1.mt.luth.se (Sven-Ove Westberg) (01/08/89)
In article <1198@hydra.riacs.edu> julian@hydra.riacs.edu.UUCP (Julian E Gomez) writes: | |Like others have pointed out, "pixmap" seems like the best bet. We're |already used to asking how many bits in a pixel, whereas we're not used |to asking how many bits in a bit. Pixmap is NOT the term we should use for this type of data structures. From ISO/DIS 8613-7, "Information processing - text and office systems - Office Document Architecture (ODA) and interchange format - Part 7: Raster graphics content architectures. pel array: A two-dimensional array of pels (Picture Elements) used to represent a pictorial image. bitmap: A two- or three-dimensional data field representing a pel array. In section 9.1 "Bitmap encoding scheme" Each element in a pel array may have one of two distinct states. These are the set state, corresponding to foreground colour and the unset state, corresponding to the background colour. For the purpose of representing such an array within a text unit, each pel may be represented by a single bit which has the value '0' or '1' depending on the state of that pel. ..... If the definition in the iso draft is valid then a bitmap can have more then two distinct states. The third dimension. But later in the encoding scheme it is clear that a bitmap is a pel array with only two distinct states. It is clear that a colour image is a pel array. But not what a bitmap is. So let us use pel array. Sven-Ove Westberg, CAD, University of Lulea, S-951 87 Lulea, Sweden. UUCP: {uunet,mcvax}!enea!cad.luth.se!sow ARPA: sow%cad.luth.se@ucbvax.berkeley.edu (only dumb ARPA mailers) Internet: sow@cad.luth.se Bitnet: sow%cad.luth.se@sekth
t-stephp@microsoft.UUCP (Stephen Poole) (01/10/89)
In article <NELSON.89Jan5101307@sun.soe.clarkson.edu> nelson@clutx.clarkson.edu writes:
The same term applies in the Mac world; it was introduced along with IM volume
V and the Mac II. I have seen several programs that refer to pixmaps on the
Mac II, so perhaps it is being used with degree of consistency.
--
-- Stephen D. Poole -- t-stephp@microsoft.UUCP -- Mac II Fanatic --
-- --
-- I'm just an Oregon Tech Software Engineering co-op at Micro- --
-- soft. Believe me, nobody here pays attention to my opinions! --anw@nott-cs.UUCP (01/11/89)
In article <8901050417.AA16210@explorer.dgp.toronto.edu> alain@dgp.toronto.edu (Alain Fournier) writes: >That leaves us without a good word for multi-bit-maps. Thereby contradicting himself. > My modest proposal >is "swatch". Sadly, "swatch" is already in use by the Swiss watch industry. Similarly, my non-computerite friends think that a VAX is a (much advertised in the UK) brand of vacuum cleaner [also a self-contradiction, how could a vacuum be dirty?]. -- Andy Walker, Maths Dept., Nott'm Univ., UK. anw@maths.nott.ac.uk
cme@cloud9.Stratus.COM (Carl Ellison) (01/11/89)
In article <1304@luth.luth.se>, sow@ulmo1.mt.luth.se (Sven-Ove Westberg) writes: > > Pixmap is NOT the term we should use for this type of data structures. > > From ISO/DIS 8613-7, "Information processing - text and office systems - > Office Document Architecture (ODA) and interchange format - > > pel array: A two-dimensional array of pels (Picture Elements) used to > represent a pictorial image. > > bitmap: A two- or three-dimensional data field representing a pel array. > > In section 9.1 "Bitmap encoding scheme" > > Each element in a pel array may have one of two distinct states. > These are the set state, corresponding to foreground colour and the > unset state, corresponding to the background colour. For the purpose of > representing such an array within a text unit, each pel may be > represented by a single bit which has the value '0' or '1' depending > on the state of that pel. ..... > There are two worlds trying to intercommunicate here. Each has its own dictionary and, unfortunately, citizens of each fail to identify their worlds -- perhaps believing that it's all one world called Computer Graphics. One world came from a background of alphanumeric CRT terminals which evolved into one bit per pixel directly addressable frame stores -- and, voila', people called that Graphics Mode. In this world (the one quoted above and the one using "bitmap"), the primitive graphics operation was the storing of a value into a pixel (originally setting or clearing of a bit; later setting a 4-bit color code (looked up in a color map for display)). Hand someone from this world a 24-bit per pixel RGB display and s/he'll set 24 bit pixel values and declare that that's "doing graphics". In that world it is. The other world (mine) came from a background of image processing and vector algebra. To us, (eg., E&S in the old days), any direct manipulation of a pixel is not graphics and a screen backed up by a single bit per pixel (eg., a Mac or Sun monochromatic screen) is not a graphics output device. To us, graphics is the creation of images from mathematically defined objects in a vector space -- especially via transformation, projection, .... In this world, there are no operations directly on individual pixels and the only raster output worth its salt has to be filtered (therefore multiple bits per color at each pixel). In my world, "bitmap" (or "pixmap") is a foreign term. It comes from those other folks who directly manipulate the pixels on screen. Meanwhile, to me as a mathematician, both "bitmap" and "pixmap" are improper because the word "map" makes no sense. In both math and computer science, "map" means a device, table, algorithm, ..., for mapping from one domain to another. (To Boy Scouts, it's a piece of paper with squiggles on it -- but that doesn't fit this case either.) --Carl Ellison ...!harvard!anvil!es!cme (normal mail address) ...!ulowell!cloud9!cme (usenet news reading) (standard disclaimer)
cberg@leadsv.UUCP (Charles R Berg) (01/11/89)
In article <17670002@hpfcdj.HP.COM> myers@hpfcdj.HP.COM (Bob Myers) writes: >For example, would you >call the image produced by a color ink-jet printer a "raster"? Yes.
elf@dgp.toronto.edu (Eugene Fiume) (01/11/89)
In article <3155@cloud9.Stratus.COM> cme@cloud9.Stratus.COM (Carl Ellison) writes: > >Meanwhile, to me as a mathematician, both "bitmap" and "pixmap" are improper >because the word "map" makes no sense. In both math and computer science, >"map" means a device, table, algorithm, ..., for mapping from one domain to >another. This is entirely incorrect. It is not hard to define a bitmap precisely as a function (taking the integers to booleans). Programming methodologists have defined the semantics of arrays (in programming languages) similarly. See Gries' book, for example. Furthermore, it is easy to define boolean rings and algebras over bitmaps. Lastly, as I stated in an earlier posting, bitmaps should be dissociated from images. Yet another "map" (mathematically precise and well defined, I might add) must be specified to associate a bitmap with an image. In this manner, one can define the semantics of things like moving a bitmap across an image, or defining a pixel aspect ratio. Once again, an image can be characterised as yet another map (this time over the reals rather than the integers). -- Eugene Fiume Dynamic Graphics Project University of Toronto elf@dgp.toronto.edu
myers@hpfcdj.HP.COM (Bob Myers) (01/13/89)
Does it strike anyone as ironic that the very group that is spending
so much time worrying about the proper use of the word "bitmap" are
also among that group who think that "input", "output", and "interface"
are verbs?
Bob Myers KC0EW HP Graphics Tech. Div.| Opinions expressed here are not
Ft. Collins, Colorado | those of my employer or any other
{the known universe}!hplabs!hpfcla!myers | sentient life-form on this planet.sean@etnibsd.UUCP (Sean McElroy) (01/14/89)
In article <8901111331.AA25073@explorer.dgp.toronto.edu>, elf@dgp.toronto.edu (Eugene Fiume) writes: > In article <3155@cloud9.Stratus.COM> cme@cloud9.Stratus.COM (Carl Ellison) writes: > >Meanwhile, to me as a mathematician, both "bitmap" and "pixmap" are improper > >because the word "map" makes no sense. In both math and computer science, > >"map" means a device, table, algorithm, ..., for mapping from one domain to > >another. > This is entirely incorrect. It is not hard to define a bitmap > precisely as a function (taking the integers to booleans). Programming Although it is certainly possible to define a functional mapping, Ellison's remarks seem to me to be concerned about the issue of whether the usage of the term bitmap implies some mapping. I think your statement, "This is entirely incorrect", is a little strong when taken within this context. I wonder if we are not trying to be too specific about the semantics of the word bitmap. Although it may have been intended to represent a narrow view at its inception, does that imply that no extensions can be applied to its original meaning? There are words whose current meaning varies widely from their original meaning. What ever happened to the use of context when plying a term? What's wrong with 24-bit color bitmap? Is there anyone out there who cannot understand that phrase? Perhaps we should consider expanding the meaning of this word rather than inventing a number of new words with narrow and specific meanings. For starters we could provide the following facets to the meaning of the term bitmap; one describing some datum (a map), another describing some procedure (a mapping). This would at least provide a compromise between Fiume and Ellison. -- ____,.,_..,__.,_.,__ Sean Philip McElroy __'..__._,_.__.__.__ Eaton Corp., SED _,___`_.'__.__.__.__ 108 Cherry Hill Dr., Beverly, MA 01922 ___`..'_,___.__.__,_ uunet!etnibsd!sean
pokey@well.UUCP (Jef Poskanzer) (01/15/89)
I get a lot of mail about my portable bitmap toolkit. I mean a *lot*
of mail - I'm approaching 1000 messages. A certain percentage of the
mail has always asked how come the package barfs when fed a color
image. Now, I named the package specifically to avoid this class of
questions. I thought *everyone* knew that a bitmap is one bit deep.
Apparently not.
The origin of the term bitmap has to do with there being a mapping, in
the strictest mathematical sense, between bits in memory and pixels on
the screen. The bits don't have to be currently displayed to qualify;
the mapping is still there even when not being used. Remember, bitmap
screens were a new idea back then - memory was $thousands / K, the
PDP-11 (memory-mapped I/O) had just been invented, and most displays
were vector-based.
It is hard for me to see how you can get from this origin to using
"bitmap" for grayscale and color images. I agree that the terminology
is still evolving in this area. Thanks to the X folks, "pixmap" is
becoming common for color images. But I haven't yet heard anyone use
an analogous term for grayscale. So I'll propose one: "graymap".
Now, is everybody happy?
---
Jef
Jef Poskanzer jef@rtsg.ee.lbl.gov ...well!pokey
Actual size.cme@cloud9.Stratus.COM (Carl Ellison) (01/15/89)
In article <975@etnibsd.UUCP>, sean@etnibsd.UUCP (Sean McElroy) writes: > What's wrong with 24-bit color > bitmap? Is there anyone out there who cannot understand that phrase? yup (or we wouldn't be having this discussion :-) --Carl
td@alice.UUCP (Tom Duff) (01/16/89)
In article <17670003@hpfcdj.HP.COM>, myers@hpfcdj.HP.COM (Bob Myers) writes:
) Does it strike anyone as ironic that the very group that is spending
) so much time worrying about the proper use of the word "bitmap" are
) also among that group who think that "input", "output", and "interface"
) are verbs?
Oh come on, everybody knows that any noun can be verbed.rogerh@arizona.edu (Roger Hayes) (01/17/89)
In article <8901111331.AA25073@explorer.dgp.toronto.edu> elf@dgp.toronto.edu (Eugene Fiume) writes: >... >This is entirely incorrect. It is not hard to define a bitmap >precisely as a function (taking the integers to booleans). >... Eugene modestly neglects to mention a reference: "A Mathematical Semantics and Theory of Raster Graphics", Eugene L. Fiume, Ph.D. dissertation, University of Toronto, May 1986. Available as U. of T. Computer Systems Research Institute Tech. Report CSRI-185. Maybe he didn't want to have to send out copies...
utoddl@ecsvax.uncecs.edu (Todd M. Lewis) (01/18/89)
Sorry to do this, folks, but I can't stand it any more.
BitPlane - a group of ordered bits corresponding one to one
to pixels in some display.
BitMap - one or more BitPlanes which, together, determine
the display characteristics of their associated
pixels.
ColorMap - usually a table which maps BitMap pixel values
to Palette colors.
And to confuse things even more, I've been calling the IBM CGA
style of mapping memory to pixels a "pixmap" for years!
--Todd
#define disc "Everyone is entitled to my opinion."