roy@phri.UUCP (Roy Smith) (12/04/86)
In article <125@ogesml0.UUCP> ciaran@ogesml0.UUCP (Ciaran O'Donnell) writes: > 6450 dhrystones/second on the MT2025 processing module of the SM90 (68020 > CPU, 25 Mhz clock, zero wait states in r/w, 1 megabyte of on-board DRAM) > > 6450 dhrystones/seconde, sur la SM90 module de traitement MT2025 (CPU 68020, > vitesse 25 Mhz, zero wait states, 1 megaoctet de RAM dynamique sur la carte) I don't know French, but it seems pretty clear that the phrases "1 megabyte" and "1 megaoctet" are equivalent. Is "octet" the proper French translation of of the English "byte"? Assuming that it is, doesn't that presuppose that a byte is 8 bits? We have enough trouble here teaching people that a byte doesn't have to be 8 bits -- if you call it an "octet", it must be even more confusing: "the VAX has 8-bit octets, but the PDP-10 has 9-bit octets." -- Roy Smith, {allegra,cmcl2,philabs}!phri!roy System Administrator, Public Health Research Institute 455 First Avenue, New York, NY 10016 "you can't spell deoxyribonucleic without unix!"
thomson@uthub.toronto.edu (Brian Thomson) (12/04/86)
According to the technical terminology book sitting on my officemate's
desk, titled "Terminologie de l'informatique" and published by the Office
de la langue francaise du Quebec, the correct translation for 'a byte'
is 'un multiplet', but 'a megabyte' is indeed 'un megaoctet'.
It seems to me that English is the sloppy one; I have never heard 'megabyte'
used to mean anything other than mega-octet, though 'byte' can certainly
have different lengths.
--
Brian Thomson, CSRI Univ. of Toronto
{linus,ihnp4,uw-beaver,floyd,utzoo}!utcsri!uthub!thomsonmarc@zeus.cs.ucla.edu (Marc Tremblay) (12/04/86)
In article <2523@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: >Is "octet" the proper French translation of of the English "byte"? In French "octet" means a 8-bit byte. Actually it comes from the latin "word" octo which means that it is related to eight. (I took latin a long time ago!) An "octet" of 9 bits should not be correct in this sense, but remember that the three musketeers were actually four...
lamy@ai.toronto.edu (12/04/86)
"byte", when used a a measure of memory capacity, usually means 8 useable bits, and gets translated as "octet". Memory should be measured in bits, or maybe in characters (maybe this is where the confusion comes from, since often 1 char <= 1 byte). Even the notion of "byte" makes little sense since it no longer corresponds to the width of the data paths, to processor register size. It may be useful when talking of the different lenghts of instructions, but I need to be convinced... "multiplet" is more general, and applies to the 9bit PDP-10 and the 6bit CDC "bytes". Jean-Francois Lamy AI Group, Dept of Computer Science CSNet: lamy@ai.toronto.edu University of Toronto EAN: lamy@ai.toronto.cdn Toronto, ON, Canada M5S 1A4 UUCP: lamy@utai.uucp
brunner@sri-spam.istc.sri.com (Thomas Eric Brunner) (12/04/86)
When I worked in Europe last year bytes were octets. The point made re: 8bits is well taken - but so what? -- Cheers! /teb brunner@spam.sri.com
vanam@pttesac.UUCP (Marnix van Ammers) (12/05/86)
In article <2523@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: >In article <125@ogesml0.UUCP> ciaran@ogesml0.UUCP (Ciaran O'Donnell) writes: >> vitesse 25 Mhz, zero wait states, 1 megaoctet de RAM dynamique sur la carte) >We have enough trouble here teaching >people that a byte doesn't have to be 8 bits -- if you call it an "octet", A byte doesn't have to be 8 bits? I thought a byte was *always* 8 bits. I thought a "word", or in C a "char" could be any number of bits, but that the "byte" was 8 bits. So what about a "nibble"? Marnix
wilkes@beatnix..UUCP (John Wilkes) (12/05/86)
In article <346@pttesac.UUCP> vanam@pttesac.UUCP (-Root Admin-) writes: [stuff deleted] > >A byte doesn't have to be 8 bits? I thought a byte was *always* >8 bits. [stuff deleted] >Marnix On the venerable PDP-10 (R.I.P, sigh) a ``byte'' could be anything from one bit to 36 bits (the word size). The byte size was defined in the ``byte pointer''. -- John (I don't care what they say, 36 bits is here to stay) Wilkes -- John Wilkes -- UUCP: {ihnp4|decwrl|pyramid}!sun!elxsi!beatnix!wilkes USPS: ELXSI, 2334 Lundy Pl., San Jose CA 95131 # My Employer appologizes for the fact that I have access to the network. # Furthermore, my Employer has absolutely no responsibility for the above # random ramblings, which are clearly the product of a deranged mind.
bzs@bu-cs.BU.EDU (Barry Shein) (12/05/86)
Byte size is a variable quantity. Octet is used extensively in the ARPAnet protocol definitions when they mean precisely eight bits. It is a fine term and anyone who is teaching people that a byte is an absolute quantity meaning 8 bits is doing those folks a disservice. That does not mean, of course, that when speaking in a context one cannot first explain that their use of the word byte will always means an octet. I use this when I teach. It's not not an octet either. -Barry Shein, Boston University
lodman@ncr-sd.UUCP (Mike Lodman) (12/05/86)
In article <348@uthub.toronto.edu> thomson@uthub.toronto.edu writes: > >It seems to me that English is the sloppy one; I have never heard 'megabyte' >used to mean anything other than mega-octet, though 'byte' can certainly >have different lengths. Although at one time this might have been a correct statement, its my quess that a reference today to anything but an 8 bit byte will cause blank stares and massive confusion. For all practical purposes the definition of a byte is eight bits. Michael Lodman Advanced Development NCR Corporation E&M San Diego
jkg@gitpyr.gatech.EDU (Jim Greenlee) (12/05/86)
In article <3284@curly.ucla-cs.UCLA.EDU> marc@zeus (Marc Tremblay) writes: >remember that the three musketeers were actually four... This has very little to do with the current discussion, but the THREE musketeers are Porthos, Athos, and Aramis. D'Artagnan was more like an apprentice. Jim Greenlee -- The Shadow...!{akgua,allegra,amd,hplabs,ihnp4,seismo,ut-ngp}!gatech!gitpyr!jkg Jryy, abj lbh'ir tbar naq qbar vg! Whfg unq gb xrrc svqqyvat jvgu vg hagvy lbh oebxr vg, qvqa'g lbh?!
byer@esunix.UUCP (12/06/86)
>In article <125@ogesml0.UUCP> ciaran@ogesml0.UUCP (Ciaran O'Donnell) writes: >> CPU, 25 Mhz clock, zero wait states in r/w, 1 megabyte of on-board DRAM) >> vitesse 25 Mhz, zero wait states, 1 megaoctet de RAM dynamique sur la carte) > Roy Smith, {allegra,cmcl2,philabs}!phri!roy writes: >megabyte" and "1 megaoctet" are equivalent. Is "octet" the proper French >translation of of the English "byte"? Assuming that it is, doesn't that >presuppose that a byte is 8 bits? We have enough trouble here teaching >people that a byte doesn't have to be 8 bits -- if you call it an "octet", >it must be even more confusing: "the VAX has 8-bit octets, but the PDP-10 >has 9-bit octets." In my (admittedly) brief experiences with Cyber 170's and French people, octet seems to be a generic term that can mean anywhere between 6 and 9 bits. The terms seems interchangable when refering to Vax bytes or Cyber bytes (6 bits). John Byer {decvax!decwrl!esunix!byer} Evans & Sutherland Computer Corp. (801)582-5847 Box 8700 Salt Lake City, Utah 84108 Anarchy Rules!
roy@phri.UUCP (Roy Smith) (12/06/86)
[Note: followups redirected to comp.std.internat only]
[I tried to send this earlier, but an overflowing /usr/tmp caused my file
to get eaten -- this is from memory and not as good as the original.]
This all started when I noticed that <<octet>> is the French
translation for the English "byte", and pointed out that this is sure to
cause confusion when you teach people that a byte does not have to be 8
bits. (This, by-the-way, was not intended as a slur against the French
language, simply an observation; English has many, greater, stupidities.)
Random people have objected to my contention that "byte != 8 bits".
Several people have pointed that <<multiplet>> is a {better, different,
alternative, what-have-you} word for <<octet>>. I like <<multiplet>>
better for the generalized concept of "smallest natural grouping of bits on
a particular machine".
In article <1257@ncr-sd.UUCP> lodman@ncr-sd.UUCP (Michael Lodman) writes:
> Although at one time [a byte is 8 bits] might have been a correct
> statement, its my quess that a reference today to anything but an 8 bit
> byte will cause blank stares and massive confusion. For all practical
> purposes the definition of a byte is eight bits.
If you speak Latin to most people in the world today, you will get
blank stares and massive confusion. Does that mean Latin is not a real
language? If I told my mother (an intelligent and educated women who
doesn't happen to grok complex algebra) that "sqrt (-1) = exp (j*pi/2)"
that would also cause blank stares and massive confusion. Does that mean
it isn't a true statement? Try telling all those people who work with
DEC-20's that a byte is 8 bits, and see how many blank stares you get then.
Being common usage doesn't make something right. I commonly talk
about making a Xerox copy, when in reality we only have Kodak and Cannon
photocopy machines at work. When I refer to the RS-232 port on a Mac-plus,
but I know full well it's not really RS-232 because (among other things) it
has the wrong connector. That people are sloppy about the phrase "Xerox
copy" doesn't cause much trouble (to me anyway; Xerox folks probably feel
differently), but people being sloppy about what they mean by "RS-232"
causes me more grief than you want to know about.
It is important to make a distinction between common speech and
speaking technicly. "Byte" is a technical term, and as such, it has a
specific meaning. I'm not going to stick my neck out and say exactly what
the definition is, but it sure doesn't include having to be 8 bits. If
people continue to use technical terms for whatever they feel like meaning
that particular day, pretty soon we'll have:
If it signals with +/- 10 volts, it's RS-232.
If it has 2 N connectors and a DB-15, it's Ethernet.
If it's 1200 baud full-duplex, it's Bell-212A.
If it's got "From:" and "To:" lines in the header, it's RFC-822.
If sending it "ESC [ <something>" makes it jump, it's X3.64.
If it prints out a "% " and runs "ls -l", it's Unix.
Unfortunately, there are all too many people in the world who would
say that some or all of the above are true. Many of those people are in
the business of building and selling computer systems. More the pity.
--
Roy Smith, {allegra,cmcl2,philabs}!phri!roy
System Administrator, Public Health Research Institute
455 First Avenue, New York, NY 10016
"you can't spell deoxyribonucleic without unix!"vanam@pttesac.UUCP (Marnix van Ammers) (12/07/86)
According to Websters:
byte: a group of adjacent binary digits often shorter than a
word that a computer processes as a unit.
So it doesn't have to be 8 bits.
--
Marnix A. van\ Ammers Work: (415) 545-8334
Home: (707) 644-9781 CEO: MAVANAMMERS:UNIX
{ihnp4|ptsfa}!pttesac!vanam CIS: 70027,70kimcm@olamb.UUCP (Kim Chr. Madsen) (12/07/86)
> I don't know French, but it seems pretty clear that the phrases "1 > megabyte" and "1 megaoctet" are equivalent. Is "octet" the proper French > translation of of the English "byte"? Assuming that it is, doesn't that > presuppose that a byte is 8 bits? We have enough trouble here teaching > people that a byte doesn't have to be 8 bits -- if you call it an "octet", > it must be even more confusing: "the VAX has 8-bit octets, but the PDP-10 > has 9-bit octets." The word "octet" is clearly defined as a bitgroup of 8 bits, which is why it is called an octet. This is done to avoid the confusion that the word "byte" brings along. You can only answer the question "How long is a byte ?" by saying: Well, look in the manual for your machine architecture, but normally it's about 8 bits long. Talking about n-bit octets are ridiculous. Talking about n-bit bytes makes sense! Kim Chr. Madsen
ihm@minnie.UUCP (Ian Merritt) (12/08/86)
>In article <2523@phri.UUCP> roy@phri.UUCP (Roy Smith) writes: >>In article <125@ogesml0.UUCP> ciaran@ogesml0.UUCP (Ciaran O'Donnell) writes: >>> vitesse 25 Mhz, zero wait states, 1 megaoctet de RAM dynamique sur la carte) >>We have enough trouble here teaching >>people that a byte doesn't have to be 8 bits -- if you call it an "octet", > >A byte doesn't have to be 8 bits? I thought a byte was *always* >8 bits. I thought a "word", or in C a "char" could be any number of >bits, but that the "byte" was 8 bits. Ah, here we have the crux of the problem; the proliferation of terminology with partial loss of meaning. A byte actually refers not to a specific number of bits, but actually to an arbitrary number of bits that is somewhat context dependent. The advent of the term "octet" (at least in its utilization in english) does refer specifically to 8 bits. For whatever it's worth, the PDP-10/DEC-20 systems do not have a fixed byte size. Instead, there is a special goodie called a byte pointer which consists of an address, an offset, and a size. Using these, any size bytes (from 1 to 36 bits) can be manipulated with the instructions provided. In the microprocessor world, and in most other newer machines, there have not been bytes of other than 8 bits, so the meaning of this term has become constrained in the minds of those not familiar with such machines as the PDP-10. Something similar has happend with the term "Baud" which actually refers to the number of symbols per second, regardless of how many bits are encoded per symbol. As such, the so-called 1200 baud modem is really 1200 bps, and (I think they are qpsk) probably 300 baud. > >So what about a "nibble"? Nybble (spelled with a "y", as in byte) usually refers to a half-byte, which in most recent cases (i.e. since the advent of the term nybble) happens to be 4 bits. Cheerz-- <>IHM<> -- uucp: ihnp4!nrcvax!ihm
chuck@amdahl.UUCP (Charles Simmons) (12/09/86)
In article <348@uthub.toronto.edu> thomson@uthub.toronto.edu writes: > It seems to me that English is the sloppy one; I have never heard 'megabyte' > used to mean anything other than mega-octet, though 'byte' can certainly > have different lengths. > -- > Brian Thomson, CSRI Univ. of Toronto > {linus,ihnp4,uw-beaver,floyd,utzoo}!utcsri!uthub!thomson I take it you've never worked on a Honeywell Computer. Some honeywell's use 9-bit bytes, and the amount of main memory (core) is measured in megabytes. chuck@amdahl
pedz@bobkat.UUCP (12/10/86)
Who gives a fuck? Why put this in this group? This has dragged on
for weeks now.
--
Perry Smith
{convex!ctvax,{ti-csl,infotel}!pollux}!bobkat!pedzphilip@axis.UUCP (Philip Peake) (12/12/86)
In article <269@bobkat.UUCP> pedz@bobkat.UUCP (Pedz Thing) writes: > >Who gives a fuck? What a nice person !
berger@imag.UUCP (Gilles BERGER SABBATEL) (12/12/86)
In article <261@esunix.UUCP> byer@esunix.UUCP writes: > >In my (admittedly) brief experiences with Cyber 170's and French people, >octet seems to be a generic term that can mean anywhere between 6 and >9 bits. The terms seems interchangable when refering to Vax bytes or >Cyber bytes (6 bits). > Definitely right - even if it can be a bit confusing. I think that flames about that should go to "Academie Francaise", and further discussion should go to the appropriate talk group. -- Gilles BERGER SABBATEL IMAG-TIM3/INPG, 46 Avenue Felix Viallet, F-38031 GRENOBLE CEDEX - FRANCE Tel: 76 47 98 55 Ext: 606 UUCP: ...!seismo!mcvax!inria!archi!berger or: berger@archi
wilkes@beatnix..UUCP (John Wilkes) (12/13/86)
In article <269@bobkat.UUCP> pedz@bobkat.UUCP (Pedz Thing) writes: > >Who gives a fuck? Why put this in this group? This has dragged on >for weeks now. > >-- >Perry Smith >{convex!ctvax,{ti-csl,infotel}!pollux}!bobkat!pedz I do, for one. If you do not share our interest, may I suggest the use of either the ``K'' key, or better yet, the ``u'' key. -- John Wilkes -- UUCP: {ihnp4|decwrl|pyramid}!sun!elxsi!beatnix!wilkes USPS: ELXSI, 2334 Lundy Pl., San Jose CA 95131 # My Employer appologizes for the fact that I have access to the network. # Furthermore, my Employer has absolutely no responsibility for the above # random ramblings, which are clearly the product of a deranged mind.
kimcm@olamb.UUCP (Kim Chr. Madsen) (12/15/86)
In article <778@axis.UUCP>, philip@axis.UUCP (Philip Peake) writes: > In article <269@bobkat.UUCP> pedz@bobkat.UUCP (Pedz Thing) writes: > > > >Who gives a fuck? > > What a nice person ! Well, well, well, Why not take it as an invitation (^: Merry X-mas Kim Chr. Madsen
scott@utcs.uucp (12/15/86)
In article <778@axis.UUCP> philip@axis.UUCP (Philip Peake) writes: > >In article <269@bobkat.UUCP> pedz@bobkat.UUCP (Pedz Thing) writes: >> >>Who gives a fuck? > >What a nice person ! not so nice, but I'd have to agree with him.... -- "I feel fine..." ...{utzoo, decvax, ihnp4, cbosgd, utcsri, mnetor}!utcs!scott scott%utcs.toronto.edu@csnet-relay.arpa scott@utoronto.bitnet scott@utcs.utoronto.bitnet Disclaimer: The above is not actually the opinion of anyone at all but especially not the administration or staff of this institution.
ihm@minnie.UUCP (Ian Merritt) (12/15/86)
> >Who gives a fuck? Why put this in this group? This has dragged on >for weeks now. > >-- >Perry Smith >{convex!ctvax,{ti-csl,infotel}!pollux}!bobkat!pedz Your choice of (ref: STIV) "colorful metaphors" leaves something to be desired. Perhaps the point is valid, though: how is this conversation still germain to this list. <>IHM<> -- uucp: ihnp4!nrcvax!ihm