Docsity
Docsity

Prepara tus exámenes
Prepara tus exámenes

Prepara tus exámenes y mejora tus resultados gracias a la gran cantidad de recursos disponibles en Docsity


Consigue puntos base para descargar
Consigue puntos base para descargar

Gana puntos ayudando a otros estudiantes o consíguelos activando un Plan Premium


Orientación Universidad
Orientación Universidad


Data Migration and Standardization in Computing: A Call for Open Data Interchange Formats, Apuntes de Informática

The challenges of data migration and standardization in the computing industry, focusing on the need for open data interchange formats to prevent data loss and theft by software manufacturers. The author argues that the current state of software, with its proprietary and unstandardized formats, is akin to horse carts being unable to keep up with the freeway, and that this issue is more important than debates over pc versus mac versus amiga. The document also touches on the importance of standardization, the limitations of niche machines, and the role of reverse engineering in copyright law.

Tipo: Apuntes

2023/2024

Subido el 16/01/2024

george-luna
george-luna 🇧🇴

29 documentos

1 / 1

Toggle sidebar

Esta página no es visible en la vista previa

¡No te pierdas las partes importantes!

bg1
FSTV
SSTV
FAX RTTY PACKET
AMTOR OSCAR FM REPEATERS
MICROWAVE AND LOTS MORE!
pat
ib le, e
xcept
by
resorting
to
lowest
-common-denominator
user
interfaces
and st icking by a
640K
memory model. Even then, it's never
really certain that the program that
runs on my PC will run on the PC ne
xt
do
or-and
getting it to print properly
on the system next
door
is well nigh
out of the question.
So, what will become of those 60
million PC 's, plus the Macs , Am igas,
Ataris, and so on? It's sad, but hu-
manity has been down this road be-
fore, and we find that horse carts,
though they were once as abundant
proportionally, simply aren't permit-
ted on t he freeway . Those computers
aren't going to make it either, though
some, notab ly the
DOS
PC, may we ll
hang on for a long wh ile as windows in
some POSIX-OS /2-Windows-DOS-
Sparc
-Mac
-PC-Amiga
machine of
the future.
(After
all, if it were worth
the trouble, it wo uldn't be all that hard
to simulate an Apple or Commodore
64 of seven years ago on a
486
PC,
high-end Mac, or high-end
Amiga-
faster than real timet)
And that raises what should be the
hottest
issue of all,
something
far
more important than ranting and rav-
in g
over
doomed
PC's
versus
doomed
Macs
versus
doomed
Amigas. The issue is our own data,
which we must somehow get out of
that DOS window and onto that ma-
chine of the fu
ture-
to say noth ing of
moving it from one application to an-
other
or
to a
supercomputer
. We
have allowed much of our data, en-
tered laboriously and at tremendous
e
xpense
, to
become
embodied
in
files that are formatted in obtuse, se-
cret , and unstandardized ways .
It is high time that the software
manufacturers who practice this are
forced to change . They are i n effect
thieves who have sto len from us what
is rightfu lly our property. It is time to
insist that all software that we buy
have the capabi lity to disgorge our
data in f ull, including such thing s as
the coordinates of graphical objects,
in an
open,
pub
li
sh
ed-d
ata-int
er
-
change format. As a sy mbo lic dis-
couragement to th
eft
and extortion
by so
ftw
are manufa
ctur
ers , and to al-
low third partie s to assist in retrieving
the stolen property without fear of
litigation, it is time to amend the
copyright law to explicitly perm it re-
verse engineering in any
progr
am
that s to res o utput but won 't prov ide it
in an o pen format, irrespective of any
COMPUTER CLASHES CONTINUED
I really do think that Mr. Cheng 's
comments in the
Apr
il " Letters" co l-
umn
went
a bit far
-and
so did
Mr
.
Ho ltz man 's rep ly . T he
ent
ire ex-
change brings to mind the o ld tale
about the b lind man a nd the e lephant ;
it really is tough to discern the big
picture through all the fog . As one
who has programmed extensively for
80X86
PC's, Am igas , and other com-
puters, I have more comments than I
can fit in a reasonably sized letter. A
few wi ll have to do .
First , on standardization: As Mr.
Holtzman pointed out, the desirable
part about standardization is that it
facilitates communication. There are
two flip sides. One is that many
soft
-
ware (as
we
ll as hardwa re) supp liers
don't want communication . Supp liers
of programs
for
CAD
and music
scor
-
ing are
espec
ially
notorious
-they
want
you locked in tight. The second
is that standards are often ages out
of date
-w
itness the
NTSC
televi-
sion standard
-to
the po int of st ifling
innovation.
The niche mach ines are a paradox.
They owe their existence to the fact
that in one important sense they are
more standardized than PC's . No two
PC's are alike, even in suc h funda -
menta ls as memory organization, vid-
eo
graph
ics
layout,
and i nt err up t
organization. As we all know,
that
causes no end of prob lems . Within
the CP U itse lf, the instr uctio n set is
an unstandardized chaos, havinq on
the
order
of a
thousand
ind ivid ual
qu
irks
. Even
the
largest.
best
-cap -
italized organ izations have been able
to cope with that, so that compiled
code tends to be a wretchedly sub-
optimaljoke. The lesson s of the old
mainframes have yet to be learned.
The upshot is that things that aren't
done often enough to support multi-
billion-dollar software companies are
often best done on niche machines,
especially if speed, large memory, or
a GUI is desired, or interrupts are
needed. Sure , a Lotus or a Microsoft
can hand le the problems caused by
dozens of graphics drivers , dozens of
extended memory drivers, more than
500
printer driv ers , dozens of s lightly
different serial and paralle l ports, hun-
dreds of quirky BIOS variants, and
hundreds
or
maybe thou sands of
hard-disk variants . The person who
writes aspecialized program for filter
analys is and sells a few hundred co p-
ies probably
can't
make it multi-com-
....
z
e
to-
-
:IE
Q
ce
OUR
24TH
YEAR
S
INCE
1967!
MCN ISA (5% added)
----
The SPEC-COM Journal
P.O. Box 1002,
Dubuque
,IA52004
(319) 557-879 1
lot II .... . ..
10
.
l
~ul
",
.,,,1,
1
~11
::=---:....-.~
_.:;:::".,
._
..
SEND
$25
CHECK
OR
MONEY
ORDER
The FCC recen tly passe d Docket 90-
55 which for th e first time allows a new
code less entry ham rad io lic ense of
tec hnician grade. Privileges 30 MHz
and above - All modes ! (See R.E. ar-
ticle in Apr il 1991 issue).
Your
Ticket
To
S'J(j(jl~SS
Over 28,000 tec hnicians have gained
admit
-
tance
wor
ldwide as certified professionals.
Let yo ur
ticket
start
open
ing do ors
for
you .
ISCET offers
Jou
rneyman certifi
cation
in
Consumer Elect ron ics, Industrial, Medical,
Communications,Radar,
Comp
uter
and
Video. For more i
nformation,
contac
t the
I
nternational
Society
of Certified Electro-
nics
Techn
icians, 27 08 West Berry
St
reet.
Fort
Worth, TX 761
09;
(817)
92 1-9101 .
Name _
Address _
City _
State Zip _
Send material
abo
ut ISCET and
--
becom ing cert ified.
Send
one
"Study
Guide for the
--
Associate Level C ET
Test
." En-
Lclose d is
$10
(inc . postage).I
-----------_-.1
r
~;;-
;O~~D-;
,
AMATEUR
RAD
IO
LICENSE
Get
all the
no-code
license
deta
ils,
study &test ing information plus a one-
year subscrip tion to one of ham radio's
longest running specia lty mode publi -
cations that will teach you all about the
new modes you w ill be able to operate!
(J)
a
Z
o
a::
I-
a
ui
...J
W
o
is _
«L
.....
'
a::
IIIiiiIiIII
-
10

Vista previa parcial del texto

¡Descarga Data Migration and Standardization in Computing: A Call for Open Data Interchange Formats y más Apuntes en PDF de Informática solo en Docsity!

FSTV SSTV FAX RTTY PACKET AMTOR OSCAR FM REPEATERS MICROWAVE AND LOTS MOR E!

pat ib le, e xcept by resorting to lowest -common-denominator user interfaces and st icking by a 640K memory model. Even then, it's never really certain that the program that runs on my PC w ill run on the PC next do or-and getting it to print properly on the system next door is well nigh out of the question. So , what wil l become of those 60 million PC 's, plus the Macs , Am igas, Ataris, and so on? It's sad, but hu- manity has been down this road be- fore, and we find that horse carts, though they were once as abundant proportionally, simply aren't permit- ted on t he freeway. Those computers aren't go ing to make it either, though some, notab ly the DOS PC, may we ll hang on for a long wh ile as windows in some POSIX -OS /2-Windows-DOS- Sparc -Mac -PC-Amiga mach ine of the future. (After all, if it were worth the trouble, it wo uldn't be all that hard to simulate an Apple or Commodore 64 of seven years ago on a 486 PC, high-end Mac, or high-end Amiga- faster than real timet) And that raises what shou ld be the hottest iss ue of all, something far more important than ranting and rav- ing over doomed PC's versus doomed Macs versus doomed Amigas. The issue is our own data, wh ich we must somehow get out of that DO S w indow and onto that ma- chine of the fu ture- to say noth ing of moving it from one app lication to an- other or to a supercomputer. We have allowed much of our data, en- tered laboriously and at tremendous e xpense , to become embodied in files that are formatted in obtuse, se- cret , and unstandardized ways. It is high time that the software manufacturers who practice this are forced to change. They are in effect thieves who have sto len from us what is rightfu lly our property. It is time to insist that all software that we buy have the capabi lity to disgorge our data in full, including such thing s as the coordinates of graphical objects, in an open, pub li sh ed -d ata-int er- change format. As a symbolic dis- couragement to th eft and extortion by so ftw are manufacturers, and to al- low third partie s to assist in retrieving the stolen prop erty without fear of lit igat ion, it is t ime to ame nd the copyright law to explicitly perm it re- vers e engineering in any progr am that stores outp ut but won 't provide it in an open format, irrespective of any

COMPUTER CLASHES CONTINUED

I really do think that Mr. Cheng 's comments in the Apr il " Letters" co l- umn went a bit far -and so did Mr. Holtzman 's rep ly. The ent ire ex- change brings to mind the old tale about the blind man and the elephant ; it really is tough to discern the big picture through all the fog. As one who has programmed extensively for 80X86 PC's, Am igas , and other com - puters , I have more comments than I can fit in a reasonably sized letter. A few wi ll have to do. First , on standardization: As Mr. Holtzman pointed out, the desirable part about standardization is that it facilitates communication. There are two flip sides. One is that many soft - ware (as well as hardwa re) supp liers don 't want communication. Supp liers of programs for CAD and music scor- ing are espec ially notorious -they want you locked in tight. The second is that standards are often ages out of date -w itness the NTSC te levi- sion standard -to the po int of st ifling innovation. The niche mach ines are a paradox. They owe their existence to the fact t hat in one important sense they are more standardized than PC's. No two PC's are alike, even in suc h funda - menta ls as memory organization, vid - eo graph ics layout, and int e rr upt organ ization. As we all know, that causes no end of prob lems. Within the CP U itse lf, the inst ruction set is an unstandardized chaos, havinq on t he order of a thousand ind ivid ual qu irks. Eve n the largest. best -cap - italized organ izations have been able to cope with t hat , so that compiled code tends to be a wretchedly sub- optima l joke. The lessons of the old mainframes have yet to be learned. The upshot is that th ings that aren't done often enough to support multi- billion-dollar software companies are often best done on niche machines, especially if speed , large memory, or a GUI is desired, or interrupts are needed. Sure , a Lotus or a Microsoft can handle the prob lems caused by dozens of graphics drivers , dozens of extended memory drivers, more than 500 printer drivers , dozens of slightly different serial and parallel ports , hun- dreds of quirky BIOS variant s, and hundreds or maybe t hou sands of hard-disk variants. The person who writes a specia lized program for filter analys is and sells a few hundred cop- ies probably can't make it multi-com-

z

e

to-

- :IE

Q

ce

OUR

24TH YEAR

SINCE

1967!

MCN ISA (5% added)

The SPEC-COM Journal P.O. Box 1002, Dubuque , IA 52004 (319) 557-879 1

lot II ..... .. 10. l ~ul ", .,,,1, 1 ~ ::=---:....-.~ .:;:::"., ...

SEND

$

CHECK

OR

MONEY

ORDER

Th e FCC recen tly passed Docket 90- 55 which for the first time allows a new code less entry ham rad io license of tec hnician grade. Privi leges 30 MHz and above - All modes! (See R.E. ar- ticle in April 1991 issue).

Your Ticket To

S'J(j(jl~SS

Over 28,000 tec hnicians have gained admit - tance wor ldwide as certified professionals. Let yo ur ticket start open ing d oors for you. ISCET offers Jou rneyman certifi cation in Consumer Elect ron ics, Industrial, Medical, Comm u nications, Radar, Comp uter and Video. For more i nformation, contac t t he I nternational Society of Certified Electro- nics Techn icians, 27 08 West Berry St reet. Fort Worth, TX 761 09; (817) 92 1-. Name _ Address _ City _ State Zip _ Send material abo ut ISCET and --becom ing cert ified. Send one "Study Gu ide for the --Associate Level CET Test ." En- L

closed is $10 (inc. postage). I

-----------_-.

r ~;;- ;O~~D-; ,

AMATEUR RAD IO LICENSE

Get al l t he no-code license deta ils, study & testing information plus a one- year subscrip tion to one of ham radio's longest running specia lty mode publi - cations that will teach you all about the new modes you will be able to operate!

(J) a

Z

o

I-a::

a ui ...J W

ois _

«L a:: IIIiiiIiIII ..... '