but the mac mini is already the incarnation of airport express AV, complete with storage, built-in quicktime decoders, and DVI/VGA out to the bigscreen...
oh, you want this dog's breakfast built into a smaller brick, even an ipod, with 802.11 b/a/n/g jammed in, too? hardly necessary, a networked ipod only needs bluetooth to scroll through content already mirroring the stuff in the ipod.
now if apple wants to debut an ancillary product to handle widescreen 480p/720p/1080p H.264 content, i suggest an apple-branded 42" 1080p tv screen, just in time for the holidays.
andover (after IPO as symbol ANDN) went to valinux (LNUX) for $800M-$900M (in stock, mostly, when LNUX was $128), including at least $60M in real money.
now slashdot parent LNUX has a market cap of $95M, so you too, can own a piece of your own blogging for less than 3X sales, including the other.org/.com businesses...
i would like to believe this, since although i'm a u.s. citizen (californian / san franciscan) i find these other lands more progressive in many ways, in addition to being quite enjoyable.
however, this semi-retired author underwent his first- ever random drugtest in australia as a plain tourist, at a roadblock in the coonawarra wine region near easter holiday. they don't do that regularly in napa valley, last time i checked. maybe since they are on half-hour time there near melbourne they do funny things...
yet, also, a certain major new zealand hotel has a breathalyzer built-in for convention-goers, where they clearly take the idea of "designated driver" seriously. perhaps i don't stay in enough stateside hotels, but i've never seen that setup here. here's to "when in rome..." and all that rot!
also, a music video version on DVD appears at cdbaby.com (click my URL).
i always found it interesting (in an "uncle tungsten" sort of way) that element #7 (nitrogen) occurs as a mathematical "fixed point" in both the table and the song.
please increment the exposure to the six-sigma stuff, as it is now making critical wall st. rounds via the vault company surveys.
i'm also ex-sun (tenure 7+ years), who survived two massive re-orgs but quit before RIF #3 precisely because of the bilious GE/Sun sigma cruft, which is garbage-in-garbage-out (GIGO) applied to metrics. so sad.
janus team laid-off? shameful. bill joy-style expertise outsourced? yikes. incompetent middle managers still holding on only because their stock options are priced at 3/4/5? horrorshow. NIH driving decision-making? gulp.
now, i'm still rooting for the good engineering work in the face of such madness, though switching my shares to AAPL has served well as household-preserving defence. will niagra save the day? yes, if price-performance becomes so right that amazon/ebay/yahoo/google beg to turn the switch from x86 white-box maintenance. too late to turn the linux tide, though... going private may be just the ticket to ride!
indeed, lossless for archival preservation is the only way, as it fits the basic rule of art restoration technology -- never apply "improvements" which cannot be reversibly undone to take advantage of future science.
ironically then, the lossless format doesn't matter.
however, at least for the instant case of dance video, the likely input (a myriad of digital tape formats) is hopelessly neanderthal -- anything having to do with DV, or MPEG, or even ATSC HDTV already tosses away much color information. (4:1:1, 4:2:0, and 4:2:2 colorspace is embarrassing to preserve "losslessly".) ditto for temporal info, with interlacing being the culprit. even film at 24fps just will not cut it for motion such as dance.
so here's to better camera technology, whether it's 10- or 12-bit 4:4:4 RGB, or something like carver mead's foveon made swift.
confessing, i can relate to those impressionable days-of-yore at the big u.
yet, dating myself severely now by admitting, (as regards to shannon-limit channel coding) i didn't fully appreciate the power of burst-error-correcting fire codes, let alone the new-fangled turbo codes which can get within a hair's breadth of the varshamov-gilbert bound.
but what really impressed me is the dumbkopf overspecification of the MPEG2 source-coding standard, which gifts to us the posterity of hardwired 9 mbits/second in the DVD groove, when we all know now that the capability of MPEG2 is 1-2 mbits/sec for 720x480 stuff.
and now that H.264 does 60fps 1920x1080p at similar quality metrics as plain ole DVD, there's hope for the world. either that or we can pack 8X the amount of simpson's cartoon family drama into the same space.
ultimately, the stone-tablet contingent is right-on about archival preservation. the medium is the message.
forgot to mention, as i took berklekamp's algebraic coding class at cal berkeley too, that even a chen-massey decoder grafted onto cross-interleaved reed-solomon syndromes just can't hack the mechanical trauma induced by any 10-year-old kid who uses these discs.
a good fraction of all dvd's i see played a dozen times or more get those sponge-bob-squarish mpeg-quantizer-give-up-the-gh ost posterization overlays surmountable only by fast-forward over the scrunge.
and now they wanna do 50GB multi-layer hybrid red-blue laser scribbling on dublous unprotected media from indonesian cleanrooms; help!
now try circumferential scratches parallel to laser motion, uncorrectable by servo or ecc, like the ones that happen at random on many library or rent-a-platters.
ya, sony used to brag about how "ecc2" could handle 1/8" drill-bit holes in CD media -- they just never thought that a 1/4" scratch on the label side might overwhelm the ECC unit with dropout storms.
we are not even talking about fingerprints or peanut butter here...
... is non-existent in the same general way as the publicly known apple/ibm & apple/microsoft patent cross-license arrangements.
so, because apple has patented the saddle point (er, "sweet spot") for a song selection control wheel, other companies must choose a different way, usually suboptimal.
as well, the sony ATRAC codec is a non-starter since it's really now just a subset of the dolby/fraunhofer/bell labs/philips/sony (yes, sony is in the Via licensing/AAC patent pool) audio standard. like many companies offering older formats which have better licensing terms for them, they try to milk the obsolete stuff as long as they can.
as well, sony & philips have longstanding differences which eventually get papered over by joint venture. time for apple/sony to do the same thing, now that they play in the same markets and both have the shared experience of microsoft treachery.
'DSC00234.jpg might as well be labeled DON'T_KNOW_DON'T_CARE.jpg'
time to get around to actually naming those 5000 pics you have lying around in iPhoto, because thumbwheel search+display appears to work better for mnemonic things.
although we've seen proof (usenix, 1983) that some of its features (e.g. "-v") are harmful, 'vi' originator WNJ himself is on record (somewhere) preferring 'cat' as a line editor.
sorry for lack of proper citations -- we'll just have to "a9" for them!
so we have the following quadruple of motivations for why microsoft is on a jihad producing for the USPTO many unpublishable, untutored implementations of "ideas" in the form of patents
- defense against other large entities
makes little sense, since they all cross-license
(MSFT/IBM, MSFT/AAPL, MSFT/SUN are publlc)
- defense against small companies or patent
holding outfits
but this can be handled via defensive discloure
means such as ye olde IBM TDB
- offense against large companies
cross-licensing negates this, but MSFT may get
better terms in leveraging by collecting more
patent cards. aim: mutual assured destruction.
- offense against small/medium commercial houses
this makes much sense, to absorb competitive
threats, modulo antitrust.
though defense-via-good-offense is a tried and true principle, note the above quadrant mainly applies to for-profit enterprise. small non-profits don't threaten, but a large free-sprited software movement does. the nexus occurs at places like NOVL/RHAT, which may be about to receive more urgent visits from the microsoft legal department.
when i was a young lad at nasa ames giving away things like free implementations of lempel-ziv compression, boyer-moore search grafted to 'egrep', thompson-style prefix coding for file search, and combinatorial anagram madness of all kinds, i.e.
we were encouraged to donate to COSMIC if development costs exceeded ten kilodollars. natch, on govt. pay many of us worked cheap, so we just put stuff up on 'uucp' as public domain...
excellent points. speaking as an ex-BSD contributor, much unix code is really work-for-hire flying under the radar.
certain open source code authors have "signed off" (i.e. dedicated to public domain, signed the FSF copyright forms, non-exclusively dual-licensed stuff to all-comers, etc.), without caring a whit about whether such artistic works are really works-for-hire...
this is a natural consequence of being a creator who writes open source code "on the side", perhaps unrelated to the business of an employer.
yet, when push comes to shove, employers with $$ can interpret the labor codes more favorably than the poor sap who works on some peripheral R&D.
the phraseology "own time, own equipment" just doesn't fly if the nexus of work & play is computer software.
Re:Is this really going to change the "big picture
on
AT&T Labs' Brain Drain
·
· Score: 1
agreed (as to the phase shift in R&D).
R -> D, largely, with R relegated to universities, despite a few rare exceptions, such as genentech in bio.
yet, this is not a problem, as academic institutions have vigorous patent out-licensing depts. which make sure the good stuff has a shot at commercialization in a win-win scenario with industry.
your bit about apple is perceptive, although one of the first acts of amelio/jobs upon returning was to decimate the corporate library as a frilly distraction. intriguingly, they were right about innovation in GUI as not needing to depend upon empire-building of SIGCHI-types like kay/tognazzini/nielsen in favor of a few good programmers overseen by managers and designers with taste (sjobs/ive).
back to bell labs (in re: the 'gucci lab') -- waxing nostalgically for the calibre of thompson/ritchie/ kernighan/pike is indeed a useful baseline. yet, why does it take the efforts of the second string (stallman/torvalds) to bring systems design borne from great R&D to ultimate relevance?
rather like gould/maiman ahead of schawlow/townes for the laser. it's not so much who is first with the practical invention as who best explains the science early on.
slighting key contributors from credit is bad enough when hidebound committees do such, but even worse when the awardees do. (e.g. rosalind franklin anticipating watson/crick for DNA seems to fit this category, as does newton erasing the memory of hookes.)
one way to help attenuate such slights is overthrow the rule that a max of three can be awarded the same prize. the requirement of being alive at award-time seems useful for some things but spells trouble in helping correct oversights.
then there's just plain incorrect politics, such as that which undermined borges from getting the nobel for literature.
I wish someone would analyze the IBM patent claims.
the '746 patent is miller & wegman's anticipation of the "welch" modification to LZ78, the lempel-ziv algorithm used in 'compress' from bsd/sysV/unixware/etc.
i've long wondered when ibm would get around to asserting this patent, essentially beating sperry/unisys to the u.s. patent office door. filed first, it has priority in any interference. yet, ibm watched as unisys boosted many an enforcement career with the GIF flap.
but now, intriguingly, while many were prematurely celebrating the expiration of the unisys filing earlier this year, '746 was overlooked. so why should this matter if it was filed first -- wouldn't it expire first?
well, no, courtesy a wrinkle in a uruguay round amendment to the berne convention. patents at one time expired 17 years from grant date, but now it's modified to *the greater* of 20 years from filing date *or* grant date + 17. looking up '746 yields the info that it gives up the ghost in year 2006...
likely ibm never asserted because they prefer cross-licensing (was AT&T bell labs, but now with the likes of sun, microsoft, and apple.)
now we see patent offense against the riff-raff like SCO. neatly, since LZW is no longer part of GNU, stallmanites need not worry about collateral damage from IBM, at least for this claim.
Slashdot Mirror
but the mac mini is already the incarnation of ...
airport express AV, complete with storage,
built-in quicktime decoders, and DVI/VGA out to the bigscreen
oh, you want this dog's breakfast built into a smaller brick,
even an ipod, with 802.11 b/a/n/g jammed in, too?
hardly necessary, a networked ipod only needs
bluetooth to scroll through content already mirroring
the stuff in the ipod.
now if apple wants to debut an ancillary product
to handle widescreen 480p/720p/1080p H.264 content,
i suggest an apple-branded 42" 1080p tv screen,
just in time for the holidays.
what, blue-laser disks aren't serialized that way?!
for those born yesterday, slashdot was sold for $10-12M
/
.org/.com businesses ...
or so to andover.net, originally, according to:
http://www.salon.com/tech/log/1999/09/17/slashdot
andover (after IPO as symbol ANDN) went to valinux (LNUX) for $800M-$900M
(in stock, mostly, when LNUX was $128), including at least $60M in real money.
now slashdot parent LNUX has a market cap of $95M, so you
too, can own a piece of your own blogging for less than 3X sales,
including the other
i would like to believe this, since although i'm
...
a u.s. citizen (californian / san franciscan)
i find these other lands more progressive in
many ways, in addition to being quite enjoyable.
however, this semi-retired author underwent his first-
ever random drugtest in australia as a plain tourist,
at a roadblock in the coonawarra wine region near easter holiday.
they don't do that regularly in napa valley,
last time i checked. maybe since they are on
half-hour time there near melbourne they do funny things
yet, also, a certain major new zealand hotel has a
breathalyzer built-in for convention-goers, where they
clearly take the idea of "designated driver" seriously.
perhaps i don't stay in enough stateside hotels, but
i've never seen that setup here. here's to "when in rome..."
and all that rot!
also, a music video version on DVD appears at cdbaby.com
(click my URL).
i always found it interesting (in an "uncle tungsten"
sort of way) that element #7 (nitrogen) occurs as a
mathematical "fixed point" in both the table and the song.
please increment the exposure to the six-sigma stuff,
... going private may be
as it is now making critical wall st. rounds via the
vault company surveys.
i'm also ex-sun (tenure 7+ years), who survived two
massive re-orgs but quit before RIF #3 precisely because
of the bilious GE/Sun sigma cruft, which is
garbage-in-garbage-out (GIGO) applied to metrics.
so sad.
janus team laid-off? shameful. bill joy-style expertise
outsourced? yikes. incompetent middle managers
still holding on only because their stock options are priced
at 3/4/5? horrorshow. NIH driving decision-making? gulp.
now, i'm still rooting for the good engineering work
in the face of such madness, though switching my shares
to AAPL has served well as household-preserving
defence. will niagra save the day? yes, if price-performance
becomes so right that amazon/ebay/yahoo/google beg to
turn the switch from x86 white-box maintenance. too late
to turn the linux tide, though
just the ticket to ride!
indeed, lossless for archival preservation is the
only way, as it fits the basic rule of art restoration
technology -- never apply "improvements" which
cannot be reversibly undone to take advantage
of future science.
ironically then, the lossless format doesn't matter.
however, at least for the instant case of dance video,
the likely input (a myriad of digital tape formats)
is hopelessly neanderthal -- anything having to do with DV,
or MPEG, or even ATSC HDTV already tosses away much
color information. (4:1:1, 4:2:0, and 4:2:2 colorspace is embarrassing
to preserve "losslessly".) ditto for temporal
info, with interlacing being the culprit. even film at
24fps just will not cut it for motion such as dance.
so here's to better camera technology, whether it's
10- or 12-bit 4:4:4 RGB, or something like
carver mead's foveon made swift.
confessing, i can relate to those impressionable
days-of-yore at the big u.
yet, dating myself severely now by admitting,
(as regards to shannon-limit channel coding)
i didn't fully appreciate the power
of burst-error-correcting fire codes,
let alone the new-fangled turbo codes
which can get within a hair's breadth of the
varshamov-gilbert bound.
but what really impressed me is the
dumbkopf overspecification of the MPEG2
source-coding standard, which gifts to us the
posterity of hardwired 9 mbits/second
in the DVD groove, when we all know now
that the capability of
MPEG2 is 1-2 mbits/sec for 720x480 stuff.
and now that H.264 does 60fps 1920x1080p at
similar quality metrics as plain ole DVD,
there's hope for the world. either that
or we can pack 8X the amount of simpson's
cartoon family drama into the same space.
ultimately, the stone-tablet contingent is
right-on about archival preservation.
the medium is the message.
forgot to mention, as i took berklekamp's
h ost posterization
algebraic coding class at cal berkeley too,
that even a chen-massey decoder grafted
onto cross-interleaved reed-solomon syndromes
just can't hack the mechanical trauma induced
by any 10-year-old kid who uses these discs.
a good fraction of all dvd's i see played a dozen
times or more get those sponge-bob-squarish
mpeg-quantizer-give-up-the-g
overlays surmountable only by fast-forward over the scrunge.
and now they wanna do 50GB multi-layer hybrid
red-blue laser scribbling on dublous
unprotected media from indonesian cleanrooms;
help!
no wonder they call it "planned obsolescence".
hub-to-rim? wrong way, jose!
now try circumferential scratches parallel to
laser motion, uncorrectable by servo or ecc,
like the ones that happen at random on many library or rent-a-platters.
ya, sony used to brag about how "ecc2" could handle
1/8" drill-bit holes in CD media -- they just never
thought that a 1/4" scratch on the label side might
overwhelm the ECC unit with dropout storms.
we are not even talking about fingerprints
or peanut butter here...
list not needed with given the existence of the archive.org
wayback machine. try on for size:
http://www.esm.psu.edu/Faculty/Gray/movies.html
... is non-existent in the same general way
as the publicly known apple/ibm & apple/microsoft
patent cross-license arrangements.
so, because apple has patented the saddle point
(er, "sweet spot") for a song selection
control wheel, other companies must choose a
different way, usually suboptimal.
as well, the sony ATRAC codec is a non-starter since it's really
now just a subset of the dolby/fraunhofer/bell labs/philips/sony
(yes, sony is in the Via licensing/AAC patent pool)
audio standard. like many companies offering older
formats which have better licensing terms for them,
they try to milk the obsolete stuff as long as they can.
as well, sony & philips have longstanding differences which
eventually get papered over by joint venture.
time for apple/sony to do the same thing, now
that they play in the same markets and both have
the shared experience of microsoft treachery.
otherwise, pride comes before the fall.
an ibm/apple joint-venture is no more farfetched
than an opposing microsoft/sun tagteam JV.
alignments in subarenas are already at hand,
e.g. the apple/cisco/ibm multimedia alliance...
it's coke vs. pepsi alright, with FSF being
some orthogonal drink, like herb tea.
from the wired magazine october 2004 piece --
'DSC00234.jpg might as well be labeled DON'T_KNOW_DON'T_CARE.jpg'
time to get around to actually naming those
5000 pics you have lying around in iPhoto, because
thumbwheel search+display appears to work
better for mnemonic things.
FYI 'vi' vs. Emacs
although we've seen proof (usenix, 1983) that some of its features (e.g. "-v") are harmful, 'vi' originator WNJ himself
is on record (somewhere) preferring 'cat' as a line editor.
sorry for lack of proper citations -- we'll just have to "a9" for them!
at least for solaris/x86, the souped-up 'lxrun' aka "project janus"
m l
http://wwws.sun.com/software/linux/janus_faq.ht
appears to address this. of course, it's hard to
beat the state-of-the-art userland shipped by the
highest-volume unix producer -- apple.
doesn't project janus for solaris 10 sidestep this,
by letting the solaris kernel run linux cmds + apps?
so we have the following quadruple of motivations
for why microsoft is on a jihad producing for the USPTO
many unpublishable, untutored implementations
of "ideas" in the form of patents
- defense against other large entities
makes little sense, since they all cross-license
(MSFT/IBM, MSFT/AAPL, MSFT/SUN are publlc)
- defense against small companies or patent
holding outfits
but this can be handled via defensive discloure
means such as ye olde IBM TDB
- offense against large companies
cross-licensing negates this, but MSFT may get
better terms in leveraging by collecting more
patent cards. aim: mutual assured destruction.
- offense against small/medium commercial houses
this makes much sense, to absorb competitive
threats, modulo antitrust.
though defense-via-good-offense is a tried and true
principle, note the above quadrant mainly applies
to for-profit enterprise. small non-profits don't
threaten, but a large free-sprited software movement does.
the nexus occurs at places like NOVL/RHAT, which
may be about to receive more urgent visits from the
microsoft legal department.
v.i.z. http://www.openchannelfoundation.org/
d =64438&cid=5977419
when i was a young lad at nasa ames giving away things
like free implementations of lempel-ziv compression,
boyer-moore search grafted to 'egrep',
thompson-style prefix coding for file search,
and combinatorial anagram madness of all kinds, i.e.
http://developers.slashdot.org/
comments.pl?si
we were encouraged to donate to COSMIC if development
costs exceeded ten kilodollars. natch, on govt. pay
many of us worked cheap, so we just put stuff up on 'uucp'
as public domain...
"This may prove useful to you someday
perhaps, in a somewhat bizarre set of circumstances."
--Tom Lehrer
corrected to:
http://www.cdbaby.com/cd/tomlehrer
appears at http://www.cdbaby.com/tomleher
excellent points.
speaking as an ex-BSD contributor, much unix code
is really work-for-hire flying under the radar.
certain open source code authors have "signed off"
(i.e. dedicated to public domain, signed the FSF
copyright forms, non-exclusively dual-licensed
stuff to all-comers, etc.), without caring a whit
about whether such artistic works are really
works-for-hire...
this is a natural consequence of being a creator
who writes open source code "on the side",
perhaps unrelated to the business of an employer.
yet, when push comes to shove, employers
with $$ can interpret the labor codes more favorably
than the poor sap who works on some peripheral R&D.
the phraseology "own time, own equipment"
just doesn't fly if the nexus of work & play
is computer software.
agreed (as to the phase shift in R&D).
R -> D, largely, with R relegated to universities, despite
a few rare exceptions, such as genentech in bio.
yet, this is not a problem, as academic institutions have
vigorous patent out-licensing depts. which make sure
the good stuff has a shot at commercialization in a
win-win scenario with industry.
your bit about apple is perceptive, although one of the
first acts of amelio/jobs upon returning was to decimate
the corporate library as a frilly distraction. intriguingly,
they were right about innovation in GUI as not needing
to depend upon empire-building of SIGCHI-types like
kay/tognazzini/nielsen in favor of a few good programmers
overseen by managers and designers with taste (sjobs/ive).
back to bell labs (in re: the 'gucci lab') --
waxing nostalgically for the calibre of thompson/ritchie/
kernighan/pike is indeed a useful baseline.
yet, why does it take the efforts of the second string
(stallman/torvalds) to bring systems design borne
from great R&D to ultimate relevance?
rather like gould/maiman ahead of schawlow/townes
for the laser. it's not so much who is first
with the practical invention as who best explains
the science early on.
slighting key contributors from credit is bad
enough when hidebound committees do such,
but even worse when the awardees do.
(e.g. rosalind franklin anticipating
watson/crick for DNA seems to fit this
category, as does newton erasing the memory
of hookes.)
one way to help attenuate such slights is
overthrow the rule that a max of three
can be awarded the same prize. the requirement
of being alive at award-time seems useful
for some things but spells trouble in helping correct oversights.
then there's just plain incorrect politics,
such as that which undermined borges from
getting the nobel for literature.
the parent pleas:
I wish someone would analyze the IBM patent claims.
the '746 patent is miller & wegman's anticipation
of the "welch" modification to LZ78, the lempel-ziv
algorithm used in 'compress' from bsd/sysV/unixware/etc.
i've long wondered when ibm would get around to
asserting this patent, essentially beating sperry/unisys
to the u.s. patent office door. filed first, it has priority
in any interference. yet, ibm watched as unisys boosted
many an enforcement career with the GIF flap.
but now, intriguingly, while many were prematurely
celebrating the expiration of the unisys filing earlier
this year, '746 was overlooked. so why should this
matter if it was filed first -- wouldn't it expire first?
well, no, courtesy a wrinkle in a uruguay round
amendment to the berne convention. patents at
one time expired 17 years from grant date, but now it's
modified to *the greater* of 20 years from filing
date *or* grant date + 17. looking up '746 yields
the info that it gives up the ghost in year 2006...
likely ibm never asserted because they prefer
cross-licensing (was AT&T bell labs, but now
with the likes of sun, microsoft, and apple.)
now we see patent offense against the riff-raff
like SCO. neatly, since LZW is no longer part of GNU,
stallmanites need not worry about collateral
damage from IBM, at least for this claim.