RFD: comp.compression.random
Metatron
REQUEST FOR DISCUSSION (RFD)
unmoderated group comp.compression.random
This is a formal Request for Discussion (RFD) for the creation of the
unmoderated newsgroup comp.compression.random.
NEWSGROUPS LINE:
comp.compression.random Random data compression algorithms and
derivates.
RATIONALE:
In the past and actuality comp.compression has become extremly
inflated
due to an increasing amount of discussion related to random data
compression and or questions regarding patents about those.
The term "random data compression" refers to quite some group of
connected
abilities a compressor may develop. I'm trying to clarify the term for
people
participating in the discussion while not being involved in or being
connected to data compression:
* able to compress (means make smaller) any string
* thus able to compress recursively (may make smaller it's own output)
* thus also able to compress "highly dense" data which isn't
compressible
otherwise (like DNA) to extreme small sizes (human genome on a
floppy,
all linux distributions of all times on a single CD, ...)
Basically the entire topic revolves around doing wonderfull things
with
some digital data, so it turns out much much smaller. The common
approach to
reach that is mostly similar to developing mechanical perpetuum-mobile
machines. I don't want to stress any thoughts about the liability or
usability or correctness of the proposed claims and algorithms. There
seem
a lot people thinking about it and I think ("right" or "wrong") they
deserve
their space to discuss their ideas, as the other compression-
enthusiasts
deserve their space to discuss undisturbed.
Absolute numbers about the percentage of random compression related
posts are difficult to produce. Just looking at the last month we have
a coverage
of 75% related posts of all posts (~250 vs. ~60), which most probably
is not
representative over the entire year. But still I expect 25% of the
group's
content to turn out to be related to random data compression.
The problem is visible for allready a very long time, even mentioned
in the
fiveteen years old comp.compression FAQ: "[this topic can] generate a
lot of
activity on comp.compression, which can last for several months."
Personally being in my tenth year of reading comp.compression feel
extremely distracted not only by the "abuse" of the general
comp.compression
group for a very specific and (without being disrespectfull) otherwise
irrelevant topic to the majority of general comp.compression inquiries
and
discussions. Moreover it affects the common acceptance and credibility
of
comp.compression as a whole, the topics and discussions exchanged and
the
people participating in it. It especially leads to automatic
hostility,
protection and pre-justification mechanism of frequent poster in the
direction of new posters which just turn out to be friendly but miss-
understood basic concepts. In that sense we should try to seek to
regain
tolerance and equilibrium for all participants, and that means IMHO
adding comp.compression.random.
The suffix .random is designed to perfectly capture the attention of
individuals in that specific area of research and the to be estimated
amount of posts-absorbtion into the new group will hopefully free
relevant discussion space in the original group. It also gives any
topic
developed in comp.compression to side-step (Follow-Up) into
comp.compression.random when it becomes off-topic. Currently
everybody just has to bear it impassively.
In accordance with the guidelines for newgroup creation, discussion
about the creation of this newsgroup will take place in
"news.groups.proposals".
I urge you to support this proposal which has the potential to
significantly assist data compression readers and those simply being
curious about compression around the world.
CHARTER:
comp.compression.random is an unmoderated group for the discussion
of random data compression algorithms. The term random data
compression referrs to all kind of compression algorithms which
ought to compress either random data, or are able to compress any
data (at times recursively), or related data compression algorithms.
The group accepts discussions about patent protection for the
mentioned
topic as well.
PROCEDURE:
For more information on the newsgroup creation process, please see:
http://www.big-8.org/dokuwiki/doku.php?idpolicies:creation
Those who wish to influence the development of this RFD and its final
resolution should subscribe to news.groups.proposals and participate
in the
relevant threads in that newsgroup. This is both a courtesy to groups
in which discussion of creating a new group is off-topic as well as
the
best method of making sure that one's comments or criticisms are
heard.
All discussion of active proposals should be posted to
news.groups.proposals
To this end, the followup header of this RFD has been set to
news.groups.proposals.
If desired by the readership of closely affected groups, the
discussion
may be crossposted to those groups, but care must be taken to ensure
that all discussion appears in news.groups.proposals as well.
We urge those who would like to read or post in the proposed newsgroup
to make a comment to that effect in this thread; we ask proponents to
keep a list of such positive posts with the relevant message ID (e.g.,
Barney Fife, <4JGdnb60fsMzHA7Z...@sysmatrix.net>).
Such lists of positive feedback for the proposal may constitute good
evidence that the group will be well-used if it is created.
DISTRIBUTION:
This document has been posted to the following newsgroups:
news.announce.newgroups
news.groups.proposals
comp.compression
PROPONENT:
Niels Fr�hling <spam...@adsignum.com>
CHANGE HISTORY:
2009-06-09 1st RFD
Martin X. Moleski, SJ
On Wed, 10 Jun 2009 12:10:14 EDT, Metatron <spam...@adsignum.com> wrote in
<28fe085e-5b06-4e66...@e24g2000vbe.googlegroups.com>:
>RATIONALE:
>In the past and actuality comp.compression has become extremly
>inflated
>due to an increasing amount of discussion related to random data
>compression and or questions regarding patents about those.
A note on format for any subsequent RFDS: please use shorter
lines to avoid ugly wrap. Not a big deal; maybe it's my
setup that is doing the ugly wrapping.
I don't have a strong view on the merits of the proposal.
I'm not going to get involved in a technical discussion
about the possibilities of compressing random data.
The biggest problem with the proposal is the question of
whether there is more than one person interested in using
the group. My first impression is that you may be the
only person who wants the new group. A newsgroup with
no news is no fun. You need to beat the bushes to see
if you can find like-minded people to post in this
thread who would both read and post to the group.
Marty
--
Co-chair of the Big-8 Management Board (B8MB) <http://www.big-8.org>
Unless otherwise indicated, I speak for myself, not for the Board.
Jim Leonard
>
> We urge those who would like to read or post in the proposed newsgroup
> to make a comment to that effect in this thread; we ask proponents to
> keep a list of such positive posts with the relevant message ID (e.g.,
> Such lists of positive feedback for the proposal may constitute good
> evidence that the group will be well-used if it is created.
By all means, I wholeheartedly support the creation of
comp.compression.random if it means the signal-to-noise ratio will
increase in comp.compression.
Kathy Morgan
> The biggest problem with the proposal is the question of
> whether there is more than one person interested in using
> the group.
This is my biggest concern, also. If there are other people who plan to
post to the group regularly if it is formed, we need for them to come
here and say that. Simply having lurkers (readers) doesn't make for a
good group. Telling people they must move to a new group also won't
work unless those people want to move.
I'm worried that this may be a NIMBY (Not In My Back Yard) proposal,
where regulars in a group want to push some of the discussion into a
different group. If the members holding that discussion want to move to
a different group, that's great! If they don't want to move, creating
the group will not help the parent group. Instead, it will just add to
the noise with extra netcopping posts, "No, you can't talk about that
here, you have to move to the other group." "No, I don't want to move,
I'll talk about it here."
--
Kathy, co-chair B8MB, speaking only for myself
Kathy Morgan
> By all means, I wholeheartedly support the creation of
> comp.compression.random if it means the signal-to-noise ratio will
> increase in comp.compression.
It sounds like you would not plan on posting in comp.compression.random,
but would like to see the "random" discussion split away from
comp.compression. Please encourage others in comp.compression to post
to news.groups.proposals regarding whether they would post to the new
group if it is created and regarding whether they support or oppose
splitting the topic out of comp.compression.
As I said in my earlier post, NIMBY groups seldom succeed and can
sometimes just add to the noise in the original group because of
netcopping posts. If that's the case here, it might be much more
helpful for us to teach members of your group about using killfiles,
netiquette, and subject tagging to reduce the noise problems and make
the group more readable for everyone.
--
Kathy, co-chair of B8MB, speaking just for myself
Metatron
> >In the past and actuality comp.compression has become extremly
> >inflated
> >due to an increasing amount of discussion related to random data
> >compression and or questions regarding patents about those.
>
> A note on format for any subsequent RFDS: please use shorter
> lines to avoid ugly wrap. �Not a big deal; maybe it's my
> setup that is doing the ugly wrapping.
Yes, I don't know I expected 80, but it seemed actualy 70.
> I don't have a strong view on the merits of the proposal.
> I'm not going to get involved in a technical discussion
> about the possibilities of compressing random data.
Me neither. But I see the need that those talks will be seperated
from the regular talks, I do describe in the proposal why. This is a
social inquiry, not one about the technical differenciation and
credibility of the subject (random compression) itself.
> The biggest problem with the proposal is the question of
> whether there is more than one person interested in using
> the group.
As I wrote since fiveteen years there are enough people interested
(for example in trying to defeat the very idea of random compression
algorithms) that the group is fill with (like last month) 75% noise
about it.
> My first impression is that you may be the
> only person who wants the new group. �A newsgroup with
> no news is no fun. �You need to beat the bushes to see
> if you can find like-minded people to post in this
> thread who would both read and post to the group.
Well do you offer me the deal to exchange comp.compression.research
by comp.compression.random? Just kiddin.
You can be well assured that it is even a (very real) possibility
that .random has more posts at the end of the year than .compression.
That is not about that there are so much people inventing and
discussion their random compression algorithms, it's about the very
persistant crowd of people that try to make those understand. You
know, there (under this topic) are two sides than can endlessly talk
about that topic without giving way, and the threads bloat like a
atomic mushrooms. It is really really difficult to maintain a regular
interest for the general group if all you see is completly off-topic!
I can't make statistics about people going away from comp.compression
because of that ...
So the thing is I care about comp.compression, I like to talk about
algorithms there and help people regarding their own inventions or
whatever, using zlib. This space should be appropriate for that kind
of talk, that's why the group exists. And it's simply necessary to be
able to move this discussion into it's own space. I don't believe in
censorship/moderated, it's not necessary ... I don't want to say them
they are wrong either, it's just the wrong space and
comp.compression.random would be the right space.
I don't know, I would be surprised if I'm the only "interested" in
the seperation ... probably I would be very sad because I'd asume
nobody else really cares about comp.compression and it's essence ...
Well, this discussion is obviously about determining the state of
mind of my fellow readers and writer and my part I think is to
convince them, that it'd be better for all.
> � � � � � � � � � � � � � � � � Marty
Ciao
Niels
Metatron
> where regulars in a group want to push some of the discussion into a
> different group. �If the members holding that discussion want to move to
> a different group, that's great! �If they don't want to move, creating
> the group will not help the parent group. �Instead, it will just add to
> the noise with extra netcopping posts, "No, you can't talk about that
> here, you have to move to the other group." "No, I don't want to move,
> I'll talk about it here."
I do understand your argument. But I actually make the proposal not
to force anybody out, it's not even that I don't want the discussion
(Chavesque) ... The problem is that earlier or later somebody comes up
with the idea to make .compression moderated, and that's the worst
case that could have success.
They do want their space (there were just two expressing agreement, I
have to send them here ...), they should also have their space, and
they must not be muted (in my opinion). Their topic is also in need of
credibility, so they need their space also to not get flamed once
somebody just sees their name.
Look, both sides are really worn down. I do believe both sides can
and want to occupy and basically also interchange opinion. Just not
under this condition (an Encyclopedia is not for publishing conspiracy
theories, that doesn't mean they are reduntant or untrue or less
worth).
So from that point of view I do not want to kick them out. I really
am trying to be their advocate.
Do you actually have experience with the NIMBY case? How was the
mediation? You got a case reverted too?
Ciao
Niels
Thomas Richter
> REQUEST FOR DISCUSSION (RFD)
> unmoderated group comp.compression.random
>
> This is a formal Request for Discussion (RFD) for the creation of the
> unmoderated newsgroup comp.compression.random.
As much as I would like the idea, I afraid it wouldn't stop the usual
suspects from posting here, causing the distraction that should be
avoided in first place. (-;
So long,
Thomas
Martin X. Moleski, SJ
<54b89728-e060-480c...@v4g2000vba.googlegroups.com>:
> Well do you offer me the deal to exchange comp.compression.research
>by comp.compression.random? Just kiddin.
Heh heh. Thanks for the laugh. :o)
> You can be well assured that it is even a (very real) possibility
>that .random has more posts at the end of the year than .compression.
>That is not about that there are so much people inventing and
>discussion their random compression algorithms, it's about the very
>persistant crowd of people that try to make those understand.
What I would like to see is a fair number of those people post
here in n.g.p or c.c that they are ready and willing to move
their discussion to c.c.r. You judge that if there was a topic
space opened up for them, they would migrate. You may be right.
I'll feel a lot more inclined to vote yes if some of the
potential migrants endorse your view.
>You
>know, there (under this topic) are two sides than can endlessly talk
>about that topic without giving way, and the threads bloat like a
>atomic mushrooms.
Some people seem to enjoy making threads mushroom. Call
it recreational decompression maybe. They are performance
artists of a sort. But they need a foil for their
"art"--and an audience.
>It is really really difficult to maintain a regular
>interest for the general group if all you see is completly off-topic!
>I can't make statistics about people going away from comp.compression
>because of that ...
Understood.
> So the thing is I care about comp.compression, I like to talk about
>algorithms there and help people regarding their own inventions or
>whatever, using zlib. This space should be appropriate for that kind
>of talk, that's why the group exists. And it's simply necessary to be
>able to move this discussion into its own space.
I understand your motivation. I'm skeptical about the good
effects of creating the "right" group for the argumentative
types to move to. I don't have any training or expertise
in the area of compression--I'm just a grateful and satisfied
customer of the end product; but my Wildly Amateurish Guess
is that the random-compressionists are not wholly rational
people. If they had an algorithm that worked, wouldn't it
be on the market? They could silence their critics by
actually writing the programs instead of just visualizing
them in their imagination.
Such people may not want to move to the rational namespace
for their work.
>I don't believe in
>censorship/moderated, it's not necessary ... I don't want to say them
>they are wrong either, it's just the wrong space and
>comp.compression.random would be the right space.
Be honest now. You go ahead and say they're wrong. I won't
hold it against you. You imply it when you say that "it's
just the wrong space." If random data can be compressed,
it falls under the general idea of compression, doesn't
it? It can only be the "wrong space" if the backers of
the idea are somewhat wrong-headed.
> I don't know, I would be surprised if I'm the only "interested" in
>the seperation ... probably I would be very sad because I'd asume
>nobody else really cares about comp.compression and its essence ...
If we opened talk.nonsense (the talk.* hierarchy is a little
light on newsgroups), maybe we could get them all to go over
there. :o(
> Well, this discussion is obviously about determining the state of
>mind of my fellow readers and writer and my part I think is to
>convince them that it'd be better for all.
1. Convince them.
2. Get them to post in the relevant threads.
3. Collect the feedback.
4. Revise your RFD in view of the discussion, if needed.
5. When you think you've got as much support as you're
going to get, submit the Final RFD / Last Call for
Comments. When that's published in n.a.n, that will
kick off a process of formal voting on the part of
the board.
Metatron
Yes, but you are seeing it from a slightly skewed angle. The point is
not that if that group exists, that in .compression all can sleep in
peace. It's not the wonderweapon against an enemy, it's no weapon and
there are no enemies! People come as they can and as naive as they may
be, but in this moment they are shot on sight. You can't even give
them a hug, and say there they can really talk about their topic,
there are others that made all the similar stuff, there are also those
that are critic and possibly even very hard and honest. They can't
really consult the archive, they can't really follow the stories of
other 'achievements'.
In this very moment you and the others and me who are actually
talking (and not lurkers), we have no way, we are pressed into
comp.compression all together and each side (because it's a technical
issue!) has the absolute conviction of being morally above the others,
because we are caged in that same space.
And probably subconciously it's also because about the space itself,
because in the absence of a moral absolutum "who has the right to
suppress/fight/disturb my desire to talk about my invention, which is
about compression after all, and this is the space about
compression ..." and vice versa.
I won't be able to promise that we'll get the honey-and-milk country
I suggested above, but how much easier will mediation become in the
presence of that group. And how much more may be talked about the
subject itself, and how much more understanding may be gained just
because you are not in automatic defense or attack. Maybe even
humor. :^)
It has been proven every time and then again, since fiveteen years
that it is of interest to a group of people, they do search to
communicate it with everybody (for or against), and that it is of
interest to even suspicious people to support that in one or another
way (even something simple as to post a link to a past experience).
And actually I don't even believe that there really are people in
comp.compression that would deny .random just because of a moral
superiority, or a logic or scientific if you wish. I think it's clear
to most that the existance of a newsgroup has nothing to do with 'the
correctness of the topic'.
You wouldn't be against it, right? :) Why would you, if?
Ciao
Niels
Jacko
> � � � � � � � � � � � REQUEST FOR DISCUSSION (RFD)
> � � � � � � �unmoderated group comp.compression.random
>
> This is a formal Request for Discussion (RFD) for the creation of the
> unmoderated newsgroup comp.compression.random.
Sounds good, should really rename this 'left' group
comp.compression.entropic to really get shut of your lovely crowd.
comp.compression is just to general a term for your specifics.
> NEWSGROUPS LINE:
> comp.compression.random Random data compression algorithms and
> derivates.
>
> RATIONALE:
>
> In the past and actuality comp.compression has become extremly
> inflated
> due to an increasing amount of discussion related to random data
> compression and or questions regarding patents about those.
>
> The term "random data compression" refers to quite some group of
> connected
> abilities a compressor may develop. I'm trying to clarify the term for
> people
> participating in the discussion while not being involved in or being
> connected to data compression:
> * able to compress (means make smaller) any string
> * thus able to compress recursively (may make smaller it's own output)
> * thus also able to compress "highly dense" data which isn't
> compressible
this is just such a debatable point :-)
> � otherwise (like DNA) to extreme small sizes (human genome on a
> floppy,
> � all linux distributions of all times on a single CD, ...)
The linux one basically would be thread the compilier output.
> Basically the entire topic revolves around doing wonderfull things
> with
> some digital data, so it turns out much much smaller. The common
> approach to
> reach that is mostly similar to developing mechanical perpetuum-mobile
> machines. I don't want to stress any thoughts about the liability or
> usability or correctness of the proposed claims and algorithms. There
> seem
> a lot people thinking about it and I think ("right" or "wrong") they
> deserve
> their space to discuss their ideas, as the other compression-
> enthusiasts
> deserve their space to discuss undisturbed.
True true, too many disturbed people in this group ;-)
> Absolute numbers about the percentage of random compression related
> posts are difficult to produce. Just looking at the last month we have
> a coverage
> of 75% related posts of all posts (~250 vs. ~60), which most probably
> is not
> representative over the entire year. But still I expect 25% of the
> group's
> content to turn out to be related to random data compression.
And who claims the SPAM shopping percentage? Proportional
representation? Or just some fanciful mandate by default brokers with
expense accounts?
> The problem is visible for allready a very long time, even mentioned
> in the
> fiveteen years old comp.compression FAQ: "[this topic can] generate a
> lot of
> activity on comp.compression, which can last for several months."
Infinite discussion?
> Personally being in my tenth year of reading comp.compression feel
> extremely distracted not only by the "abuse" of the general
> comp.compression
> group for a very specific and (without being disrespectfull) otherwise
> irrelevant topic to the majority of general comp.compression inquiries
> and
> discussions. Moreover it affects the common acceptance and credibility
> of
Credability is so important in these un credited times ;-)
> comp.compression as a whole, the topics and discussions exchanged and
> the
> people participating in it. It especially leads to automatic
> hostility,
> protection and pre-justification mechanism of frequent poster in the
> direction of new posters which just turn out to be friendly but miss-
> understood basic concepts. In that sense we should try to seek to
> regain
> tolerance and equilibrium for all participants, and that means IMHO
> adding comp.compression.random.
The basic concepts are important. Who but an idiot would attempt
improvements without the mainstays of the subject and it's history.
> The suffix .random is designed to perfectly capture the attention of
> individuals in that specific area of research and the to be estimated
> amount of posts-absorbtion into the new group will hopefully free
> relevant discussion space in the original group. It also gives any
> topic
> developed in comp.compression to side-step (Follow-Up) into
> comp.compression.random when it becomes off-topic. Currently
> everybody just has to bear it impassively.
True.
> In accordance with the guidelines for newgroup creation, discussion
> about the creation of this newsgroup will take place in
> "news.groups.proposals".
Yes another quango group, why move shop when the issue is here?
> I urge you to support this proposal which has the potential to
> significantly assist data compression readers and those simply being
> curious about compression around the world.
Got my vote.
> Barney Fife, <4JGdnb60fsMzHA7ZnZ2dnUVZ_rWdn...@sysmatrix.net>).
> Such lists of positive feedback for the proposal may constitute good
> evidence that the group will be well-used if it is created.
>
> DISTRIBUTION:
>
> This document has been posted to the following newsgroups:
>
> � news.announce.newgroups
> � news.groups.proposals
> � comp.compression
>
> PROPONENT:
>
> Niels Fr�hling <spamt...@adsignum.com>
>
> CHANGE HISTORY:
>
> 2009-06-09 � � 1st RFD
cheers jacko
Steve Bonine
> I won't be able to promise that we'll get the honey-and-milk country
> I suggested above, but how much easier will mediation become in the
> presence of that group. And how much more may be talked about the
> subject itself, and how much more understanding may be gained just
> because you are not in automatic defense or attack. Maybe even
> humor. :^)
You are making two assumptions that, if they are correct, will improve
the discussion of the topic of compression.
1. You're assuming that all of the people who want to discuss the
subtopic of compressing random data will move to the new group, where
they will be able to discuss this topic without flames,
2. You're assuming that none of the flamers will bother with the new group.
> You wouldn't be against it, right? :) Why would you, if?
Because, if your two assumptions above are incorrect, creating the new
group will result in an entry in the list of groups, and nothing more.
It could, in fact, result in a degradation in the discussion in
comp.compression.
What happens if the new group is created, and no one uses it? In this
case we have a dead newsgroup in the list, which is not a big deal as it
joins hundreds of others. But what is the result for the existing
comp.compression group? In addition to all the noise that you've got
today, you'll have additional noise from people trying to get the
discussion moved to the "correct" newsgroup. The net result is negative.
And look at it from the perspective of the prospective users of
comp.compression.random. Why should they move? Do you really believe
that the new newsgroup will be "honey-and-milk"? Many of the folks who
feel that it's their moral duty to rebuke the folks who want to discuss
compression of random data will simply take their flames to the new
group. What's in it for the folks you're trying to get rid of? Nothing.
The success of the proposed group depends on whether people who want to
discuss compression of random data will move their discussion to the new
group. It does NOT depend on whether the people who do NOT want to
discuss that topic think that the new group is a good idea.
Frankly, given what I've seen so far (and it's early in the process), I
have to conclude that the people discussing compression of random data
would not move to a new group. The fact that they're currently
discussing the topic in a hostile environment suggests to me that
they're not particularly rational, and the fact that the new environment
would likely be just as hostile doesn't enhance the probability that
they will move.
This conclusion will change if I see a few people say that they will
post in the proposed group. So far that number is a good solid zero.
Kathy Morgan
> So the thing is I care about comp.compression, I like to talk about
> algorithms there and help people regarding their own inventions or
> whatever, using zlib. This space should be appropriate for that kind
> of talk, that's why the group exists. And it's simply necessary to be
> able to move this discussion into it's own space. I don't believe in
> censorship/moderated, it's not necessary ... I don't want to say them
> they are wrong either, it's just the wrong space and
> comp.compression.random would be the right space.
It sounds to me like the real problem is not the posts by people who
want to discuss compression of random strings, it's the resultant flames
and responses from people who do not believe random strings can be
compressed.
So far, no one has spoken up to say they want to use the proposed group;
all I have seen is people saying they want the random compression
discussion to leave comp.compression. I don't believe that would happen
even if we did create comp.compression.random.
It seems to me you have few viable choices: 1) You can maintain the
current status quo, with a low signal/noise ratio, 2) Those who do not
want to discuss random compression can learn to ignore the posts about
random compression and educate newcomers on how to respond only to posts
that do not involve random compression. (The second choice is made much
easier by using a good newsreader with killfile or scoring capabilities;
you and other regulars might need to abandon use of Google Groups and
learn to use a good newsreader.)
Another possibility would be to create a moderated companion group to
which posters like you, Jim Leonard and Thomas Richter could move.
There does not seem to be a desire for that so I don't believe it would
succeed.
--
Kathy, speaking only for myself
Metatron
> You are making two assumptions that, if they are correct, will improve
> the discussion of the topic of compression.
Yes, but I'm also aware that I can't through sheer will make it
happen and become good. I know I can't be the placeholder for those
who'd like to have the group, they have to stand for their own in the
final instance. But I hope I can kickstart those that are maybe
intimidated.
> The success of the proposed group depends on whether people who want to
> discuss compression of random data will move their discussion to the new
> group. �It does NOT depend on whether the people who do NOT want to
> discuss that topic think that the new group is a good idea.
I know, that's why I express the argument about the utility of the
new group in a positive way. I'm mean there is the chance in it, just
need to be taken.
> Frankly, given what I've seen so far (and it's early in the process), I
> have to conclude that the people discussing compression of random data
> would not move to a new group. �The fact that they're currently
> discussing the topic in a hostile environment suggests to me that
> they're not particularly rational, and the fact that the new environment
> would likely be just as hostile doesn't enhance the probability that
> they will move.
Well, I guess an additional presumption of mine is that people would
behave differently (in .random context) while still trying to
enlighten them. I think it's easier to be kind if you respect the
other sides position (and if just as a temporary hypothesis). But I'm
going in circles. :^)
> This conclusion will change if I see a few people say that they will
> post in the proposed group. �So far that number is a good solid zero.
Paul had something to say but apparently has technical difficulties:
-- snipp --
I have replied to group in the news.groups.proposals newsgroup, but I
never see my posts show up there. I get emails from the moderators of
the groups when I do, saying they received them, but I never see them
posted. So I just wanted to say that I support the desire for the new
group but not sure my voice is getting heard over in the
news.groups.proposals newsgroup.
Paul
-- snipp --
Ciao
Niels
Martin X. Moleski, SJ
<6866bdde-d82c-489b...@y9g2000yqg.googlegroups.com>:
>Paul had something to say but apparently has technical difficulties:
Paul has had three posts make it to n.g.p:
Message-ID: <fBSXl.1368$Gn4...@newsfe10.iad>
Message-ID: <TR8Yl.3990$gz5....@newsfe07.iad>
Message-ID: <TR8Yl.3987$gz5....@newsfe07.iad>
Thomas Richter
> On 10 Jun., 23:19, Thomas Richter <t...@math.tu-berlin.de> wrote:
>> Metatron wrote:
>>> REQUEST FOR DISCUSSION (RFD)
>>> unmoderated group comp.compression.random
>>> This is a formal Request for Discussion (RFD) for the creation of the
>>> unmoderated newsgroup comp.compression.random.
>> As much as I would like the idea, I afraid it wouldn't stop the usual
>> suspects from posting here, causing the distraction that should be
>> avoided in first place. (-;
>>
>> So long,
>> Thomas
>
> Yes, but you are seeing it from a slightly skewed angle. The point is
> not that if that group exists, that in .compression all can sleep in
> peace. It's not the wonderweapon against an enemy, it's no weapon and
> there are no enemies!
Please notice the "(-;" at the end - no there is no enemy, there are only
a) naive people, and
b) trolls
which causes a rather bad signal/noise ratio here, or in the terms of
the group, a rather low entropy that makes it hard to extract relevant
information.
> You wouldn't be against it, right? :) Why would you, if?
All you say is correct, but those issues cannot be addressed by a new
group. In fact, a group for the "reverse", namely "serious people from
research" does exist. It is comp.compression.research. It's just that
nobody uses it. IOW, comp.compression *is* already the "loony group" you
want to create (-: (Note the smiley!) in first place. Rather, the "we
believe in science" folks should move out and leave comp.compression for
others, since there *is* a group for all the die-hards that "believe"
that mathematics work and 1+1 = 2 holds true, even on Jupiter and on
Monday afternoons.
However, like it or not, c.c.research is *empty*. So what can be done
about it? The problem has been detected correctly, but unfortunately the
solution doesn't fit it.
There is a FAQ (nobody cared), there is Wikipedia, there are textbooks,
education in math in schools and in university... Neither will help,
only continuous effort *here* will. Except for a minority of
scam-artists, many posters claiming compression of i.i.d.
equal-distributed data - after all - never return, and I think that this
*is* a success; more for them then for the quality of the group. It *is*
the quality of the group since apparently those people learned
something, but most are (a very human move) too proud to admit their error.
IOW, why change something if it works?
So long,
Thomas
Metatron
> > >> Kathy, if the new group gets created, I will use it. �I'm one of those
> > >> people that get flamed often here in comp.compression.
>
> > > And rightly so. �I too am for this new group. �If we can sequester the
> > > troll population the world will be a better place.
>
> > > Tom
>
> > Exactly, Tom. �Good, we can count on you as another supporter. �;-) �Why
> > don't you go ahead and post in the news.groups.proposals group and tell them
> > your desire for it as well, that is unless your just trolling here.
>
> > Paul
>
> This is an excellent idea.
> ===================
>
> --jg
Jules, we are here together for a very long time, even maybe we
didn't intersect in our talks. Would you see a reason for a new group?
Or you allready got stone-skin after all this, and don't care? :)
Ciao
Niels
Metatron
> So far, no one has spoken up to say they want to use the proposed group;
> all I have seen is people saying they want the random compression
> discussion to leave comp.compression. �I don't believe that would happen
> even if we did create comp.compression.random.
I just want to hint hat Paul and Jules try to express their opinion,
but for the moment the posts only make it to comp.compression
(proposals seem not to be in the receipient-list). I hope you give
their opinion and willingness to talk about the new group credit, even
though the posts don't appear here. :)
I could redirect their posts, but I sort of have the feeling that
wouldn't be exactly right. We don't need to discuss this "only" here?
> you and other regulars might need to abandon use of Google Groups
I don't read news with GG, I just don't have a post-server, so I'm
always jumping there to write my posts. I can kill and filter, ain't
not giving me "better feelings".
Ciao
Niels
Metatron
> Kathy, if the new group gets created, I will use it. �I'm one of those
> people that get flamed often here in comp.compression.
>
> Paul
Which people would you hope to find there? You allready know what
you'd like to share or exchange or chat in a new group? Why do you
want the group?
Ciao
Niels
Metatron
wrote:
> >Paul had something to say but apparently has technical difficulties:
>
> Paul has had three posts make it to n.g.p:
Ah, okay. Very well, but strange. They don't appear on two of the
news-servers I'm able to consult. Anyway, if they arrive ...
Thanks
Niels
Kathy Morgan
> On Thu, 11 Jun 2009 21:23:55 CST, Metatron <spam...@adsignum.com> wrote in
> <6866bdde-d82c-489b...@y9g2000yqg.googlegroups.com>:
>
> >Paul had something to say but apparently has technical difficulties:
>
> Paul has had three posts make it to n.g.p:
>
> Message-ID: <fBSXl.1368$Gn4...@newsfe10.iad>
>
> Message-ID: <TR8Yl.3990$gz5....@newsfe07.iad>
>
> Message-ID: <TR8Yl.3987$gz5....@newsfe07.iad>
Hmmm....very interesting. None of those posts appear to have made it to
the Individual.Net server.
Niels, I would appreciate it if you would quote Paul's posts into ngp,
because I have not seen them here.
Thanks,
Kathy Morgan
> All you say is correct, but those issues cannot be addressed by a new
> group. In fact, a group for the "reverse", namely "serious people from
> research" does exist. It is comp.compression.research. It's just that
> nobody uses it.
Comp.compression.research is a moderated group whose moderator appears
to have left his post, either voluntarily or involuntarily. Does anyone
know who the most recent moderator of record is and whether he might be
willing to hand the group over to someone else? As long as there is no
active moderator to approve posts, no one can use
comp.compression.research. Would members of comp.compression be
interested in using comp.compression.research for a higher signal/noise
ratio if the group were again functional?
--
Kathy, speaking just for myself
Goldy
comp.compression love to flame those who are new to the topic of
"compressing random data".
People can't even ask a question the topic without getting 50+
messages from people telling them to stop, their wasting their time,
and that they are "trools" for posting such a thing.
In all honesty, I gave up on comp.compression for any real answers as
those who flame the new comers have become the trolls that are
addicted to running the group.
It is suggestions such as this which give me hope of a usenet group
without flaming trolls.
People should be able to come together, share ideas and information,
without having to deal with the assholes online that "know everything
about everything" and try to squash the dreams of others.
If my vote has a count, I think this should happen.
Hey Paul, I feel ya.
Goldy
> Because, if your two assumptions above are incorrect, creating the new
> group will result in an entry in the list of groups, and nothing more.
> It could, in fact, result in a degradation in the discussion in
> comp.compression.
>
> What happens if the new group is created, and no one uses it? �In this
> case we have a dead newsgroup in the list, which is not a big deal as it
> joins hundreds of others. �But what is the result for the existing
> comp.compression group? �In addition to all the noise that you've got
> today, you'll have additional noise from people trying to get the
> discussion moved to the "correct" newsgroup. �The net result is negative.
>
First off, I've gotten the impression that those who get flamed in
comp.compression is because people have NO OTHER NEWS GROUP to talk
about it in, hence this proposal.
Second, I can see the flamers in c.c telling those who end up in c.c
asking about R.D.C. being something along the lines of "You need to
join the nut jobs in comp.compression.random", not because their
assholes, but because they to are sick and tired of hearing people
talk about it. Both sides are sick of the other.
> And look at it from the perspective of the prospective users of
> comp.compression.random. �Why should they move? �Do you really believe
> that the new newsgroup will be "honey-and-milk"? �Many of the folks who
> feel that it's their moral duty to rebuke the folks who want to discuss
> compression of random data will simply take their flames to the new
> group. �What's in it for the folks you're trying to get rid of? �Nothing.
>
Probably not, their are always the assholes out their, and that will
never change, and some might come flame in the new group......shit
happens.
"moral duty to rebuke the folks..".....WTF? It's about numbers, not
morals. Morals and numbers are like oil and water.
> The success of the proposed group depends on whether people who want to
> discuss compression of random data will move their discussion to the new
> group. �It does NOT depend on whether the people who do NOT want to
> discuss that topic think that the new group is a good idea.
>
I WILL GLADLY MAKE THAT MOVE!
> Frankly, given what I've seen so far (and it's early in the process), I
> have to conclude that the people discussing compression of random data
> would not move to a new group. �
You think people don't want a place to go where they can meet other
like minded people? You are way wrong about that.
You think people enjoy being excited about their ideas to only be
flamed by "those who know"? Not the case.
> The fact that they're currently
> discussing the topic in a hostile environment suggests to me that
> they're not particularly rational, and the fact that the new environment
> would likely be just as hostile doesn't enhance the probability that
> they will move.
>
Again, the reason people end up in comp.compression is because no
other news group exist that they could choose from...ie, they didn't
have much of a choice.
It's perfectly rational, because MOST of the folks that came to
comp.compression asking about R.D.C. don't come back, BECAUSE THEY GET
FLAMED!
> This conclusion will change if I see a few people say that they will
> post in the proposed group. �So far that number is a good solid zero.
I WILL! Now it's one.
Take Paul, that makes two.
I can recall years and years ago when I started down the path of
research R.D.C.
All I could find was comp.compression and a few web sites. None of
the web sites had a chat or forum about the topic. So the only
logical choice was to go to comp.compression.
Now if I had to choose between comp.compression and
comp.compression.random, guess which one I would
choose......hrm.....see my point.
You can't say "they have a choice" when it's either comp.compression
or nothing.
Yes, their are other newsgroups, but comp.compression is the best fit
name that new comers find, and that's where they always end up because
they don't have much of a choice.
So before you state that people won't make the change, why don't we
get the group created, then see what happens after a year or so. Does
it hurt anyone to make a new news group???
If nobody post to the group, then feel free to come back and say I
told you so. Until then, please, keep an open mind and lets give it a
try.
Steve Bonine
The articles in question were posted from Highwinds and do not contain
Approved: headers. I can see them via Giganews, but they do not appear
at Google or at individual.net. My assumption is that Highwinds' server
has news.groups.proposals configured as a non-moderated newsgroup, and
that the Giganews servers don't require an Approved: header for incoming
articles from peers for moderated newsgroups.
Martin X. Moleski, SJ
<68f3808d-9aea-40bc...@y34g2000prb.googlegroups.com>:
> ... Does
>it hurt anyone to make a new news group?
As Steve Bonine indicated, the only problem with
adding a dead-on-arrival group is that it will
join hundreds of other dead groups on our
list.
There seem to be a sufficient number of servers
willing to go along with creation requests that
the group can probably be propagated enough to
consider it "created."
Getting rid of dead groups seems to be less
effective.
>If nobody post to the group, then feel free to come back and say I
>told you so. Until then, please, keep an open mind and let's give it a
>try.
It doesn't take a large number to persuade me to
vote for giving it a try. "2" is too small.
I'd like to see a minyan--at least ten.
A newsgroup with no news is no fun.
Marty
Metatron
> Comp.compression.research is a moderated group whose moderator appears
> to have left his post, either voluntarily or involuntarily. �Does anyone
> know who the most recent moderator of record is and whether he might be
> willing to hand the group over to someone else?
Ross Williams made the RFD and Voting, and initially the group was
proposed unmoderated. Peter Gutman was/is first moderator. I wrote
both a querry. Let's see.
Besides that I'd agree to take an active role if people would
seriously want to use .reseach for bleeding-edge-discussion, means I
would volunteer for moderator (allthough there are more 'qualified'
fellas, it's just adinistrative work after all). But first I'd
honestly try to make .research unmoderated. I do research, and I often
think my posts may be misplaced. .research has in the moment sort of
the image of the altar you look up to or something. More or less.
Probably because it is empty and moderated. LOL (I don't want to hurt
anybody's feeling about the church, but that's funny).
> As long as there is no
> active moderator to approve posts, no one can use
> comp.compression.research. �Would members of comp.compression be
> interested in using comp.compression.research for a higher signal/noise
> ratio if the group were again functional?
There are at least posts that qualify, there are several people
proposing new ideas and create and publish original work. I can even
imagine that posts which ask for an explanation of the inner workings
about compressors (not the implementation self-understood) would be
well seen there as well. Just to make a rough estimate I would not
expect the group to have more frequency than .random (I guess the two
group of people are equally big and animate (phun implied) ;^) ) There
once was .theory discussion but rejected, my archives doesn't reach
that far, so I can't say if the reasoning against that group is void
now.
How are the others seeing it? Mihai, Dave, Nemo, for example?
But that's not really related to this RFD, we'd have to make another
and try to revert the moderation-state, right?
Ciao
Niels
Kathy Morgan
Kathy Morgan
> On 12 Jun., 08:26, kmor...@spamcop.net (Kathy Morgan) wrote:
>
> > Comp.compression.research is a moderated group whose moderator appears
> > to have left his post, either voluntarily or involuntarily. Does anyone
> > know who the most recent moderator of record is and whether he might be
> > willing to hand the group over to someone else?
>
> Ross Williams made the RFD and Voting, and initially the group was
> proposed unmoderated. Peter Gutman was/is first moderator. I wrote
> both a querry. Let's see.
Thanks! Let us know the results.
> Besides that I'd agree to take an active role if people would
> seriously want to use .reseach for bleeding-edge-discussion, means I
> would volunteer for moderator (allthough there are more 'qualified'
> fellas, it's just adinistrative work after all). But first I'd
> honestly try to make .research unmoderated.
That is extremely unlikely to happen. The problem is that some servers
would honor a request to unmoderated the group, but others wouldn't.
This seriously fractures the discussion, because people using a server
that shows the group as moderated will not see any of the posts made by
people using servers that correctly show the group as unmoderated and
those servers would not propagate posts from servers that show it as
unmoderated. Since there would no longer be a moderator to approve
their posts, they're also unlikely to see their own messages to the
group. (A robot could be set up to automatically approve all messages,
but that means finding a volunteer to operate the 'bot.)
--
Kathy, member of B8MB, speaking only for myself.
Doug Freyburger
> kmor...@spamcop.net (Kathy Morgan) wrote:
>
> > Comp.compression.research is a moderated group whose moderator appears
> > to have left his post, either voluntarily or involuntarily. �Does anyone
> > know who the most recent moderator of record is and whether he might be
> > willing to hand the group over to someone else?
>
> seriously want to use .reseach for bleeding-edge-discussion, means I
> would volunteer for moderator (allthough there are more 'qualified'
> fellas, it's just adinistrative work after all).
You could post a Moderator Vacancy Investigation request.
> But first I'd
> honestly try to make .research unmoderated ...
Changing the moderator of a group is a matter of process.
Changing the moderation status of a newsgroup is completely
out of the question. It's been tried and the result was over a
decade of chaos. No such request will be approved.
> �How are the others seeing it? Mihai, Dave, Nemo, for example?
> �But that's not really related to this RFD, we'd have to make another
> and try to revert the moderation-state, right?
No, do an MVI and volunteer to take over the moderation.
If the .research group went active again, would your expected
content be natural in that group? The advantage of moderation
is doing carefully can ensure a flame-free environment in the
group. The disadvantage of moderation is the narrower topic
always results in lower traffic - The traffic level of moderated
newsgroups across UseNet is low and it becomes easy to
compare traffic and make incorrect conclusion based on that.
Metatron
> That is extremely unlikely to happen. �The problem is that some servers
> would honor a request to unmoderated the group, but others wouldn't.
> This seriously fractures the discussion, because people using a server
> that shows the group as moderated will not see any of the posts made by
> people using servers that correctly show the group as unmoderated and
> those servers would not propagate posts from servers that show it as
> unmoderated. �Since there would no longer be a moderator to approve
> their posts, they're also unlikely to see their own messages to the
> group. �(A robot could be set up to automatically approve all messages,
> but that means finding a volunteer to operate the 'bot.)
Uhm, how complicated. I have running dedicated servers I can run the
bot on, and I would. I'm also responsible enough to report on system-
failures, be it mine or of my resources (well except I get shot,
that's not such a weird idea in the country I'm living). Is just
missing the POST-server. I'd like to prevent direct cost (I'd
obviously absorb all my voluntary "cost"), so if possibly a University
or I don't know who could gift us that?
Ciao
Niels
Metatron
> You could post a Moderator Vacancy Investigation request.
> No, do an MVI and volunteer to take over the moderation.
Okay, I give Ross and Peter some time, and well whatever comes we
figure it our then.
> If the .research group went active again, would your expected
> content be natural in that group?
My expected content ... ? As I said there is content produced in c.c.
which fits perfectly. I wished at times I've could have been able to
post to c.c.r. Natural in the full sense of the name and charter, yes.
And with content comes probably also a better perception of where one
sees a post fit better, in c.c. or c.c.r.
> The advantage of moderation
> is doing carefully can ensure a flame-free environment in the
> group. �The disadvantage of moderation is the narrower topic
> always results in lower traffic - The traffic level of moderated
> newsgroups across UseNet is low and it becomes easy to
> compare traffic and make incorrect conclusion based on that.
The fun is that I think flaming is extremely unlikely in .research,
except one assumes/d that the random compression discussion would have
migrated there. Then flaming would possibly be even more agressive
(defending credibility).
Ciao
Niels
Metatron
> There is a FAQ (nobody cared)
Just a quick note about this. Under Mark's initiative we reworked the
FAQ, I even added relevant new sections; and I update sections; but I
feel incapable of maintaining the entire FAQ as a whole. I'm not
versatile/up-to-date/professional in each and every topic regarding
compression.
Still people post links to the outdated FAQ-archive. [Because] There
is also nobody sending FAQ-reminders (well Mark did, but yearly), and/
or the FAQ itself. Kathy and you are right in that we seem in need of
becoming a bit more structured, in that I also take responsability in
not having at least reacted to the obvious appropriately within the
group.
But that are three seperate things, determining what/how/where/
why .random (this is upto those supporting it for themself), what's up
with .research (this is a loong pending issue) and becoming
constructive in .compression. "Solving" one of it doesn't "solve" any
of the other.
Ciao
Niels
Dave Sill
>
> Changing the moderation status of a newsgroup is completely
> out of the question. It's been tried and the result was over a
> decade of chaos. No such request will be approved.
Let's just say it's very unlikely.
> No, do an MVI and volunteer to take over the moderation.
Yes, and there's nothing that says you can't set up a modbot that
automatically approves everything.
-Dave
Thomas Richter
> On 12 Jun., 01:24, Thomas Richter <t...@math.tu-berlin.de> wrote:
>
>> There is a FAQ (nobody cared)
>
> Just a quick note about this. Under Mark's initiative we reworked the
> FAQ, I even added relevant new sections; and I update sections; but I
> feel incapable of maintaining the entire FAQ as a whole. I'm not
> versatile/up-to-date/professional in each and every topic regarding
> compression.
Nobody is, not even the followers of the top-rated conferences in this
regard (DCC being one of them). However, I don't think that this is even
required. Most discussion here is on a level that doesn't require a top
scientist of the field, nor should it be. Research questions are
discussed by other means in first place. I'm perfectly fine with the FAQ
as is, and I don't want to imply that it requires updating. It requires
*being read*, that is the point. In a nutshell, 90% of the discussion
here would be gone if the following two principles would be encouraged:
o) Any function from a finite larger set onto a finite smaller set is
not one-to-one.
o) The core of science is independent verification.
I think that's all. Really! Everything beyond that can and should be
discussed here if related to compression. I would be glad if we would
have a couple substantial questions even for the commonly known methods
(Huffman, Lempel-Ziv, Burrows-Wheeler, JPEG), to re-answer them even on
a monthly schedule, and I'm also glad not being able to answer others
(How to design efficient multi-channel coding?), or helping with my
expertise (how to change progression in a JPEG 2000 file). All that is
on-topic, but only a minor percentage of the traffic here.
So long,
Thomas
Mark Adler
> This is a formal Request for Discussion (RFD) for the creation of the
> unmoderated newsgroup comp.compression.random.
Let's give it a shot. If it ends up in the realm of dead newsgroups,
so be it. Even a small chance of drawing the random compression
discussions away from comp.compression is a good thing.
I for one promise to never look or post there, in order to leave the
random compression posters in peace. Everyone else who understands the
mapping of finite sets should do the same. We can't expect all of the
random compression developers to remain sequestered there, since some
of them are trolling for responses from the stodgy minds stuck in the
boring world of mathematical proof, in order to bolster their
self-image of maverick free-thinking leading to fame and riches.
However some of them also seem to actually want to discuss whatever it
is they want to discuss without constantly being interrupted by those
who feel impelled for some reason to point out how they are provably
wrong. I find this apparently relentless motivation to teach the
unteachable to be almost as perplexing as the pursuit of random
compression itself. Just let it go.
As it currently stands, the signal to noise ratio on comp.compression
is about minus 20 dB. Create the group, and let's see if we can gain
at least a few dB.
Now if we could get all the spammers to post to
comp.compression.advertisements ...
Mark
Paul W. Schleck
Hash: SHA1
>Mark
Sounds like "random" compression as proposed by one side in this debate
is perceived by the other side to be junk science, either provably wrong
or at least "non-falsifiable" (i.e., not testable or (dis)provable).
This seems analogous to arguing over the capabilities of perpetual
motion machines (or communications systems that exceed the information
and error rates predicted by Shannon's Law), when mainstream science
views them as impossible.
Earlier in Usenet's history, when the contents of a given newsgroup
about science (such as comp.ai or talk.origins) descended into arguments
over things that were perceived by their communities to not be rigorous
science, the solution was to moderate the newsgroup in-place.
Since that is not longer possible, the choices now are to either
reactivate a dead moderated newsgroup for compression,
comp.compression.research (hoping that the "rigorous" compression
participants move there) or make a throwaway newsgroup for discussion
about the junk science (hoping that the "nuts" move there and take up a
new home, possibly giving them a platform for recruiting naive others
into their nuttiness). Lovely.
If random compression is the perpetual motion machine of compression,
then I don't see the point trying to make it the topic of a Big-8
newsgroup. I presume that the Big-8 Board would hesitate to approve a
newsgroup whose participants intended to incite, defend, or conspire to
commit unlawful activity, for example. I would hope also that they
would hesitate to approve a newsgroup that would be about purely junk
science, and in essence, a "con-job." One question that I would have
would be, how long has this conflict been going on, and how likely is
random compression to be a short-lived fad that will blow over and
eventually leave the comp.compression newsgroup in peace, at least until
the next fad comes along? Another question would be, is there any
interest in reactivating the moderated newsgroup
comp.compression.research for discussion of compression theory that is
rigorous and supported by the mainstream scientific community?
Assistance would be available if interested volunteers can be found.
The moderated newsgroup would also solve the SPAM problem.
With regard to what is provably wrong or theoretically impossible in
compression, I defer to one of the luminaries in the field, Khalid
Sayood, who has participated very rarely on comp.compression (and sadly,
not recently, likely due in part to some of these conflicts in the
current makeup of the newsgroup). The most salient point that he makes
below is, "Claims are generally of the form 'given ..... this is the
best way ..' If you change the givens, the claim no longer holds."
Improvements over Huffman compression, via arithmetic encoding, were
possible because arithmetic coding was able to find a way around one of
the givens (a codebook of size 2^n), while still obeying the laws of
information theory. Are both sides of the random compression debate at
least talking about the same "given"'s?
>Newsgroups: comp.compression
>Path: sparky!uunet!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu!news.unomaha.edu!unlinfo.unl.edu!eecom2!ksayood
>From: ksay...@eecom2.unl.edu (Khalid Sayood)
>Subject: Re: 16:1 compression - why not?
>Message-ID: <1992Apr19....@unlinfo.unl.edu>
>Sender: n...@unlinfo.unl.edu
>Nntp-Posting-Host: eecom2.unl.edu
>Organization: University of Nebraska-Lincoln
>References: <39229.11...@kcbbs.gen.nz>
>Date: Sun, 19 Apr 1992 20:34:57 GMT
>Lines: 42
In article <39229.1109223...@kcbbs.gen.nz> Peter_Gutm...@kcbbs.gen.nz (Peter Gutmann) writes:
>Reply-To: pg...@cs.aukuni.ac.nz
>
>I've been reading the string of postings about why the WEB compressor can't
>achieve 16:1 compression on all files "because the laws of information theory
>say so". Well until a few years ago you couldn't get any better encoding of
>data than Huffman encoding "because the laws of information theory say so"
>(until various people discovered arithmetic coding, that is).
[stuff deleted]
>
>(OK, that's poured out the petrol, someone wanna throw on a match? :-)
>
Anything is better than grading so here goes:
Let H be the entropy of the source which you want to code n letters at
a time. Huffman coding gaurantees you a coding rate R where
H<= R <= H + /n. (this bound can be tightened but the form a/n is still
the same). By making n large we can come as close to H as we want.
Unfortunately making n large we also make the Huffman codebook size very
large. For example if our source alphabet is binary, for n=1, the codebook
size is 2, for n=2, the codebook size is 4 ( 00 01 10 11) and for n = k
the codebook size is 2**k. Arithmetic coding gaurantees you a coding rate
R where H<= R <= 2/n. At first glance this seems worse than Huffman coding.
However in the case of Arithmetic coding, to encode a string of length n
you do not need to build a codebook of size 2**n. That's all. The
appearance of Arithmetic Coding did not violate any previous "laws".
A good reference for all of this is "Elements of Information Theory,"
by Cover and Thomas, published by Wiley.
As far as I know nobody ever said you could not get an encoding of data
better than Huffman encoding "because of the laws of information theory".
Claims are generally of the form "given ..... this is the best way .."
If you change the givens, the claim no longer holds.
As far as the WEB claim is considered, I suppose anything is possible
in an infinite universe :^)
Khalid
- --
Paul W. Schleck
psch...@novia.net
http://www.novia.net/~pschleck/
Finger psch...@novia.net for PGP Public Key
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (SunOS)
iD8DBQFKM/cA6Pj0az779o4RApHVAJ92V9CIc9cvS3RJ5gerLRu0nj5XNgCeOprO
QXJ/ViKZubMmFpOtVhqz/d0=
=MHSm
-----END PGP SIGNATURE-----
Adam H. Kerman
Does it? What will the Board do about getting Highwinds to process
checkgroups?
Steve Bonine
We (the moderators of news.groups.proposals) are addressing this
situation, now that we know that it exists. The chance of success in
this endeavor is higher for someone who is a customer of Highwinds, so
if anyone would like to make a positive contribution and take on the
challenge of contacting Highwinds' support as a customer, that would be
much appreciated.
Disclaimer: At this point we have three articles that don't contain an
Approved: header that appeared to come from Highwinds. That suggests,
but certainly doesn't prove, that there's a misconfiguration there.
Steve Bonine
head moderator for news.groups.proposals
Martin X. Moleski, SJ
>Sounds like "random" compression as proposed by one side in this debate
>is perceived by the other side to be junk science, either provably wrong
>or at least "non-falsifiable" (i.e., not testable or (dis)provable).
>This seems analogous to arguing over the capabilities of perpetual
>motion machines (or communications systems that exceed the information
>and error rates predicted by Shannon's Law), when mainstream science
>views them as impossible.
It sounds that way to me, too.
> ... the choices now are to either
>reactivate a dead moderated newsgroup for compression,
>comp.compression.research (hoping that the "rigorous" compression
>participants move there) ...
I would prefer that to making "a throwaway newsgroup for discussion
about the junk science" (the next line in your post).
> ... I presume that the Big-8 Board would hesitate to approve a
>newsgroup whose participants intended to incite, defend, or conspire to
>commit unlawful activity, for example.
True.
>I would hope also that they
>would hesitate to approve a newsgroup that would be about purely junk
>science, and in essence, a "con-job."
If there was a large enough core who wanted to use the group,
I would probably vote in favor of it. Let the reader beware!
I don't see that the random compressionists want to make the
move (except for two, I think).
I've done my share of helping to create groups that were
DOA. But when I voted for them, I thought there was at
least some hope that they might thrive eventually. I
am not seeing any indication that c.c.r fits that bill.
Adam H. Kerman
>Adam H. Kerman wrote:
>>Kathy Morgan <kmo...@spamcop.net> wrote:
>>>Steve Bonine <s...@pobox.com> wrote:
>>>>Metatron wrote:
>>>>>On 12 Jun., 01:24, "Martin X. Moleski, SJ" <mole...@canisius.edu> wrote:
>>>>>>>Paul had something to say but apparently has technical difficulties:
>>>>>>Paul has had three posts make it to n.g.p:
>>>>>Ah, okay. Very well, but strange. They don't appear on two of the
>>>>>news-servers I'm able to consult. Anyway, if they arrive ...
>>>>The articles in question were posted from Highwinds and do not contain
>>>>Approved: headers. I can see them via Giganews, but they do not appear
>>>>at Google or at individual.net. My assumption is that Highwinds' server
>>>>has news.groups.proposals configured as a non-moderated newsgroup, and
>>>>that the Giganews servers don't require an Approved: header for incoming
>>>>articles from peers for moderated newsgroups.
>>>Thanks, Steve. That explanation makes sense.
>>Does it? What will the Board do about getting Highwinds to process
>>checkgroups?
>We (the moderators of news.groups.proposals) are addressing this
>situation, now that we know that it exists.
The Board was the proponent of this newsgroup. It's an issue for the
proponent to deal with. That's why I put the question that way to
Kathy. You aren't at fault here.
>The chance of success in this endeavor is higher for someone who is
>a customer of Highwinds, so if anyone would like to make a positive
>contribution and take on the challenge of contacting Highwinds' support
>as a customer, that would be much appreciated.
Highwinds does not have individual subscribers.
Highwinds is a backend Usenet provider. Newshosting.com, formerly a
backend provider and in-house reseller, merged into Highwinds some time
ago and is now only an in-house reseller of Highwinds. easynews.com and
Usenetserver.com are now Highwinds in-house resellers too. To complicate
matters further, the three in-house resellers are also intermediaries to
other resellers.
Highwinds has different servers farms. To track down the error, the
error would have to be reported to the particular reseller that the
author subscribes to, so they can get it fixed.
fwiw, the propagation search engine checks easynews but the moderation
flag is missing. anchordude http://www.newsgroupservers.net/#biglist
has never changed its message that EasyNews continues to be screwed up.
Martin X. Moleski, SJ
>The Board was the proponent of this newsgroup.
I'm pretty sure the proponent is:
Message-ID: <28fe085e-5b06-4e66...@e24g2000vbe.googlegroups.com>
> Niels Fr�hling <spam...@adsignum.com>
He's also known as Megatron in his posts to this list.
Adam H. Kerman
>>The Board was the proponent of this newsgroup.
>I'm pretty sure the proponent is:
>Message-ID: <28fe085e-5b06-4e66...@e24g2000vbe.googlegroups.com>
>>Niels Fr�hling <spam...@adsignum.com>
>He's also known as Megatron in his posts to this list.
Sigh.
The affected newsgroup with the misset moderation flag is
news.groups.proposals. The Board was the proponent of
news.groups.proposals, not Steve Bonine. Getting misset moderation flags
corrected is the proponent's job. It's the Board's responsibility
to promote implementation of checkgroups, generally.
Someone on the Board needs to follow through and not drop the ball.
Contact the author of the message that wasn't submitted for approval to
learn which Highwinds reseller he subscribes to, then report the error
through that reseller. Different Highwinds resellers use different
server farms. It's going to take a lot of follow through to resolve the
issue. It would be lovely if the Board got itself a technical contact at
the appropriate server farm to promote checkgroups.
Mark Adler
> Co-chair of the Big-8 Management Board (B8MB) <http://www.big-8.org>
> Unless otherwise indicated, I speak for myself, not for the Board.
I was curious what the "Big-8" board is, but that link returns an empty
response after redirecting to a php page.
Mark
Mark Adler
> One question that I would have
> would be, how long has this conflict been going on, and how likely is
> random compression to be a short-lived fad that will blow over and
> eventually leave the comp.compression newsgroup in peace, at least until
> the next fad comes along?
It has been going on as long as I've been reading the newsgroup, almost
20 years. There is zero probability that it is a short-lived fad. As
the old proponents fade away or die, new younger ones replace them. At
the same time, eager and vociferous naysayers are continually
replenished as the older ones get tired.
> Another question would be, is there any
> interest in reactivating the moderated newsgroup
> comp.compression.research for discussion of compression theory that is
> rigorous and supported by the mainstream scientific community?
As far as I know, no one is stepping up to be a moderator. Also there
needs to be a comp.compression on which newbies with legitimate
questions can come get legitimate answers and not be drowned in a sea
of noise. comp.compression.research was and would be for specialists.
> Are both sides of the random compression debate at
> least talking about the same "given"'s?
No. For some odd reason the other side doesn't seem to understand that
you can't map a larger set one-to-one to a smaller set. Either that,
or they don't understand that for a compressor to be lossless, you also
have to be able to decompress and get the original bits back.
Mark
Martin X. Moleski, SJ
>On 2009-06-13 14:31:38 -0700, "Martin X. Moleski, SJ"
><mol...@canisius.edu> said:
>> <http://www.big-8.org>
>I was curious what the "Big-8" board is, but that link returns an empty
>response after redirecting to a php page.
The sites seems to be down.
I've e-mailed our service provider.
I did a draft of an article on big-8 management
a couple of years ago. It's here:
<http://en.wikipedia.org/wiki/User:Mxmsj/Usenet>
After working on it for a few days, I realized it
wouldn't pass Neutral Point of View standards and
moved it to the (presently zoned out) big-8 wiki.
Marty
--
Martin X. Moleski, SJ
> ... The Board was the proponent of
>news.groups.proposals, not Steve Bonine. ...
Thanks for the clarification and instructions.
Metatron
> > �Ross Williams made the RFD and Voting, and initially the group was
> > proposed unmoderated. Peter Gutman was/is first moderator. I wrote
> > both a querry. Let's see.
>
> Thanks! �Let us know the results.
Okay, I have very nice result to report. Peter is active and alive,
and would make me moderator or co-moderator of c.c.r. I would boy POST
on "news.individual.net" (or any news-server you can suggest me for
text-group moderation, I have no interest in any alt.* or binary
access whatsoever).
Then I'm going to set up assisted robo-moderatrion and will go ahead
and try to reastablich reminders for the "new" FAQ as well as actively
inviting people to use c.c.r. for research-related on-topic
discussions. I have anyway other projects under the (voluntar)
supervision that defenitely fullfill c.c.r. charter which means we're
going to at least have some repetitive auto-activity to bring to mind
that that group indeed is not dead!
Very much thanks for triggering my initiative into this direction.
Ciao
Niels
P.S.: It's Me_t_atron (look in the Kabalah, he's the keeper of
information/librarian, and that's what I am), I don't feel Me_g_a, and
it's not my choice yet (GG). I'll be back, with real-name. ;)
Metatron
> Since that is not longer possible, the choices now are to either
> reactivate a dead moderated newsgroup for compression,
> comp.compression.research (hoping that the "rigorous" compression
> participants move there) or make a throwaway newsgroup for discussion
> about the junk science (hoping that the "nuts" move there and take up a
> new home, possibly giving them a platform for recruiting naive others
> into their nuttiness). �Lovely.
Well, let's say in a wide sense of the charter and/or the name you
can discuss any "legitimite" compression and or coding of random
(appearing) sources in the group. That means also you can discuss ways
to not expand random sources too much with a specific compressor, or
which criteria lead to real "bad" expansion. That's up to the people
how they feel to put into context their exchange.
And even if that one claim of recursive compression may be critizised
or prooven wrong, that doesn't mean there must never come out
something else practical out of that area of research, or that there
is no respectable talent in those who embraced the topic. That's why I
think we shouldn't fight about the truthfullness of that one claim
(out of much related) and play ursupator of the topic as a whole (and
that's what's happening, there is no space left to talk about chances,
in the end all is just filled up with denial).
> Another question would be, is there any
> interest in reactivating the moderated newsgroup
> comp.compression.research for discussion of compression theory that is
> rigorous and supported by the mainstream scientific community?
On it's way.
Ciao
Niels
Martin X. Moleski, SJ
<be9a5aa0-7d27-47de...@g19g2000yql.googlegroups.com>:
> P.S.: It's Me_t_atron (look in the Kabalah, he's the keeper of
>information/librarian, and that's what I am), I don't feel Me_g_a, and
>it's not my choice yet (GG). I'll be back, with real-name. ;)
My apologies!
I'm not all done with stupid. There's lots more where
that came from. :o(
Metatron
> So what is the verdict? �Are we going to get the comp.compression.random
> group? �If not, then I can open my forum up to a section devoted to this and
> bypass usenet all together. �If there is that much red tape it may not be
> worth it. �I already have my own random compression project established
> there as a thread (www.tretbase.com/forum) and I could open up a subforum
> dedicated to the subject.
>
> Paul
The discussion period is maybe a month (round-about I think),
definitely longer than the 6 days we've had yet. This is to give
infrequent readers and posters a chance to participate in the
discussion.
After that we ask the board if we need to collect last comments, and/
or we can proceed to vote. You have to convince the board and/or
enough people in comp.compression which convince the board that the
intentions in creating the newsgroup are not only serious but also
sensefull.
Several board-members stated that they will support the creation of
the group if/when there comes enough positive feedback from
proponents, and not so much "storage track" feedback from RDC
opponents.
That means you should for example try to find discussion partners in-
and outside of comp.compression which express their (active) support
for the group. Martin suggested an amount of ca. 10 people (which is
not very much, comp.compression.research was (almost) unanimously
agreed with 312 yes-sayers, and the group hasn't been used since
(sarc.). :^)
You see it's not much, but at least that should be it. Otherwise (and
I actually agree on that) even the call-for-vote may be rejected.
Ciao
Niels
Earl_Colby_Pottinger
working magic compressor) will talk to?
All of them first start off starting they can't talk about the details
of thier design because they are going to make a fortune off the
patent.
Will it be full of people all boasting they can do it but they can't
tell?
How many, even at this present day peak they are only 3-4 people
posting about thier magic compressors.
How will supply them any help?
At-least half of them show up here claiming to prove their ideas on
paper or on a spreadsheet and are begging for someone to write the
code for them, if they can't get programming help now how is this new
group going to find them programmers?
Personal needs?
Most magic compressor authors seem to come here to boast how smart
they are (if not why don't they just write code, patent it and make
money?), who can they boast to if the group has only a handful of
users and they all boast of being able to do the same thing?
Scammers?
The few who seem to be scammers come to comp.compression to either
drum up business or to claim that this group is researching thier
product. A smaller (unknown) group is not going have much pull with
them.
Niels Fröhling
> REQUEST FOR DISCUSSION (RFD)
> unmoderated group comp.compression.random
>
> This is a formal Request for Discussion (RFD) for the creation of the
> unmoderated newsgroup comp.compression.random.
>
> comp.compression.random Random data compression algorithms and
> derivates.
Okay, I think some time passed. I also think in regard to support for the
group we've been left sort of unconclusive. As much as I'd like to cite humility
as sufficient reason to create a group, it's not. Also I'd much like to at least
make the experiment/experience with the group, but I understand the character of
usenet as well.
Having said that you may understand that I still am a promoter of the group.
I've tried to leave space for those which really would be in need of it.
Let's have a look at Marty's list:
1. Convince them.
2. Get them to post in the relevant threads.
3. Collect the feedback.
4. Revise your RFD in view of the discussion, if needed.
5. When you think you've got as much support as you're
going to get, submit the Final RFD / Last Call for
Comments. When that's published in n.a.n, that will
kick off a process of formal voting on the part of
the board.
1. Well ...
2. Nope, not much ...
3. Pretty negative (negative in the sense of embracement)
4. ?
5. That's what I'd like to have your opinion about. Is it worth? Honestly I
think sadly the message (being a chance instead of a damnation) didn't find
consensus. We would need at least a hundred to vote (right?), I have no
experience with making an estimate about that, but it appears unlikely even that.
I though have to say we've got another chance, which is
comp.compression.research - for having triggered that I've to thank you all who
participated in the discussion.
Ciao (with realname now ;^)
Niels
Martin X. Moleski, SJ
> ... We would need at least a hundred to vote (right?) ...
No.
I've updated the FAQ:
<http://www.big-8.org/dokuwiki/doku.php?id=faqs:creation>
Steve Bonine
> Okay, I think some time passed. I also think in regard to support for
> the group we've been left sort of unconclusive. As much as I'd like to
> cite humility as sufficient reason to create a group, it's not. Also I'd
> much like to at least make the experiment/experience with the group, but
> I understand the character of usenet as well.
I agree that there is no clearcut evidence either that the proposed
newsgroup is a good or bad idea. It would have been nice to see an
outpouring of support from folks who stated that they would post in the
new group; instead we've mostly seen people who (paraphrasing in a
somewhat negative way) would love to see the discussion move away from
the existing newsgroup, but have no intention of participating in the
new group.
> Having said that you may understand that I still am a promoter of the
> group. I've tried to leave space for those which really would be in need
> of it.
Indeed, the question boils down to whether the folks currently
discussing the topic of random compression will use the new group. Do
they really want to conduct a serious discussion with people who hold
similar ideas, or are they more interested in baiting folks who disagree
with them? I suspect that both motives exist.
> Honestly I think sadly the message (being a chance instead of a
> damnation) didn't find consensus. We would need at least a hundred to
> vote (right?), I have no experience with making an estimate about that,
> but it appears unlikely even that.
No, you don't need 100. You need four. There are currently seven
people on the big-8 board, and you need a majority for the group to be
created. If you still feel that there is a reasonable chance for the
group to succeed, I'd suggest that you submit an RFC/LCC and let the
board decide whether to create the newsgroup. After all, that's how
they earn their handsome paychecks.
> I though have to say we've got another chance, which is
> comp.compression.research - for having triggered that I've to thank you
> all who participated in the discussion.
In any case, comp.compression.research sounds like a good idea. But it
doesn't solve the problem that you set out to solve in the unmoderated
group. Nothing you can do will _solve_ that problem, but it might be an
interesting experiment to create the new newsgroup and observe what
effect it does have on the traffic in comp.compression.
Niels Fröhling
>> Honestly I think sadly the message (being a chance instead of a
>> damnation) didn't find consensus. We would need at least a hundred to
>> vote (right?), I have no experience with making an estimate about
>> that, but it appears unlikely even that.
>
> No, you don't need 100. You need four. There are currently seven
> people on the big-8 board, and you need a majority for the group to be
> created. If you still feel that there is a reasonable chance for the
> group to succeed, I'd suggest that you submit an RFC/LCC and let the
> board decide whether to create the newsgroup. After all, that's how
> they earn their handsome paychecks.
Uh, okay. So no public votes anymore.
Before I send the LCC, are there any objections about the charter? I
understand the rationale is informal, anyway somebody likes to see a more
refined rationale containing more "positive" reasoning? Are there formal
objections (newsgroupline fe.)?
>> I though have to say we've got another chance, which is
>> comp.compression.research - for having triggered that I've to thank
>> you all who participated in the discussion.
>
> In any case, comp.compression.research sounds like a good idea. But it
> doesn't solve the problem that you set out to solve in the unmoderated
> group.
That's right.
> Nothing you can do will _solve_ that problem, but it might be an
> interesting experiment to create the new newsgroup and observe what
> effect it does have on the traffic in comp.compression.
Okay. We have a little snapshot of the situation, some opinions, and we let
the board think and decide about the matter.
Thanks
Niels