Sun currently ships with CDE, the Common Desktop Environment, which is based on Motif and the Motif widget set. I find it quite clunky and cumbersome -- FVWM has it licked on most counts. SGI's Irix is similarly based on a Motif-like system though with some nice chrome. I believe the display mechanisms are different though (is Irix X based?).
The only really outstanding commercial Unix desktop predating Linux that I'm aware of is NeXT, and I'm currently running the highly NeXT-influenced WindowMaker. Note that I didn't run NeXT, I just find wm quite the bomb.
Sun previously shipped with OpenLook / OpenWindows (also X based) and NeWS (Network Extensible Windowing System). Hewlett Packard had it's VUE (Visual User Environment) which highly influenced CDE. There's more general history here.
Frankly, though, I have to agree with the general assessment that there was no really good integrated graphical environment for Unix prior to Linux and the FSF. For all the comparisons made between Linux and consumer desktops, you'll note that there are none made between GNOME/KDE and commercial Unix offerings, or to Windows 3.x. The development path for a best-of-breed desktop for Linux, bar none, is clearly established.
FWIW, and so long as we're talking fluff, my first impression of a publicity shot of Noah Wyle was he's more of a Linus than a Jobs. Maybe next year when we get "Attack of the Screaming Penguin Hoardes"....
A signed binary buys you a degree of assurance that the binary is what you think it is. It is useful for the person installing the binary on their machine.
A signed binary does not buy others confidence in the results of the binary -- they can see neither the binary itself nor its signature. This is a general problem in distributed computing, SETI@Home is just an interesting case.
My understanding of digital signatures (the probable solution, IMO) is that they rely on a secret maintained by a user, where the integrity of the secret can be trusted. With a binary, you have an object which may be accessed by several users, and either the object or the user may be trying to hoodwink you. What would be required is some way to at least verify that the the binary itself is what it purports to be. I see a challenge-response of some sort on data transfer as the only real way for this to happen, however the mechanism would have to rely on something not apparent in either the source code or the binary itself.
Possibly a twist on public key encryption, where the compilation host generates a key pair for each binary (or for a collective set of binaries), in which one portion of the key is coded into the binary and its pair is retained by a trusted host. This would allow for public distribution of code, however production results would have to come from a binary compiled on a known trusted host.
I'm talking through my hat and making this up as I go. The idea sounds intriguing however.
I was reading with some interest the notes at SETI@Home discussing why an open source client wasn't being distributed. Apparently there are concerns with what would happen if an OSS client were hacked to provide wrong answers. I have two key thoughts.
First, closing the source apparently isn't preventing exploits based on the existing client. Whether or not anyone's tried to reverse engineer the code or not I'm not sure, but it's probably a matter of time before an RE or exploit-capable client is produced.
Second, this problem is something which OSS must face in general, particularly in a distributed computing environment. While digital signatures can be used to validate individuals and email, I'm wondering if similar means can be used to verify a program and its results. What SETI@Home needs is a way to distribute its source (to get benefits of OSS development), but to be able to mark the "canonized" version of the code in such a way that a non-forgeable signature can be attached to results and not just the code.
I'd be interested in knowing how or whether this issue is being persued elsewhere.
Huh? I've taken quite a few economics courses in my day, but I don't get this statement, unless of course you are saying something similar to car theft reduces auto prices, because thieves get free cars.
The earlier respondant points in the general direction, and my earlier essay discusses this in depth, but to reiterate:
Software piracy is not the same as theft of real property -- rather than redistributing a fixed quantity of goods (theft), piracy introduces a new supply of goods (the pirated goods) with a lower production curve (the production costs are lower). In a market sense, you've created a larger supply with a lower cost -- in an S-D (supply-demand) diagram, the pirate supply curve is to the south-west (lower and to the left) of the legitimate supply.
Assuming demand (economic sense, not SIIA's) doesn't change, the price must be lower. Depending on the shape of the pirate supply curve, either more or fewer copies of the software will be distributed. The only way (in a free market) the legitimate distributer can increase his own sales and profits is to lower his price (moving down on his supply curve). Result: fewer legitimate SW sales, lower cost.
Other alternatives are to affect market demand (make piracy a less attractive option) or pirate supply (make the cost of business for a pirate too high). Interestingly, one option is to periodically undercut the pirates -- presumably they're out to make money, so by removing their profit incentive, piracy is reduced.
Logic, and pointers to references (this is all standard microeconomics) are in the previously referenced essay.
WRT business deductions for losses -- what is the logic that says piracy losses are not a deductable business loss?
While software piracy has real economic effects, the methods and conclusions stated by the BSA, SIIA (formerly the SPA) and other software industry groups are grossly overstated. Moreover, piracy actually reduces the cost of consumer software -- this is just simple economics at work. Yes, piracy also reduces software revenues, but the consumer impact is often either misstated or unstated.
I wrote a response to last year's SPA report. As the SIIA is repeating its rediculous revenue loss figures, I will continue to promote the piece. Specifically, SIIA's definitions of "supply" and "demand" have absolutely no relationship to the same terms as used in economics, and the "loss" estimates are merely the street value of pirated softare, not the lost business opportunity.
Here's a point to ponder. If software companies are booking these losses as tax writeoffs, this is a tremendous fraud being practiced to the cost of US taxpayers.
Karsten M. Self What part of "Gestalt" don't you understand?
...can be run as a swapfile. Note: parameters may be varied, particularly the 'count' option. 16MB is a relatively minimal amount of swap, but sufficient for testing purposes.
You'll probably want to add the last to your/etc/rc.d/rc.system file to automatically turn on swap at boot.
...yes, performance suffers mildly, but on today's quick li'l boxen, the casual user won't know any better or notice, and converting to full-on Linux might provide an added benefit or two....
does just this. It's an option on the SuSE 5.2+ installs.
It allows a "demo" install of Linux on a DOS or VFAT filesystem, or a cohabitating Linux installation without requiring a repartition of your harddrive. It allows me to say my mom has Linux on her computer (does she use it, well, that's another story...).
"So why not make a "slimmed-down" Linux for them."
I'm having a similar discussion on another front, though the issue is modular tools in general, not merely Linux.
First response is that Red Hat (and SuSE, IIRC) allow you to roll out different configurations of Linux, including some relatively basic workstation configs.
The kicker though (these aren't my words, I'm your view) is that people will frequently (often?) go for the gusto and install absolutely everything. Especially first timers. We've been told that more is more so often that "less is more" sounds false.
IMO, stripped Linux installs will be the only way to roll for a corporate desktop. The home user, however, will probably feel cheated if that's the way it's presented. However, existing products such as the Cobal Qube, Trinux, and the Netwinder are in fact tailored Linux distros, so you're predicting the past to a certain extent.
Things will work out eventually.
Unix shell optimized for pessimum keyboard
on
The Myth of QWERTY
·
· Score: 2
Ditto on the vi comments. The Unix shell and utilities are optimized for a pessimal keyboard layout. While Dvorak's splitting of vowel and consonant keys between right and left hands works well for standard written text (and is slanted toward English), a both the tendency to abbreviate by dropping vowels, and the choice of balanced keystroke commands, strongly favors use of QWERTY.
Just typing 'ls' (r-ring, l-ring to r-pinky+, r-pinky) a half-dozen times killed the joy. vi and other utilities which rely on a "logical" placement of movement keys are utterly broken by Dvorak. Such is the power of lock-in.
That said, I found I had a more natural rhythem using Dvorak for text and documents.
In the same way, Lesstif is a rip-off of Motif, gcc is a reimplementation of c. There are times when a strict reimplementation is justified -- impelementing a truely free development tool or toolkit is one.
Proprietary software has done this as well -- there's more than one C compiler out there. But it's also done this for end-user apps. Word and Quattro both had emulation modes for WordPerfect and Lotus 1-2-3, respectively. Borland's implementation of the Lotus UI was supported by the Supreme Court in a landmark case.
I've also got it on good authority that proprietary developers are gleening features from free software. Idea appropriation goes both ways.
Moderation indicates both preferences and controversy. I would expect that there will be some abuse of the moderation system, in the sense that moderators will (conciously or unconciously) vote their personal preferences. But with 400 people out there, there's a lot of opinion to be divided up. It should work out.
Still, I think the current score is only part of the story. It would be really nice to have the number of moderation votes as an additional indicator. Comments with a high score and high number of votes are highly favored. Comments with a very low score and a high number of votes are pretty unambiguously bad. But a netral (~= 0) score with a large number of votes indicates a strong set of differences in the moderators' community. These articles are important, even if they aren't agreeable to all.
You should realize that it's quite probable that moderators won't be seeing posts which have slipped below their own threshold settings. Once a post goes below see level, so to speak, it won't keep getting thwacked. A really beligerent post might get swatted simultaneously by several moderators (threshholds don't apply until you reload the page), but this should not be a major issue.
I would hope that moderators have better things to do with their time then finding out what the lowest scored comment in a discussion is and drilling it into the floor.
I've suggested a couple of different moderation schemes in the past. Most of the serious needs are addressed on the client side, not the user side -- what you are asking for is the ability to view articles which have provoked strong reactions up or down. To this end, an "absolute score" filter would be good. Posts which are merely offensive or annoying would likely not get a strong reaction, just a mild whack-a-troll.
It's also not clear how conflicting moderation gets mitigated -- if you have four moderators scoring an article down 1 point, and two scoring it up two, does it net out to zero? Moderators get limited points, so there's a disincentive to vote on something which has already been moderated in the direction the moderator would have it done.
Your floating point suggestion is good, but it's only part of the story. Ideally, I'd like to see a score given a few different ways:
Aggregate score (-1 = score Scorings (number of times the article has been moderated): 1, 2, 3,...
Aggreement ratio -- number of posts in agreement or out of agreement. High level of moderation and low level of agreement (~= 50%). Controversial articles are likely to have relatively low absolute scores but a high number of moderators.
A combination of net score and number of scores would be a good indication of how good and/or controversial a post was. High score, many scores -- good post. Low score, many scores -- bad post. Middlin score (~=0), many scores -- raging controversy.
I was thinking about this this morning, what I'd really like to see is MUSENET -- moderated USENET. This would be, like USENET, a distributed database, but of ratings, not articles. Backward compatible, natch. Completely independent of the client software, so if you really wanted a feature, you could add it. I'd like it to be more of an 'attributes database', where attributes could be arbitrarily created. Attribute attributes might be (key, value, reference, owner, expiry, permissions). Not sure what else would be required. Basically, what you want to assign, the weight assigned, what it refers to, who created it, when it expires (if ever), and who may modify it (if anyone). Expiry and permissions might be advisory only, but the addition of a signature might add some additional assurances to the system.
It's probably a lot of baggage to carry around, but I'd invite others to kick the idea around.
Code Complete by Steve McConnell (Microsoft Press, 1993) is both a good introduction to software quality practices (on the applied, not theoretical, level), and has a very good bibliography of prior works, including PCP, MMM, Peopleware, and other classics.
Despite the irony of being published by Microsoft, this really is quite a solid book. It's the best of the three titles produced by McConnell (others are Rapid Development and Software Project Survival Guide). McConnell maintains related information at his web site.
If you care to research the issue, this is known as the "resale doctrine" of copyright. Licensed software is arguably not a sale of copyrighted materials (the software is covered by copyright, but you have merely licensed, not bought it). Pick a fight with a lawyer or go look at misc.int-property or Findlaw for more information.
Subject: [svlug] Other report posted on freebsd list Date: 15 Feb 1999 23:59:26 GMT From: marc_news@merlins.org (Marc MERLIN) Organization: Private Linux Box proudly running a modified RedHat Linux 5.2 To: svlug@svlug.org
From: Gregory Sutter To: freebsd-advocacy@FreeBSD.ORG, freebsd-chat@FreeBSD.ORG, announce@bafug.org Subject: Windows Refund Day: Bay Area Report Date: Mon, 15 Feb 1999 15:14:38 -0800
FreeBSD folk,
This report is being written from the CoffeeNet, where we've all met after the event in Foster City. We didn't exactly "storm the Gates", but both the Linux and FreeBSD communities turned out in force, accompanied by a couple of others (I saw a Sun and an SCO person). We got to the parking garage where the Microsoft "welcome table" was set up. There were Microsoft-sponsored refreshments and a lot of press, including many major stations.
A few of us grabbed some press members with cameras and headed for the main entrance. We arrived and were greeted by a friendly Microsoft security guard, who told us that we were not allowed to go up to the ninth floor (where the Microsoft offices are). When the press members began questioning her, she called another person, who apparently called security, because they showed up long before any Microsoft representative would talk to us.
I was interviewed by several press members at this time, right from the lobby of the Microsoft office. Several others had arrived by this time and we were becoming more forceful in our attempts to get to the Microsoft Office. People began getting on the elevators, only to find that they had been locked down so that nobody could visit the ninth floor. those going to the tenth and using the stairs found that the stair doors were locked from the inside as well, so nobody actually got into the office. (Later, they locked the elevators down completely.)
After twenty minutes or so of increased numbers of refund attempters, a person claiming to be a Microsoft representative appeared. While several of us attempted to get some straight answers out of him, he would only give us the typical Microsoft doubletalk. He handed out a sheet explaining Microsoft's statement on the refund policy[1] and would only echo what was contained on that paper. Although several tried, we were unable to get him to admit even that this was Microsoft policy and not just a random statement.
Around this time, the rest of the group arrived and the press began seriously interviewing everyone. I saw Eric Raymond and many others being questioned repeatedly on the purpose of the gathering and whether Microsoft had issued a statement. There was a _lot_ of mass media presence at the event, and Microsoft's attempts to stonewall us at the door didn't impress anyone. I look forward to the news reports tonight and tomorrow.
Toward the end, we all just stood in the courtyard and kibitzed before finally breaking up around 1:45 to return to our regularly scheduled activities (and a fine gathering at the CoffeeNet). Gregory Sutter, reporting for OSS News[2], signing off.
[1] For the full document, see after 20:00 PST today.
[2] and Daemon News.
Greg -- Gregory S. Sutter My reality check just bounced. mailto:gsutter@pobox.com http://www.pobox.com/~gsutter/ PGP DSS public key 0x40AE3052 ------------------------------------------------ ----------------------------
Marc -- "Microsoft is to software what McDonalds is to gourmet cooking"
Home page: http://marc.merlins.org/ (friendly to non IE browsers) Finger marc_f@merlins.org for PGP key and other contact information
-- echo "unsubscribe svlug" | mail majordomo@svlug.org ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^ to unsubscribe see http://www.svlug.org/mdstuff/lists.shtml for posting guidelines.
I'm assuming you used the tag to space out your chart. If this appeared correctly in preview mode, but was rendered incorrectly in the submitted comment, be aware that Slashdot tends to translate character codes into the resolved character, then submit these. I've had similar problmes with the > and <& (< and >) tags.
The trick is to preview your post, then go back to the original edit screen and clean up any errors. You can't do 'edit' => 'preview' => 'submit'.
The 212% growth for Linux cited in various places refers I believe to unit sales, not unit installations. Actual unit growth appears to be about 100% (7.5 million to 15 million), according to the Red Hat survey numbers. See this article for more details.
This is the problem when looking at growth rates: you have to keep track of what it is that's growing (and what the percentage basis is).
Credit card -- consumer protection
on
Toshiba and EULA
·
· Score: 1
Don't quote me on this. Check your credit card customer agreement before trying this at home.
Some credit cards may offer you a "consumer protection plan", where if you dispute a charge for specified reasons (I'm not sure how broad), then you may be able to decline charges through the credit card company.
I repeat: Check your credit card customer agreement before trying this at home.
If this is the case, then the handy thing is that the payment dispute is between your credit card bank and the vendor. Simplifies the hassles.
Things may not go completely smoothly, and you stand a good chance of fscking up your credit record by trying to pull a stunt like this.
It would also be interesting to see performance figures for the tested systems (as well as OS/2, and Novell) where custom tuning has been applied to the systems and applications -- runtime options in the case of IIS, compilation options and runtime options in the case of Samba and Apache.
Isn't one mode of this compromise mirroring software which doesn't validate what it is mirroring? It's bad enough that one site throws up compromised source. The cracker is relatively assured his/her work will be propogated worldwide to mirror sites in a matter of hours.
Mirroring software must check PGP signatures.
Hell, the FTP sites must check PGP sigs.
And the installation software must check PGP sigs.
Bruce Perens is quite right that crypto is the solution.
Slashdot Mirror
Sun currently ships with CDE, the Common Desktop Environment, which is based on Motif and the Motif widget set. I find it quite clunky and cumbersome -- FVWM has it licked on most counts. SGI's Irix is similarly based on a Motif-like system though with some nice chrome. I believe the display mechanisms are different though (is Irix X based?).
The only really outstanding commercial Unix desktop predating Linux that I'm aware of is NeXT, and I'm currently running the highly NeXT-influenced WindowMaker. Note that I didn't run NeXT, I just find wm quite the bomb.
Sun previously shipped with OpenLook / OpenWindows (also X based) and NeWS (Network Extensible Windowing System). Hewlett Packard had it's VUE (Visual User Environment) which highly influenced CDE. There's more general history here.
Frankly, though, I have to agree with the general assessment that there was no really good integrated graphical environment for Unix prior to Linux and the FSF. For all the comparisons made between Linux and consumer desktops, you'll note that there are none made between GNOME/KDE and commercial Unix offerings, or to Windows 3.x. The development path for a best-of-breed desktop for Linux, bar none, is clearly established.
FWIW, and so long as we're talking fluff, my first impression of a publicity shot of Noah Wyle was he's more of a Linus than a Jobs. Maybe next year when we get "Attack of the Screaming Penguin Hoardes"....
This pic shows the resemblance pretty well....
If you're still listening....
Do you have more info on just how this happens? The FAQ and webpages at Netrek don't really get into details. Sounds like it's up my alley though.
Respond via email (unmunge address).
A signed binary buys you a degree of assurance that the binary is what you think it is. It is useful for the person installing the binary on their machine.
A signed binary does not buy others confidence in the results of the binary -- they can see neither the binary itself nor its signature. This is a general problem in distributed computing, SETI@Home is just an interesting case.
My understanding of digital signatures (the probable solution, IMO) is that they rely on a secret maintained by a user, where the integrity of the secret can be trusted. With a binary, you have an object which may be accessed by several users, and either the object or the user may be trying to hoodwink you. What would be required is some way to at least verify that the the binary itself is what it purports to be. I see a challenge-response of some sort on data transfer as the only real way for this to happen, however the mechanism would have to rely on something not apparent in either the source code or the binary itself.
Possibly a twist on public key encryption, where the compilation host generates a key pair for each binary (or for a collective set of binaries), in which one portion of the key is coded into the binary and its pair is retained by a trusted host. This would allow for public distribution of code, however production results would have to come from a binary compiled on a known trusted host.
I'm talking through my hat and making this up as I go. The idea sounds intriguing however.
I was reading with some interest the notes at SETI@Home discussing why an open source client wasn't being distributed. Apparently there are concerns with what would happen if an OSS client were hacked to provide wrong answers. I have two key thoughts.
First, closing the source apparently isn't preventing exploits based on the existing client. Whether or not anyone's tried to reverse engineer the code or not I'm not sure, but it's probably a matter of time before an RE or exploit-capable client is produced.
Second, this problem is something which OSS must face in general, particularly in a distributed computing environment. While digital signatures can be used to validate individuals and email, I'm wondering if similar means can be used to verify a program and its results. What SETI@Home needs is a way to distribute its source (to get benefits of OSS development), but to be able to mark the "canonized" version of the code in such a way that a non-forgeable signature can be attached to results and not just the code.
I'd be interested in knowing how or whether this issue is being persued elsewhere.
Huh? I've taken quite a few economics courses in my day, but I don't get this statement, unless of course you are saying something similar to car theft reduces auto prices, because thieves get free cars.
The earlier respondant points in the general direction, and my earlier essay discusses this in depth, but to reiterate:
Software piracy is not the same as theft of real property -- rather than redistributing a fixed quantity of goods (theft), piracy introduces a new supply of goods (the pirated goods) with a lower production curve (the production costs are lower). In a market sense, you've created a larger supply with a lower cost -- in an S-D (supply-demand) diagram, the pirate supply curve is to the south-west (lower and to the left) of the legitimate supply.
Assuming demand (economic sense, not SIIA's) doesn't change, the price must be lower. Depending on the shape of the pirate supply curve, either more or fewer copies of the software will be distributed. The only way (in a free market) the legitimate distributer can increase his own sales and profits is to lower his price (moving down on his supply curve). Result: fewer legitimate SW sales, lower cost.
Other alternatives are to affect market demand (make piracy a less attractive option) or pirate supply (make the cost of business for a pirate too high). Interestingly, one option is to periodically undercut the pirates -- presumably they're out to make money, so by removing their profit incentive, piracy is reduced.
Logic, and pointers to references (this is all standard microeconomics) are in the previously referenced essay.
WRT business deductions for losses -- what is the logic that says piracy losses are not a deductable business loss?
While software piracy has real economic effects, the methods and conclusions stated by the BSA, SIIA (formerly the SPA) and other software industry groups are grossly overstated. Moreover, piracy actually reduces the cost of consumer software -- this is just simple economics at work. Yes, piracy also reduces software revenues, but the consumer impact is often either misstated or unstated.
I wrote a response to last year's SPA report. As the SIIA is repeating its rediculous revenue loss figures, I will continue to promote the piece. Specifically, SIIA's definitions of "supply" and "demand" have absolutely no relationship to the same terms as used in economics, and the "loss" estimates are merely the street value of pirated softare, not the lost business opportunity.
Here's a point to ponder. If software companies are booking these losses as tax writeoffs, this is a tremendous fraud being practiced to the cost of US taxpayers.
Karsten M. Self
What part of "Gestalt" don't you understand?
They'd have few points to begin with and spend them rather quickly, with 4,000 moderators out there to play whack-a-mole. Not a problem, IMO.
# dd if=/dev/zero of=/swapfile bs=1024 count=16384 /swapfile /swapfile
# sync
# mkswap
# swapon
You'll probably want to add the last to your /etc/rc.d/rc.system file to automatically turn on swap at boot.
...yes, performance suffers mildly, but on today's quick li'l boxen, the casual user won't know any better or notice, and converting to full-on Linux might provide an added benefit or two....
It allows a "demo" install of Linux on a DOS or VFAT filesystem, or a cohabitating Linux installation without requiring a repartition of your harddrive. It allows me to say my mom has Linux on her computer (does she use it, well, that's another story...).
"So why not make a "slimmed-down" Linux for them."
I'm having a similar discussion on another front, though the issue is modular tools in general, not merely Linux.
First response is that Red Hat (and SuSE, IIRC) allow you to roll out different configurations of Linux, including some relatively basic workstation configs.
The kicker though (these aren't my words, I'm your view) is that people will frequently (often?) go for the gusto and install absolutely everything. Especially first timers. We've been told that more is more so often that "less is more" sounds false.
IMO, stripped Linux installs will be the only way to roll for a corporate desktop. The home user, however, will probably feel cheated if that's the way it's presented. However, existing products such as the Cobal Qube, Trinux, and the Netwinder are in fact tailored Linux distros, so you're predicting the past to a certain extent.
Things will work out eventually.
Just typing 'ls' (r-ring, l-ring to r-pinky+, r-pinky) a half-dozen times killed the joy. vi and other utilities which rely on a "logical" placement of movement keys are utterly broken by Dvorak. Such is the power of lock-in.
That said, I found I had a more natural rhythem using Dvorak for text and documents.
BTW: pessimum = ! optimum. The Yellow Book.
In the same way, Lesstif is a rip-off of Motif, gcc is a reimplementation of c. There are times when a strict reimplementation is justified -- impelementing a truely free development tool or toolkit is one.
Proprietary software has done this as well -- there's more than one C compiler out there. But it's also done this for end-user apps. Word and Quattro both had emulation modes for WordPerfect and Lotus 1-2-3, respectively. Borland's implementation of the Lotus UI was supported by the Supreme Court in a landmark case.
I've also got it on good authority that proprietary developers are gleening features from free software. Idea appropriation goes both ways.
Moderation indicates both preferences and controversy. I would expect that there will be some abuse of the moderation system, in the sense that moderators will (conciously or unconciously) vote their personal preferences. But with 400 people out there, there's a lot of opinion to be divided up. It should work out.
Still, I think the current score is only part of the story. It would be really nice to have the number of moderation votes as an additional indicator. Comments with a high score and high number of votes are highly favored. Comments with a very low score and a high number of votes are pretty unambiguously bad. But a netral (~= 0) score with a large number of votes indicates a strong set of differences in the moderators' community. These articles are important, even if they aren't agreeable to all.
You should realize that it's quite probable that moderators won't be seeing posts which have slipped below their own threshold settings. Once a post goes below see level, so to speak, it won't keep getting thwacked. A really beligerent post might get swatted simultaneously by several moderators (threshholds don't apply until you reload the page), but this should not be a major issue.
I would hope that moderators have better things to do with their time then finding out what the lowest scored comment in a discussion is and drilling it into the floor.
Some good points.
I've suggested a couple of different moderation schemes in the past. Most of the serious needs are addressed on the client side, not the user side -- what you are asking for is the ability to view articles which have provoked strong reactions up or down. To this end, an "absolute score" filter would be good. Posts which are merely offensive or annoying would likely not get a strong reaction, just a mild whack-a-troll.
It's also not clear how conflicting moderation gets mitigated -- if you have four moderators scoring an article down 1 point, and two scoring it up two, does it net out to zero? Moderators get limited points, so there's a disincentive to vote on something which has already been moderated in the direction the moderator would have it done.
Your floating point suggestion is good, but it's only part of the story. Ideally, I'd like to see a score given a few different ways:
A combination of net score and number of scores would be a good indication of how good and/or controversial a post was. High score, many scores -- good post. Low score, many scores -- bad post. Middlin score (~=0), many scores -- raging controversy.
I was thinking about this this morning, what I'd really like to see is MUSENET -- moderated USENET. This would be, like USENET, a distributed database, but of ratings, not articles. Backward compatible, natch. Completely independent of the client software, so if you really wanted a feature, you could add it. I'd like it to be more of an 'attributes database', where attributes could be arbitrarily created. Attribute attributes might be (key, value, reference, owner, expiry, permissions). Not sure what else would be required. Basically, what you want to assign, the weight assigned, what it refers to, who created it, when it expires (if ever), and who may modify it (if anyone). Expiry and permissions might be advisory only, but the addition of a signature might add some additional assurances to the system.
It's probably a lot of baggage to carry around, but I'd invite others to kick the idea around.
Despite the irony of being published by Microsoft, this really is quite a solid book. It's the best of the three titles produced by McConnell (others are Rapid Development and Software Project Survival Guide). McConnell maintains related information at his web site.
If you care to research the issue, this is known as the "resale doctrine" of copyright. Licensed software is arguably not a sale of copyrighted materials (the software is covered by copyright, but you have merely licensed, not bought it). Pick a fight with a lawyer or go look at misc.int-property or Findlaw for more information.
Subject:
[svlug] Other report posted on freebsd list
Date:
15 Feb 1999 23:59:26 GMT
From:
marc_news@merlins.org (Marc MERLIN)
Organization:
Private Linux Box proudly running a modified RedHat Linux 5.2
To:
svlug@svlug.org
-----------------------------------------------
From: Gregory Sutter
To: freebsd-advocacy@FreeBSD.ORG, freebsd-chat@FreeBSD.ORG, announce@bafug.org
Subject: Windows Refund Day: Bay Area Report
Date: Mon, 15 Feb 1999 15:14:38 -0800
FreeBSD folk,
This report is being written from the CoffeeNet, where we've all met
after the event in Foster City. We didn't exactly "storm the Gates",
but both the Linux and FreeBSD communities turned out in force,
accompanied by a couple of others (I saw a Sun and an SCO person).
We got to the parking garage where the Microsoft "welcome table" was
set up. There were Microsoft-sponsored refreshments and a lot of
press, including many major stations.
A few of us grabbed some press members with cameras and headed for
the main entrance. We arrived and were greeted by a friendly
Microsoft security guard, who told us that we were not allowed to go
up to the ninth floor (where the Microsoft offices are). When the
press members began questioning her, she called another person, who
apparently called security, because they showed up long before any
Microsoft representative would talk to us.
I was interviewed by several press members at this time, right from
the lobby of the Microsoft office. Several others had arrived by this
time and we were becoming more forceful in our attempts to get to the
Microsoft Office. People began getting on the elevators, only to find
that they had been locked down so that nobody could visit the ninth
floor. those going to the tenth and using the stairs found that the
stair doors were locked from the inside as well, so nobody actually got
into the office. (Later, they locked the elevators down completely.)
After twenty minutes or so of increased numbers of refund attempters, a
person claiming to be a Microsoft representative appeared. While
several of us attempted to get some straight answers out of him, he
would only give us the typical Microsoft doubletalk. He handed out a
sheet explaining Microsoft's statement on the refund policy[1] and would
only echo what was contained on that paper. Although several tried, we
were unable to get him to admit even that this was Microsoft policy and
not just a random statement.
Around this time, the rest of the group arrived and the press began
seriously interviewing everyone. I saw Eric Raymond and many others
being questioned repeatedly on the purpose of the gathering and whether
Microsoft had issued a statement. There was a _lot_ of mass media
presence at the event, and Microsoft's attempts to stonewall us at the
door didn't impress anyone. I look forward to the news reports tonight
and tomorrow.
Toward the end, we all just stood in the courtyard and kibitzed before
finally breaking up around 1:45 to return to our regularly scheduled
activities (and a fine gathering at the CoffeeNet). Gregory Sutter,
reporting for OSS News[2], signing off.
[1] For the full document, see
after 20:00 PST today.
[2] and Daemon News.
Greg
--
Gregory S. Sutter My reality check just bounced.
mailto:gsutter@pobox.com
http://www.pobox.com/~gsutter/
PGP DSS public key 0x40AE3052
-----------------------------------------------
Marc
--
"Microsoft is to software what McDonalds is to gourmet cooking"
Home page: http://marc.merlins.org/ (friendly to non IE browsers)
Finger marc_f@merlins.org for PGP key and other contact information
--
echo "unsubscribe svlug" | mail majordomo@svlug.org
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
see http://www.svlug.org/mdstuff/lists.shtml for posting guidelines.
Interesting, never knew of this. Obligatory links: 1, 2.
Note that Hanson was President of the United States in Congress Assembled", not "President of the United States", the current office.
The trick is to preview your post, then go back to the original edit screen and clean up any errors. You can't do 'edit' => 'preview' => 'submit'.
This is the problem when looking at growth rates: you have to keep track of what it is that's growing (and what the percentage basis is).
Some credit cards may offer you a "consumer protection plan", where if you dispute a charge for specified reasons (I'm not sure how broad), then you may be able to decline charges through the credit card company.
I repeat: Check your credit card customer agreement before trying this at home.
If this is the case, then the handy thing is that the payment dispute is between your credit card bank and the vendor. Simplifies the hassles.
Things may not go completely smoothly, and you stand a good chance of fscking up your credit record by trying to pull a stunt like this.
The value of OSS is the code. Use it!
Mirroring software must check PGP signatures.
Hell, the FTP sites must check PGP sigs.
And the installation software must check PGP sigs.
Bruce Perens is quite right that crypto is the solution.