I think everyone should listen to "American Woman" at least one, and I don't see a GOOD reason why they should have to pay for it. That money surely isn't going as an incentive for Hendrix to produce new music.
Jimi did a cover version of "American Woman"? Hold on a second... ssh, ssh, cd blah, fire up giFTcurs... search for American Woman... while that's running, try google... oh wait, am I going to get sued for this?
(The only versions of "American Woman" I've heard are the original by the Guess Who, and the Lenny Kravitz cover. If there's a Jimi cover, I want it, and eventually I'll find it... but so far I'm not seeing any evidence of one.)
Just about every P2P uses hashes to uniquely identify files on the network.
Correct. Except that many of the popular networks only hash the first 300k or so of the file. More mature networks hash the entire file.
So if you rip Nirvana's smells like teen spirit from the orginal cd into mp3 and then put it into your share directory for your P2P client it gets a unique hash assigned by your P2P client.
A hash isn't assigned. It's computed. It's a mathematical function that calculates a "checksum" of the input data. Same input, same hash. Different input, probably a different hash. The goal of a good hash function is to produce a different output for every input, insofar as that is possible. Changing one byte of the input file should produce a wildly different hash value. Hash collisions are to be avoided.
Now lets say a friend of yours does the exact same thing but his mp3 will have a different hash than yours because his P2P client creates a unique hash for his rip.
No. If his rip is byte-for-byte identical to yours, then it will have the same hash as yours. That's how hash functions work. The reason his rip gets a different hash is because it's a different rip. Reading CDDA is nondeterministic; you get single-bit errors in the data. And even if you had precisely the same raw audio data during the rip, you'd probably use a different encoder, or a different version of the encoder, or different encoder settings. Or you'd put different tags on the encoded file (even changing the title inside an ID3v2 or Vorbis tag from "Home By The Sea" to "Home by the Sea" will produce a different hash).
(Insert description of "swarming" here.)
Along comes RIAA or an agent for it. They catalog a list of all the known hashes for the song smell's like teen spirit [...]
I will grant that they may have accumulated a catalog (not a complete one!) of hashes of encodings of "Smells Like Teen Spirit" which happen to be in widespread circulation. The rest of your comment makes sense with that adjustment.
The upshot of this is that if you ripped your own, or even changed the tags on the one you downloaded, you would be safe as long as your copy doesn't become widely circulated and catalogued by the bad guys.
So how evil is the RIAA? Sounds pretty fair to me.
Your ethical code leaves something to be desired (but I'm sure you'd have a negative opinion of mine as well). How is it fair that they have undermined the Constitution of the United States, destroyed the public domain, tricked, manipulated and bankrupted musicians, and lined their own pockets with money extorted from twelve-year-old girls in single-parent homes? How is it fair that they have destroyed their own market position by attacking the very fans that are buying their music? Oh, wait... that part is fair. They deserve their own self-inflicted destruction. I just wish it would come sooner, and without so much collateral damage.
No. The way to avoid screwing over your data when you accidentally delete stuff is to make backups of your data on a regular basis.
And learn not to type "rm" so hastily. Your LD_PRELOAD and "alias rm='rm -i'" bullshit tricks will come back and bite you in the ass as soon as you move to another system where they aren't there.
there's a real hunger and need for an ultra powerful email/usenet/scheduler/contact/pda manager.
Or not.
<Anecdote>I tried installing Thunderbird on one of my Linux boxes at home, for my wife to use while we wait for the replacement for the Windows box which has died (hardware failure). All seemed OK until she tried to send mail. Thunderbird simply hung while attempting to write her message to the "sent mail" folder. There were no error messages to be found anywhere. I went poking around in the directories, trying to figure out what was going on... couldn't make sense of a damned thing! There were all sort of weird files, most of them opaque binary ones (Berkeley DB files were quite common). Couldn't find a "sent mail" folder. Couldn't find a log file.
My best guess at this point is that it was attempting to lock a file. Home directories are NFS-mounted from an OpenBSD box with the nolock option, in order to pacify mutt (when writing to mboxes) and GNOME (which dies a horrible, agonizing, screaming death if gconf can't lock something).
Uninstalled Thunderbird, and set up PHP4 and squirrelmail on the mail server. Surfed to the squirrelmail login URL and bookmarked it for her in Firebird (haven't upgraded that box to Firefox yet). She hasn't complained yet.</Anecdote>
Firefox, ( otherwise very NICE ) takes about 15 seconds to start on my P4.
There's something seriously wrong with your computer, then. Perhaps your disk drives aren't using DMA mode? Or you've got insufficient RAM or a seriously heavy workload at the time? The start-up time for Firebird/Firefox/Mozilla on my Linux box at home is nowhere near that long. (And it's an Athlon XP 2000+.)
Granted, I'm not at home right now to actually time it, but I'm sure it's more on the order of 5 seconds for me -- after a cold boot with nothing cached in RAM yet. On the second or subsequent startup, with lots of files cached in RAM, I think it's more like 2 seconds.
Make ESC stop animated GIFs. Just fucking do it already! It's been, what, 5 years now? How long could it possibly take to fix thatbug?
When mousing over a link, with Javascript disabled, I should always, always, always be able to see the URL in the bottom of the window. ALWAYS. ALWAYS. I should never have to right-click on it, select "Copy Link Location", and paste it into a terminal window. (This is a Firefox regression, apparently. I haven't seen the URL displaying fail in Mozilla 1.4 or in Firebird.)
I should be able to configure the behavior of mailto: links to run a mail agent of my choice. (E.g.rxvt -e mutt %s.) And I shouldn't have to learn 3 or 4 new Mozilla-developed programming languages to do so. It should be on a damned menu somewhere.
I should be able to use an external text editor for writing very large comments in text boxes... you know, just like the one I'm typing in right now. Wouldn't it be divine if I could press something -- Alt-E, or whatever -- and have the text I'm writing get saved to a temp file, fire up rxvt -e vim %s on it, and then reload it from the temp file when the editor terminates? Oh well, at least it doesn't crash nearly as often as Netscape 4.x did when writing very large messages to web forums.
Microsoft has already moved on and is creating virtual machine run-times and a DirectX hardware-accelerated desktop. Linux is still trying to get a desktop off the ground with "cute" names like KDE and GNOME
Yeah, like "DirectX" isn't a cute name....
the snappy responsiveness OS X and Windows XP have
Windows XP isn't snappy. It spends at least half a second doing some stupid "fade in" instead of opening up a damned widget immediately. I guess this is supposed to make it less intimidating or something. That, and the Fisher Price colors. I don't get it, personally.
(because to Linux users, registries magically = bad because Windows happens to have something called the "registry")
No, registries are bad because they're opaque and complex. They fail, and when they fail, it's catastrophic. (AIX has a registry, too. It's called the ODM (Object Database M-something). I've been a professional AIX sysadmin, beginning with 3.1. I've seen ODM failures. They're not pretty.)
I don't want one on my real computer. I'll let Windows have one, because you need one for Windows, and you need Windows for gaming. But my real computer should work the way I want it to, and for me that's ~/.xsession, fvwm, rxvt, mutt and vim. I know most of you don't share my view, and that's fine. You can have your CORBAs and your "Let's save the state of every application so we can bring them all back up when the user logs in, and they'll never have to edit a file to customize their desktop!", and I hope you enjoy it. But when it breaks, I'll just sit there and look at you. I won't have to say "I told you so". But you'll know that I'm thinking it.
For some people. The rest of us shun printers as the foul abominations from hell that they truly are.
if I have to spend more than 2 minutes setting up a printer, I am outta there.
Man. I can see you've never set up printing on a real Unix system before. Setting up printers is the most horrendous, arcane, maniacally twisted experience you'll ever have on a Unix box, except for sendmail.
What do you mean, you can't handle editing/etc/printcap? It's just like writing a termcap entry! What? You've never written your own termcap before? Oh yeah, and you can't do filtering on remote queues. So you have to set up two queues for each printer: one to filter locally and redirect to the other queue, and the second one to actually deliver the data to the remote host. Why can't you filter and deliver in one queue? Well, uh... because you can't! It's been that way for 15 years, and we're not going to change it now![1]
And that's just the BSD side. The System V side lets you write your own shell script printer "drivers". Which is great, because everyone should have the luxury of getting to use stty for something other than making backspace work, and making @ no longer be the kill key (so you can type those newfangled Internet style e-mail addresses with @ signs in them instead of ! UUCP separators). And you can handle a 100-line shell script, can't you? Sure, piece of cake. Just follow the templates!
I haven't used CUPS yet. But I have absolutely no reason going in to expect it to be simple. If ESR got fooled into thinking it would be, then it's due to lack of experience in Unix printing. (He didn't even know what a JetDirect is. That's quite telling.)
Do you know what truly scared me when I read this article? Learning that CUPS allegedly broadcasts crap over the network every 30 seconds. Nothing more needs to be said.
[1] Yes, LPRng fixes it. But have you seen the size of the LPRng source tarball lately? It's bigger than the early Linux kernels. And yes, the complexity of the administration scales in proportion to the size of the source code, at least in this case. It's bad. Really, really bad. You get all the ickyness of the BSD printcap file, plus some brand new Next Generation ickyness to worry about on top of that!
Try reading the subject of the post you reply to. JWZ != ESR.
Oh, oh, oh! I can rant on this topic!
If you are writing something online -- whether it be an e-mail to a mailing list, or a post to a Usenet newsgroup, or a message to a web forum -- please, for the love of all that is holy, do not assume that the reader has read the subject line. Repeat your subject line in the body of the message, if necessary. A reader should be able to look at the body of your message and figure out what the hell you are talking about.
Yes, this also means you should quote some context from previous messages if it's required for your message to make sense. But don't just blindly copy the entire set of previous messages in that thread. Choose only the relevant sections.
Many of us who "grew up" in the old Usenet days, or on mailing lists, have learned not to read the Subject lines by default. They're usually either uninformative ("Subject: HELP NEEDED!!") or long past their expiration dates ("Subject: Re: qmail vs sendmail (was Re: why is my sendmail relaying spam?)" -- and the thread has even moved on past the qmail vs. sendmail tangent and is now a flame war about text editors).
That's why we don't read Subject lines, and that's why the grandparent message in this thread is misleading. The article is about ESR's rant at poor Free/Open Source software interface design, triggered by his experience with CUPS. Seeing a Slashdot comment with no explanatory details, about someone who writes software and complains about usability, and not knowing anything about XScreensaver off hand, it's natural to assume that a reference to ESR is being made.
I think the distinction you're looking for here is "drop-in replacement" versus "replacement". Y is clearly not a "drop-in replacement" for X, but it is a replacement.
One of the reasons OpenBSD tends to be more secure is because it ships with *almost* everything off. However, there's a solid 10+ default user accounts, 3-4 default services (sshd, sendmail, inetd/portmap)
To clarify: sendmail does not listen on port 25 of all network interfaces by default in OpenBSD. If you want to receive e-mail from the Internet using sendmail, you have to edit rc.conf to allow it.
*Sigh*... you can get precompiled binaries for Windows x86, Linux x86, Solaris SPARC and Solaris x86. Whoopie. Gee, how exciting.
Oh, and the downloads page actually confused me for a second. They have a giant section labelled "32-bit/64-bit for Windows/Linux/Solaris SPARC 32-bit for Solaris x86". I kept thinking, "Why do they offer a Linux SPARC version but not a Linux x86 version??" Then I figured out how to parse it. When they write "Linux", they mean "Linux x86", because there aren't any other Linux architectures on the whole planet, right? Sheesh.
Excuse me while I continue not using Java(R)(TM) on my HP-UX workstation at work, or on my OpenBSD x86 box at home (but Kaffe sort of runs on the latter).
Why can't Sun just release a god-damned TARBALL OF THE SOURCE CODE and say, "Here you go. Here's Java 1.5.0 release candidate 1. We'd like to have some help porting it to FreeBSD, OpenBSD, NetBSD, BeOS, Mac OS X, AIX, HP-UX, Tru64, IRIX, and so on. Patches are welcome. In fact, we've released the entire thing under the new BSD license (without advertising clause), so you can integrate it with your applications, include it on official Debian CDs, etc."?
Could you expand on this, please? Or provide a link?
Plus, it has non-trivial deployment issues
On the receiving MTA side, yes. The sending MTA side (DNS TXT record) is trivial.
and a set of drawbacks associated with it.
Such as...? And, doesn't everything have drawbacks? Eating food has the drawback that I might become poisoned, infected or obese, but it's the only way I know to avoid starvation.
I think the only thing that will resolve the spam issue is abandonment of SMTP as we know it, and an adoption of a new protocol
You may be right on this part.
that enforces traceability.
But I don't know whether you're right on that part.
Have you looked at DJB's Internet Mail 2000 yet? I don't know whether that will "solve spam", or whether it's even viable. But anyone who's interested in these issues should definitely look at it.
A similar system will need to exist for mail, that will require some sort of a registration and compliance to join the "mail provider" network, whatever that will be.
A centrally managed e-mail network? *shudder* No thanks!
If you, dead reader, happen to control the DNS for your own site
I'm not dead! 'Ere. He says he's not dead! Yes, he is! I'm not! He isn't? Well, he will be soon. He's very ill. I'm getting better! No you're not. You'll be stone dead in a moment.
The finger deamon can be rewritten slighly to return an affirmative if the user actually sent an email to the fingering domain.
You're not thinking this through. My mail server receives a message from someone at IP address A. It's addressed to me (greg@wooledge.org). It's got return envelope sender address S.
That's ALL the information available. I can look up the IP address and decide what to do based on what DNS says. I can look up the sender address and decide what to do based on what DNS says. I can attempt to contact the IP address that sent the message to me, or the MX server for the sender address, and decide what to do based on what those machines say in response to my questions.
How, then, does running a for-fuck's-sake finger daemon gain you a thing? If the spammer needs to have an ident daemon give back replies for whatever username they put in the sender address, then they'll run an ident daemon. It they need a finger daemon to give a response for whatever username they make up, they'll run a finger daemon. If they need... oh, I don't know, something just as lame and useless as finger, like, say, daytime... if they need a friggin' daytime server to give the right time of day, then they'll uncomment the daytime line in their inetd.conf file.
How does this fight spam? It doesn't!
The SMTP server can drop a line in the.plan
Dude, put down the crack pipe and step away from the keyboard.
Think about it, for $1/song and a typical 15-song album, that's $15.
I've always been a bit curious about this one. 15 songs is a lot higher than the track count on the average album in my collection. I've got albums with only one track on them: the first that comes to mind is Jethro Tull's "Thick as a Brick". One song, 42 minutes, one track on the CD. It's got a period of silence in the middle, which obviously had to be present so it could fit on double-sided media (vinyl, cassette), since it's from 1972. If I bought this album from iTunes, it would cost 99 cents, right? Or is it 10 dollars?
How about Dream Theater's "Six Degrees of Inner Turbulence"? It's a two-CD album, with 6 songs total. There are 5 songs on the first CD, and 1 on the second. The song on the second CD is divided into 8 tracks, with individual titles, but they're continuous. It's really one 40-minute song in 8 movements. What does this album cost? 6 dollars (1 dollar per song)? 10 dollars (10 dollars per album)? 13 dollars (1 dollar per track)? 20 dollars (10 dollars per CD)?
What about classical music? A CD with Beethoven's 9th symphony on it will probably have 4 tracks, one for each movement. Is it 1 dollar, or 4 dollars, or 10 dollars?
I think the online sweetspot wil be around $0.50.
As long as we're talking about 3-to-5-minute pop songs, I think that's a pretty good guess (though I'd believe $0.25 also). But not all music comes in individually wrapped slices.
Here's another question: a lot of the people who buy songs online will then want to burn them to CD-Rs for listening in their car, etc. In some countries such as Canada, people have paid a tax on blank CD-Rs, which is given (from what I understand) to the Canadian record companies according to whoever's most popular on the record charts at the time. So someone who buys legitimate online music in these countries gets hit by a double whammy. Is this going to change?
Well, i own about 25 cd's, all of which were bought, when you could get them for 15DM (about 7.50 EUR). I will not buy anymore untill i can get them for that price or lower again AND they stop behaving like the mafia.
Bravo! I don't know how many CDs I have, but it's a hell of a lot more than 25. Let me put it this way: I've got a wooden shelf-thing that allegedly holds 530 CDs. It was a Christmas present. So I spent the Sunday after Christmas -- not "Sunday Morning" or "Sunday Afternoon", but Sunday -- organizing my CDs alphabetically and putting them on the shelf. The shelf holds A-L and the first part of M. The rest had to go on some of the plastic CD holders (which hold on the order of 100-250 CDs each, and aren't as good, because they have standard-sized slots that don't hold double-size CDs, etc.).
In short, I've probably got somewhere between 1000 and 1500 CDs. This counts full albums, double albums, CD singles, etc. Some of these were purchased at full price, some were used, and some were free. Over my lifetime, I've been a hell of a good customer for the RIAA and their member companies, not to mention retailers.
I haven't bought a new CD since October. Used CDs only (or non-RIAA record companies, if I happen to find any worth buying from). No more money for the RIAA, until they start being a music company instead of a litigation company. I want my money to go to musicians, not to lawyers.
Have I stopped using peer-to-peer software? Take a guess.
I've been hearing quite a lot of this lately, from liberal and independent media sources. I don't think anyone is using the word capitalism correctly any more. Your particular use may have been sarcastic (it's hard to tell), but I'm reacting in a more general sense to all of the "anti-capitalist forces" out there who have been perverting a perfectly good word and turning it into a label for "the status quo", which they dislike. I can rant on this for a while, but I'll try to keep this one focused.
Let's start with the definition:
From WordNet (r) 2.0 [wn]:
capitalism
n : an economic system based on private ownership of capital
[syn: {capitalist economy}] [ant: {socialism}]
This also relies on the definition of capital which is basically wealth used to produce more wealth -- in other words investment.
So, what does peer-to-peer file sharing have to do with capitalism? As far as I can see, no such relationship exists. However, if there were any such relationship -- excuse me for a moment as I voyage out into the hypothetical -- it would be just the opposite of what your phrase implies.
What is copyright? Copyright is the right to copy something. That sounds really simple, doesn't it? In fact, that's too simplistic. What good is the right to copy? Isn't that just like the right to breathe?
See, what copyright really is, is the right to prevent other people from copying something. Now, that's something worth talking about! Your government (I assume that the Gentle Reader is from a nation which has a copyright statute of some sort) has decided that it would be a grand idea to let one person say to another: "You may not copy this, for I hold the power of copyright over it. I alone may copy it."
How does one go about enforcing such a thing? Back in the days when such laws were first created, it was a simple matter, because copying the items in question (books) required substantial resources -- a printing press, for one. So you could just keep an eye on everyone with a printing press and make sure they didn't produce copies of books that are copyrighted by someone else.
This was feasible because there weren't very many printing presses. A printing press is an investment, used to produce wealth (printed books, which have more value than their constituent ink and paper). So, a printing press is capital by the economic definition. The use of a (privately owned) printing press to make copies of a work is a capitalist activity.
How does copyright interact with this? Copyright says that, for a given work (book, phonorecord, etc.), a ban on copying shall exist until a specific amount of time has passed. That is, a governmentally enforced monopoly is granted, to one person or group of people, for the production and distribution of this work.
This is anti-capitalist. In a pure capitalist marketplace, everyone would be allowed to produce copies of the work, and distribute them in a competitive fashion. Because of copyright, a capitalist market for this work is not allowed to flourish. Let met say it again for those of you who are just skimming: Copyright is anti-capitalist.
So, what does that make peer-to-peer software? Let's take a specific work as an example: say, a copy of the studio recording of Yesterday by The Beatles. Despite the fact that half the band is deceased, this recording is still covered by copyright in most countries, and will continue to be so covered for another few generations. But how many copies of this recording can you find? There's one on a plastic disc in your local record store -- that's a government-approved one. iTunes probably has one in encumbered AAC format -- and that's also government-approved, at least if you don't do anything with it. There's probably one -- or a thousand -- in MP
Slashdot Mirror
I think everyone should listen to "American Woman" at least one, and I don't see a GOOD reason why they should have to pay for it. That money surely isn't going as an incentive for Hendrix to produce new music.
Jimi did a cover version of "American Woman"? Hold on a second... ssh, ssh, cd blah, fire up giFTcurs... search for American Woman... while that's running, try google... oh wait, am I going to get sued for this?
(The only versions of "American Woman" I've heard are the original by the Guess Who, and the Lenny Kravitz cover. If there's a Jimi cover, I want it, and eventually I'll find it... but so far I'm not seeing any evidence of one.)
Just about every P2P uses hashes to uniquely identify files on the network.
Correct. Except that many of the popular networks only hash the first 300k or so of the file. More mature networks hash the entire file.
So if you rip Nirvana's smells like teen spirit from the orginal cd into mp3 and then put it into your share directory for your P2P client it gets a unique hash assigned by your P2P client.
A hash isn't assigned. It's computed. It's a mathematical function that calculates a "checksum" of the input data. Same input, same hash. Different input, probably a different hash. The goal of a good hash function is to produce a different output for every input, insofar as that is possible. Changing one byte of the input file should produce a wildly different hash value. Hash collisions are to be avoided.
Now lets say a friend of yours does the exact same thing but his mp3 will have a different hash than yours because his P2P client creates a unique hash for his rip.
No. If his rip is byte-for-byte identical to yours, then it will have the same hash as yours. That's how hash functions work. The reason his rip gets a different hash is because it's a different rip. Reading CDDA is nondeterministic; you get single-bit errors in the data. And even if you had precisely the same raw audio data during the rip, you'd probably use a different encoder, or a different version of the encoder, or different encoder settings. Or you'd put different tags on the encoded file (even changing the title inside an ID3v2 or Vorbis tag from "Home By The Sea" to "Home by the Sea" will produce a different hash).
(Insert description of "swarming" here.)
Along comes RIAA or an agent for it. They catalog a list of all the known hashes for the song smell's like teen spirit [...]
I will grant that they may have accumulated a catalog (not a complete one!) of hashes of encodings of "Smells Like Teen Spirit" which happen to be in widespread circulation. The rest of your comment makes sense with that adjustment.
The upshot of this is that if you ripped your own, or even changed the tags on the one you downloaded, you would be safe as long as your copy doesn't become widely circulated and catalogued by the bad guys.
So how evil is the RIAA? Sounds pretty fair to me.
Your ethical code leaves something to be desired (but I'm sure you'd have a negative opinion of mine as well). How is it fair that they have undermined the Constitution of the United States, destroyed the public domain, tricked, manipulated and bankrupted musicians, and lined their own pockets with money extorted from twelve-year-old girls in single-parent homes? How is it fair that they have destroyed their own market position by attacking the very fans that are buying their music? Oh, wait... that part is fair. They deserve their own self-inflicted destruction. I just wish it would come sooner, and without so much collateral damage.
Why haven't I seen any comments about using Freenet for this yet? Where's Ian hiding today?
No. The way to avoid screwing over your data when you accidentally delete stuff is to make backups of your data on a regular basis.
And learn not to type "rm" so hastily. Your LD_PRELOAD and "alias rm='rm -i'" bullshit tricks will come back and bite you in the ass as soon as you move to another system where they aren't there.
there's a real hunger and need for an ultra powerful email/usenet/scheduler/contact/pda manager.
Or not.
<Anecdote>I tried installing Thunderbird on one of my Linux boxes at home, for my wife to use while we wait for the replacement for the Windows box which has died (hardware failure). All seemed OK until she tried to send mail. Thunderbird simply hung while attempting to write her message to the "sent mail" folder. There were no error messages to be found anywhere. I went poking around in the directories, trying to figure out what was going on... couldn't make sense of a damned thing! There were all sort of weird files, most of them opaque binary ones (Berkeley DB files were quite common). Couldn't find a "sent mail" folder. Couldn't find a log file.
My best guess at this point is that it was attempting to lock a file. Home directories are NFS-mounted from an OpenBSD box with the nolock option, in order to pacify mutt (when writing to mboxes) and GNOME (which dies a horrible, agonizing, screaming death if gconf can't lock something).
Uninstalled Thunderbird, and set up PHP4 and squirrelmail on the mail server. Surfed to the squirrelmail login URL and bookmarked it for her in Firebird (haven't upgraded that box to Firefox yet). She hasn't complained yet.</Anecdote>
Firefox, ( otherwise very NICE ) takes about 15 seconds to start on my P4.
There's something seriously wrong with your computer, then. Perhaps your disk drives aren't using DMA mode? Or you've got insufficient RAM or a seriously heavy workload at the time? The start-up time for Firebird/Firefox/Mozilla on my Linux box at home is nowhere near that long. (And it's an Athlon XP 2000+.)
Granted, I'm not at home right now to actually time it, but I'm sure it's more on the order of 5 seconds for me -- after a cold boot with nothing cached in RAM yet. On the second or subsequent startup, with lots of files cached in RAM, I think it's more like 2 seconds.
That's enough for this morning.
An XFree86 logo during the X server startup would get noticed!
Only after crashing/restarting X. How often do you do that? (Unless of course you're a laptop user....)
1 / Email the dev and tell him (gently) what's wrong in your opinion and what should be done. If he has the time, he'll fix it.
Not always.
Microsoft has already moved on and is creating virtual machine run-times and a DirectX hardware-accelerated desktop. Linux is still trying to get a desktop off the ground with "cute" names like KDE and GNOME
Yeah, like "DirectX" isn't a cute name....
the snappy responsiveness OS X and Windows XP have
Windows XP isn't snappy. It spends at least half a second doing some stupid "fade in" instead of opening up a damned widget immediately. I guess this is supposed to make it less intimidating or something. That, and the Fisher Price colors. I don't get it, personally.
(because to Linux users, registries magically = bad because Windows happens to have something called the "registry")
No, registries are bad because they're opaque and complex. They fail, and when they fail, it's catastrophic. (AIX has a registry, too. It's called the ODM (Object Database M-something). I've been a professional AIX sysadmin, beginning with 3.1. I've seen ODM failures. They're not pretty.)
I don't want one on my real computer. I'll let Windows have one, because you need one for Windows, and you need Windows for gaming. But my real computer should work the way I want it to, and for me that's ~/.xsession, fvwm, rxvt, mutt and vim. I know most of you don't share my view, and that's fine. You can have your CORBAs and your "Let's save the state of every application so we can bring them all back up when the user logs in, and they'll never have to edit a file to customize their desktop!", and I hope you enjoy it. But when it breaks, I'll just sit there and look at you. I won't have to say "I told you so". But you'll know that I'm thinking it.
printing is a basic fact of computer life
/etc/printcap? It's just like writing a termcap entry! What? You've never written your own termcap before? Oh yeah, and you can't do filtering on remote queues. So you have to set up two queues for each printer: one to filter locally and redirect to the other queue, and the second one to actually deliver the data to the remote host. Why can't you filter and deliver in one queue? Well, uh... because you can't! It's been that way for 15 years, and we're not going to change it now![1]
For some people. The rest of us shun printers as the foul abominations from hell that they truly are.
if I have to spend more than 2 minutes setting up a printer, I am outta there.
Man. I can see you've never set up printing on a real Unix system before. Setting up printers is the most horrendous, arcane, maniacally twisted experience you'll ever have on a Unix box, except for sendmail.
What do you mean, you can't handle editing
And that's just the BSD side. The System V side lets you write your own shell script printer "drivers". Which is great, because everyone should have the luxury of getting to use stty for something other than making backspace work, and making @ no longer be the kill key (so you can type those newfangled Internet style e-mail addresses with @ signs in them instead of ! UUCP separators). And you can handle a 100-line shell script, can't you? Sure, piece of cake. Just follow the templates!
I haven't used CUPS yet. But I have absolutely no reason going in to expect it to be simple. If ESR got fooled into thinking it would be, then it's due to lack of experience in Unix printing. (He didn't even know what a JetDirect is. That's quite telling.)
Do you know what truly scared me when I read this article? Learning that CUPS allegedly broadcasts crap over the network every 30 seconds. Nothing more needs to be said.
[1] Yes, LPRng fixes it. But have you seen the size of the LPRng source tarball lately? It's bigger than the early Linux kernels. And yes, the complexity of the administration scales in proportion to the size of the source code, at least in this case. It's bad. Really, really bad. You get all the ickyness of the BSD printcap file, plus some brand new Next Generation ickyness to worry about on top of that!
Try reading the subject of the post you reply to. JWZ != ESR.
Oh, oh, oh! I can rant on this topic!
If you are writing something online -- whether it be an e-mail to a mailing list, or a post to a Usenet newsgroup, or a message to a web forum -- please, for the love of all that is holy, do not assume that the reader has read the subject line. Repeat your subject line in the body of the message, if necessary. A reader should be able to look at the body of your message and figure out what the hell you are talking about.
Yes, this also means you should quote some context from previous messages if it's required for your message to make sense. But don't just blindly copy the entire set of previous messages in that thread. Choose only the relevant sections.
Many of us who "grew up" in the old Usenet days, or on mailing lists, have learned not to read the Subject lines by default. They're usually either uninformative ("Subject: HELP NEEDED!!") or long past their expiration dates ("Subject: Re: qmail vs sendmail (was Re: why is my sendmail relaying spam?)" -- and the thread has even moved on past the qmail vs. sendmail tangent and is now a flame war about text editors).
That's why we don't read Subject lines, and that's why the grandparent message in this thread is misleading. The article is about ESR's rant at poor Free/Open Source software interface design, triggered by his experience with CUPS. Seeing a Slashdot comment with no explanatory details, about someone who writes software and complains about usability, and not knowing anything about XScreensaver off hand, it's natural to assume that a reference to ESR is being made.
I think the distinction you're looking for here is "drop-in replacement" versus "replacement". Y is clearly not a "drop-in replacement" for X, but it is a replacement.
One of the reasons OpenBSD tends to be more secure is because it ships with *almost* everything off. However, there's a solid 10+ default user accounts, 3-4 default services (sshd, sendmail, inetd/portmap)
To clarify: sendmail does not listen on port 25 of all network interfaces by default in OpenBSD. If you want to receive e-mail from the Internet using sendmail, you have to edit rc.conf to allow it.
*Sigh*... you can get precompiled binaries for Windows x86, Linux x86, Solaris SPARC and Solaris x86. Whoopie. Gee, how exciting.
Oh, and the downloads page actually confused me for a second. They have a giant section labelled "32-bit/64-bit for Windows/Linux/Solaris SPARC
32-bit for Solaris x86". I kept thinking, "Why do they offer a Linux SPARC version but not a Linux x86 version??" Then I figured out how to parse it. When they write "Linux", they mean "Linux x86", because there aren't any other Linux architectures on the whole planet, right? Sheesh.
Excuse me while I continue not using Java(R)(TM) on my HP-UX workstation at work, or on my OpenBSD x86 box at home (but Kaffe sort of runs on the latter).
Why can't Sun just release a god-damned TARBALL OF THE SOURCE CODE and say, "Here you go. Here's Java 1.5.0 release candidate 1. We'd like to have some help porting it to FreeBSD, OpenBSD, NetBSD, BeOS, Mac OS X, AIX, HP-UX, Tru64, IRIX, and so on. Patches are welcome. In fact, we've released the entire thing under the new BSD license (without advertising clause), so you can integrate it with your applications, include it on official Debian CDs, etc."?
*Sigh*
why not just send the email header only - and require the originating server to hold the email content?
You're half-way to re-inventing Internet Mail 2000, then.
SPF is easy to defeat.
Could you expand on this, please? Or provide a link?
Plus, it has non-trivial deployment issues
On the receiving MTA side, yes. The sending MTA side (DNS TXT record) is trivial.
and a set of drawbacks associated with it.
Such as...? And, doesn't everything have drawbacks? Eating food has the drawback that I might become poisoned, infected or obese, but it's the only way I know to avoid starvation.
I think the only thing that will resolve the spam issue is abandonment of SMTP as we know it, and an adoption of a new protocol
You may be right on this part.
that enforces traceability.
But I don't know whether you're right on that part.
Have you looked at DJB's Internet Mail 2000 yet? I don't know whether that will "solve spam", or whether it's even viable. But anyone who's interested in these issues should definitely look at it.
A similar system will need to exist for mail, that will require some sort of a registration and compliance to join the "mail provider" network, whatever that will be.
A centrally managed e-mail network? *shudder* No thanks!
If you, dead reader, happen to control the DNS for your own site
...
I'm not dead!
'Ere. He says he's not dead!
Yes, he is!
I'm not!
He isn't?
Well, he will be soon. He's very ill.
I'm getting better!
No you're not. You'll be stone dead in a moment.
The finger deamon can be rewritten slighly to return an affirmative if the user actually sent an email to the fingering domain.
.plan
You're not thinking this through. My mail server receives a message from someone at IP address A. It's addressed to me (greg@wooledge.org). It's got return envelope sender address S.
That's ALL the information available. I can look up the IP address and decide what to do based on what DNS says. I can look up the sender address and decide what to do based on what DNS says. I can attempt to contact the IP address that sent the message to me, or the MX server for the sender address, and decide what to do based on what those machines say in response to my questions.
How, then, does running a for-fuck's-sake finger daemon gain you a thing? If the spammer needs to have an ident daemon give back replies for whatever username they put in the sender address, then they'll run an ident daemon. It they need a finger daemon to give a response for whatever username they make up, they'll run a finger daemon. If they need... oh, I don't know, something just as lame and useless as finger, like, say, daytime... if they need a friggin' daytime server to give the right time of day, then they'll uncomment the daytime line in their inetd.conf file.
How does this fight spam? It doesn't!
The SMTP server can drop a line in the
Dude, put down the crack pipe and step away from the keyboard.
Richland Middle School
HEAD / HTTP/1.0
HTTP/1.1 200 OK
Server: Microsoft-IIS/5.0
MicrosoftOfficeWebServer: 5.0_Pub
Content-Location: http://10.0.1.149/Default.html
Date: Thu, 08 Jan 2004 14:08:10 GMT
Content-Type: text/html
Accept-Ranges: bytes
Last-Modified: Mon, 20 May 2002 20:40:17 GMT
ETag: "80cecc8c3e0c21:a5d"
Content-Length: 801
Think about it, for $1/song and a typical 15-song album, that's $15.
I've always been a bit curious about this one. 15 songs is a lot higher than the track count on the average album in my collection. I've got albums with only one track on them: the first that comes to mind is Jethro Tull's "Thick as a Brick". One song, 42 minutes, one track on the CD. It's got a period of silence in the middle, which obviously had to be present so it could fit on double-sided media (vinyl, cassette), since it's from 1972. If I bought this album from iTunes, it would cost 99 cents, right? Or is it 10 dollars?
How about Dream Theater's "Six Degrees of Inner Turbulence"? It's a two-CD album, with 6 songs total. There are 5 songs on the first CD, and 1 on the second. The song on the second CD is divided into 8 tracks, with individual titles, but they're continuous. It's really one 40-minute song in 8 movements. What does this album cost? 6 dollars (1 dollar per song)? 10 dollars (10 dollars per album)? 13 dollars (1 dollar per track)? 20 dollars (10 dollars per CD)?
What about classical music? A CD with Beethoven's 9th symphony on it will probably have 4 tracks, one for each movement. Is it 1 dollar, or 4 dollars, or 10 dollars?
I think the online sweetspot wil be around $0.50.
As long as we're talking about 3-to-5-minute pop songs, I think that's a pretty good guess (though I'd believe $0.25 also). But not all music comes in individually wrapped slices.
Here's another question: a lot of the people who buy songs online will then want to burn them to CD-Rs for listening in their car, etc. In some countries such as Canada, people have paid a tax on blank CD-Rs, which is given (from what I understand) to the Canadian record companies according to whoever's most popular on the record charts at the time. So someone who buys legitimate online music in these countries gets hit by a double whammy. Is this going to change?
Well, i own about 25 cd's, all of which were bought, when you could get them for 15DM (about 7.50 EUR). I will not buy anymore untill i can get them for that price or lower again AND they stop behaving like the mafia.
Bravo! I don't know how many CDs I have, but it's a hell of a lot more than 25. Let me put it this way: I've got a wooden shelf-thing that allegedly holds 530 CDs. It was a Christmas present. So I spent the Sunday after Christmas -- not "Sunday Morning" or "Sunday Afternoon", but Sunday -- organizing my CDs alphabetically and putting them on the shelf. The shelf holds A-L and the first part of M. The rest had to go on some of the plastic CD holders (which hold on the order of 100-250 CDs each, and aren't as good, because they have standard-sized slots that don't hold double-size CDs, etc.).
In short, I've probably got somewhere between 1000 and 1500 CDs. This counts full albums, double albums, CD singles, etc. Some of these were purchased at full price, some were used, and some were free. Over my lifetime, I've been a hell of a good customer for the RIAA and their member companies, not to mention retailers.
I haven't bought a new CD since October. Used CDs only (or non-RIAA record companies, if I happen to find any worth buying from). No more money for the RIAA, until they start being a music company instead of a litigation company. I want my money to go to musicians, not to lawyers.
Have I stopped using peer-to-peer software? Take a guess.
the anti-capitalist P2Pers
Newspeak! Double-plus un-good!
I've been hearing quite a lot of this lately, from liberal and independent media sources. I don't think anyone is using the word capitalism correctly any more. Your particular use may have been sarcastic (it's hard to tell), but I'm reacting in a more general sense to all of the "anti-capitalist forces" out there who have been perverting a perfectly good word and turning it into a label for "the status quo", which they dislike. I can rant on this for a while, but I'll try to keep this one focused.
Let's start with the definition:
From WordNet (r) 2.0 [wn]:
capitalism
n : an economic system based on private ownership of capital
[syn: {capitalist economy}] [ant: {socialism}]
This also relies on the definition of capital which is basically wealth used to produce more wealth -- in other words investment.
So, what does peer-to-peer file sharing have to do with capitalism? As far as I can see, no such relationship exists. However, if there were any such relationship -- excuse me for a moment as I voyage out into the hypothetical -- it would be just the opposite of what your phrase implies.
What is copyright? Copyright is the right to copy something. That sounds really simple, doesn't it? In fact, that's too simplistic. What good is the right to copy? Isn't that just like the right to breathe?
See, what copyright really is, is the right to prevent other people from copying something. Now, that's something worth talking about! Your government (I assume that the Gentle Reader is from a nation which has a copyright statute of some sort) has decided that it would be a grand idea to let one person say to another: "You may not copy this, for I hold the power of copyright over it. I alone may copy it."
How does one go about enforcing such a thing? Back in the days when such laws were first created, it was a simple matter, because copying the items in question (books) required substantial resources -- a printing press, for one. So you could just keep an eye on everyone with a printing press and make sure they didn't produce copies of books that are copyrighted by someone else.
This was feasible because there weren't very many printing presses. A printing press is an investment, used to produce wealth (printed books, which have more value than their constituent ink and paper). So, a printing press is capital by the economic definition. The use of a (privately owned) printing press to make copies of a work is a capitalist activity.
How does copyright interact with this? Copyright says that, for a given work (book, phonorecord, etc.), a ban on copying shall exist until a specific amount of time has passed. That is, a governmentally enforced monopoly is granted, to one person or group of people, for the production and distribution of this work.
This is anti-capitalist. In a pure capitalist marketplace, everyone would be allowed to produce copies of the work, and distribute them in a competitive fashion. Because of copyright, a capitalist market for this work is not allowed to flourish. Let met say it again for those of you who are just skimming: Copyright is anti-capitalist.
So, what does that make peer-to-peer software? Let's take a specific work as an example: say, a copy of the studio recording of Yesterday by The Beatles. Despite the fact that half the band is deceased, this recording is still covered by copyright in most countries, and will continue to be so covered for another few generations. But how many copies of this recording can you find? There's one on a plastic disc in your local record store -- that's a government-approved one. iTunes probably has one in encumbered AAC format -- and that's also government-approved, at least if you don't do anything with it. There's probably one -- or a thousand -- in MP