There's a VERY simple program, called tcpbridge (do a websearch on tcpbridge.c), which does simple forwarding of TCP/IP sockets. It has a few problems, though; for example, under certain circumstances (such as a socket unexpectedly closing or whatever), it goes into a CPU-thrashing state. Also, it has no logging facilities, and any connection will look like it came from the bridge box (this actually caused me some grief, when someone caused some general mischeif and I had no idea where it came from). Something at the protocol/packet level would be much nicer than such an inelegant solution.
Now, I have seen references to IPchains being used for forwarding, but they didn't go into detail and so I don't know if they're just spread misconceptions (which I am guilty of perpetrating, myself, as is anyone else who reads too much into the description of IPchains).
I'm with you... I recently realized that I was totally dependent on caffeine (based on the fact that I drank a pint of Dew in the morning to wake up, had to drink more Dew and Coke and Dr. Pepper constantly at work to stay awake, and was completely incapable of concentrating on anything, and wasn't sleeping at all well) and went cold turkey for a week. I had massive withdrawal headaches, but it was worth it. Now I drink caffeine again, but only in moderation (1-2 sodas a day, tops) and am sleeping much better and able to actually get work done.
Caffeine poisoning just isn't worth it in the long run. --- "'Is not a quine' is not a quine" is a quine.
<joke>What I heard was that they sold a bunch, but then went out of business when nobody needed more for another 20 years...</joke> --- "'Is not a quine' is not a quine" is a quine.
Ahh, that does me a lot of good now, after I've completely switched over to Debian.;)
Seriously, though, my point was that the swap partition issue is a bit annoying, and hoped to also convey that it's part of why new users have a hard time adopting Linux. If RedHat were to just use swapfiles instead, that would remove a LOT of the confusion, and make it much more flexible later on as well. One shouldn't have to go into expert mode to make it easier; if anything, you should have to use Expert Mode to use a swap partition.
Why does Redhat *require* a swap partition? Yes, if you have 4 megs of RAM you need one, and if you have 16 or less you'd probably want one for programs which are resource-intensive *cough*Netscape*cough*KDE*cough*. However, my system has 128 megs of RAM, which is becoming increasingly more common.
Debian doesn't require a swap partition. Slackware doesn't require a swap partition. Sure, they recommend one, but they don't absolutely require one to install. Can't Redhat use a swap file? Yes, it's a bit slower, but when you're swapping, chances are a few additional microseconds of latency in paging are the least of your performance problems.
The main reason that Redhat's insistence annoys me, aside from the fact that it makes new users squeamish, is that when I was upgrading to Redhat from Slackware, I had no decent way to repartition my drive to make the minimal 8 meg swap partition, and so I ended up using - get this - a SCSI zip drive, for lack of a better means of satiating Redhat's installer.
I got lucky in that my current employer (whom I've been working for for the last month) saw my plaintext resume on Headhunter.net and was impressed enough to contact me. It also helps that this is a mostly UNIX house and so they were more than happy to view my complete resume (which I keep in HTML format) without the assistance of Microsoft products.:) --- "'Is not a quine' is not a quine" is a quine.
There's somewhat of a scientific flaw in the 'bits as atoms' argument they're using. They're assuming that when I send a bit from my computer to a server somewhere else, it's the exact same electron(s) which carried that bit that end up in the server somewhere else. This is a common misconception, but the electrons travel literally at a snail's pace; it's the *signal* which goes at relativistic speeds.
As I commented previously, I'm split on this issue. It's beneficial to me, being a resident of Virginia, to be able to sue someone for libelling/slandering/spamming/etc. me since their bits eventually reach my home computer in Virginia (and since I use a personal proxying/masquerading firewall to have all my and my roommates' systems online, it could be argued that a Virginia server is used, without even having to piss over the difference between server and workstation), but I don't like this precedent as it applies to the Internet in terms of long-arm jurisdiction and the like.
Also, would defamation online be constituted as libel or slander? It's "spoken" in a way, but it's also "printed" in a way (libel relates to untrue printed information, slander is to spoken information), but in reality it's neither. But if 'printing' is constituted by the act of publishing, well, technically anything you say on Usenet is eventually 'published' (on the web) thanks to Deja et al.
Complicated issue, not at all helped by the fact that our government and legal system are still set up in terms of late 1700s/early 1800s technology.
Perhaps the best way to look at this whole argument is how it would be on a national television broadcast. If someone in Texas is on CNN and broadcasts his message about how someone in Virginia is a stupidhead because he thinks such-and-such about the purported JFK conspiracy, where would a slander case be held? (Since in this case it's definitely speech at issue, it'd definitely be slander, not libel.) Woudln't it be a federal case? Take it one step further; if someone in Iraq slanders the American president and military and broadcasts it (and some of Saddam Hussein's speeches were, I might add, rebroadcast in the USA as part of the US media's anti-propaganda propaganda) and Clinton decided to sue that certain Iraqi for slander, where would a case, if any, exist?
Um, dude, the four corners area is Arizona, Colorado, Utah, and New Mexico.:) (I'm a former New Mexican, but as of recently a resident of Virginia, oddly enough. The 'four corners area' is where the Hanta virus outbreaks keep happening. Not the whole damned state.)
I'm split on this issue, given that I'm now a resident of Virginia. I'm unsure as to support this ruling (it's got both its good and bad points; good because it leaves lots of loopholes to get around shitty rulings, but bad because long-arm jurisdiction sucks, and puts in more shitty rulings to need to get around), but at least it means now I can threaten a lawsuit to AOL spammers with legal grounds to stand on.:) (I'm not sue-happy, but I'm sick and tired of spam. blah.) --- "'Is not a quine' is not a quine" is a quine.
ExploreZip is an executable. Melissa was a macro. This worm will work on a system that doesn't have MS Office; Melissa won't.
Not to be overly pedantic, but isn't an executable file technically a macro for execution on an x86 Von Neumann virtual machine? It just has less levels of indirection.:) (I'm using 'virtual machine' in the pure CS sense, i.e. anything which can take in instructions and perform actions based on them.)
All in all this is a very good article. I have one problem with it, though, based on a silly little thing which doesn't have anything to do with the topic of the article:
Last summer also witnessed the debut of "Back Orifice," which grants unauthorized users remote access to machines running either Windows 95 or Windows 98, the operating system of choice for most home computers.
Last I checked it wasn't the OS of choice, but of taciturn consent. Didn't work for Louis XIV... Ahwell. Otherwise I'd love to see more articles like this.
One big partition makes it difficult to try different distributions of Linux, and in the case of some distributions (such as Slackware) it can be downright difficult to even upgrade if everything's on one big partition. By keeping/home and/usr/local on separate partitions, you can nuke the rest of the system and not have to worry about going through the pain of recovering user files, reinstalling non-packaged programs, etc. (though I generally just keep/home since changing distributions gives me a good reason to seek out later versions of/usr/local programs anyway). --- "'Is not a quine' is not a quine" is a quine.
I wonder if people will go to the theatre for SW1 and, forgetting where they are, leave in disgust when they don't show the SW1 trailer before the film.:) --- "'Is not a quine' is not a quine" is a quine.
Well, these days, 'class A' means/8, 'class B' means/16, and 'class C' means/24. It's a lot more convenient, IMO, to still refer to those common denominations as such. --- "'Is not a quine' is not a quine" is a quine.
Well, it was my impression that the kernel itself was going to incorporate ALSA. Generally individual projects which have been included in the kernel don't make much mention of it if any (see the joystick driver homepage, the bttv (one of the Video for Linux drivers) homepage, etc. for good examples of this. That is, I'd be more inclined to use "what I've heard from people who read linux-kernel" as an indication that it is than to use the fact it's not mentioned on one particular page that it isn't. --- "'Is not a quine' is not a quine" is a quine.
Well, the first and most important one I can think of is ALSA replacing OSS as the soundsystem. FINALLY some decent, modular multimedia and MIDI functionality in Linux!:) --- "'Is not a quine' is not a quine" is a quine.
Re:End User Apps - cooperation on file formats
on
Pair of KDE Stories
·
· Score: 1
I totally agree with you here. The only two differences, really, between Gnome and KDE are the communication mechanism and the enforced application toolkit. The latter is why I prefer Gnome; although GTK isn't nearly as mature or stable as Qt, GTK *does* have bindings for many languages (C, C++, Python, PERL if I'm not mistaken, and several others, and its design allows it to be bound to practically any language), whereas Qt is C++, end of story. It's easy to wrap C in basically any language, but C++ has a lot of issues (such as the vtable, name-mangling, and other implementation-specific things) to deal with.
Another advantage of Gnome is that it uses CORBA, so it could theoretically be ported to systems which don't use X. I don't know what inter-client communication mechanism KDE uses; could someone please clarify that? I'm too lazy to find it on kde.org.:) Of course, this is just, as you said, another interface like a file. In the case of TCP/IP sockets, the interface doesn't even need to appear different except in the way the file is opened, at least in UNIX. Dunno how it is in Windows; I'd imagine that Windows, which is based on interfaces specific to the hardware instead of the nice consistent general file interface for everything, has its own set of APIs which would need to be wrapped.
Whatever. Anyway. Cooperation between the various usability projects is vital. --- "'Is not a quine' is not a quine" is a quine.
Even moreso, it doesn't even matter if people use the same suite (KDE or Gnome) as long as they use open file formats and can be made compatible with one another. Cooperation, and not competition, is VERY important to both KDE and Gnome. Myself, I don't care much for either, and I don't want to have to run KOffice under KDE to read someone's document any more than anyone wants to have to run MS Office under Win'98.
UNIX is great in that file formats have always been *open* and relatively easy to deal with. I mean, there's several dozen commonly-used mail programs out there, and practically all of them use the Sendmail/var/spool/mail fileformat to store everything. PINE, ELM, Mutt, mail, etc. can all read and write to my various mail folders without a problem, and so switching programs isn't a daunting task. (Wish the same could be said for Netscape Communicator, but I've not used Netscape for reading mail ever since Netscape 3.0 consistently corrupted all my messages and did other unpleasant things.)
I mean, okay, I don't need to run KDE or Gnome to use KDE or Gnome applications, but that's not the point. The point I'm trying to make is that just because a program is available now and is free doesn't mean one should be locked into that single program. It'd suck if I were stuck using PINE forever and ever, for example. (I want to try Mutt someday.)
That said, perhaps some work should be done in deciding/drafting formal document format specs. We already have a perfectly good word processing format (namely TeX) and many word processors which can use it (such as LyX/KLyX, and even some lesser-known Windows word processors such as Scientific Workplace)... but what about spreadsheets? Are the formats of GNUmeric and (whatever KOffice's spreadsheet program is) compatible? What about StarOffice? Whatever happened to 'sc'?
What about groupware programs? It's hard enough to deal with KDE/Gnome flamewars on Slashdot. What about in the workplace where the manager wants everyone to use a KDE-based groupware app but all the Gnome zealots want to use a Gnome-based one, and then all the desktop-agnostics (which I currently am) want to just use one which works under fvwm2?
(Yes, I know there's better WMs than fvwm2. But I have it setup to work nicely with my Datahand and I don't feel like learning how to configure another WM right now.)
I've somehow managed to avoid getting spam for a while. I use the RBL religiously, I don't put my email address online promiscuously in a way that harvesters catch, and I'm constantly adding things to my.procmail/rc.spam filter. Yaaay. Still the occasional piece of spam reaches me, at which point I get horribly medieval on the spammer, the spammer's upstream, all the relays used, InterNIC, etc. I don't even do it with a script, as I take some sort of insane pleasure in dealing with it by hand. --- "'Is not a quine' is not a quine" is a quine.
Would an IBM 3270 terminal qualify for this? It had a pretty funky protocol for doing that sort of thing with respects to its window interface. --- "'Is not a quine' is not a quine" is a quine.
Slashdot Mirror
BeDope is satire, dude.
---
"'Is not a quine' is not a quine" is a quine.
There's a VERY simple program, called tcpbridge (do a websearch on tcpbridge.c), which does simple forwarding of TCP/IP sockets. It has a few problems, though; for example, under certain circumstances (such as a socket unexpectedly closing or whatever), it goes into a CPU-thrashing state. Also, it has no logging facilities, and any connection will look like it came from the bridge box (this actually caused me some grief, when someone caused some general mischeif and I had no idea where it came from). Something at the protocol/packet level would be much nicer than such an inelegant solution.
Now, I have seen references to IPchains being used for forwarding, but they didn't go into detail and so I don't know if they're just spread misconceptions (which I am guilty of perpetrating, myself, as is anyone else who reads too much into the description of IPchains).
---
"'Is not a quine' is not a quine" is a quine.
I'm with you... I recently realized that I was totally dependent on caffeine (based on the fact that I drank a pint of Dew in the morning to wake up, had to drink more Dew and Coke and Dr. Pepper constantly at work to stay awake, and was completely incapable of concentrating on anything, and wasn't sleeping at all well) and went cold turkey for a week. I had massive withdrawal headaches, but it was worth it. Now I drink caffeine again, but only in moderation (1-2 sodas a day, tops) and am sleeping much better and able to actually get work done.
Caffeine poisoning just isn't worth it in the long run.
---
"'Is not a quine' is not a quine" is a quine.
<joke>What I heard was that they sold a bunch, but then went out of business when nobody needed more for another 20 years...</joke>
---
"'Is not a quine' is not a quine" is a quine.
I dunno, verbing words is fun. Though, at time, verbing weirds language... :)
---
"'Is not a quine' is not a quine" is a quine.
So was he.
---
"'Is not a quine' is not a quine" is a quine.
Ahh, that does me a lot of good now, after I've completely switched over to Debian. ;)
Seriously, though, my point was that the swap partition issue is a bit annoying, and hoped to also convey that it's part of why new users have a hard time adopting Linux. If RedHat were to just use swapfiles instead, that would remove a LOT of the confusion, and make it much more flexible later on as well. One shouldn't have to go into expert mode to make it easier; if anything, you should have to use Expert Mode to use a swap partition.
---
"'Is not a quine' is not a quine" is a quine.
Why does Redhat *require* a swap partition? Yes, if you have 4 megs of RAM you need one, and if you have 16 or less you'd probably want one for programs which are resource-intensive *cough*Netscape*cough*KDE*cough*. However, my system has 128 megs of RAM, which is becoming increasingly more common.
Debian doesn't require a swap partition. Slackware doesn't require a swap partition. Sure, they recommend one, but they don't absolutely require one to install. Can't Redhat use a swap file? Yes, it's a bit slower, but when you're swapping, chances are a few additional microseconds of latency in paging are the least of your performance problems.
The main reason that Redhat's insistence annoys me, aside from the fact that it makes new users squeamish, is that when I was upgrading to Redhat from Slackware, I had no decent way to repartition my drive to make the minimal 8 meg swap partition, and so I ended up using - get this - a SCSI zip drive, for lack of a better means of satiating Redhat's installer.
Bloody annoying.
---
"'Is not a quine' is not a quine" is a quine.
I got lucky in that my current employer (whom I've been working for for the last month) saw my plaintext resume on Headhunter.net and was impressed enough to contact me. It also helps that this is a mostly UNIX house and so they were more than happy to view my complete resume (which I keep in HTML format) without the assistance of Microsoft products. :)
---
"'Is not a quine' is not a quine" is a quine.
Bet you really love my .signature, then. :)
---
"'Is not a quine' is not a quine" is a quine.
RichardWare, or 'DickWare' for short? I don't think so. :)
---
"'Is not a quine' is not a quine" is a quine.
There's somewhat of a scientific flaw in the 'bits as atoms' argument they're using. They're assuming that when I send a bit from my computer to a server somewhere else, it's the exact same electron(s) which carried that bit that end up in the server somewhere else. This is a common misconception, but the electrons travel literally at a snail's pace; it's the *signal* which goes at relativistic speeds.
As I commented previously, I'm split on this issue. It's beneficial to me, being a resident of Virginia, to be able to sue someone for libelling/slandering/spamming/etc. me since their bits eventually reach my home computer in Virginia (and since I use a personal proxying/masquerading firewall to have all my and my roommates' systems online, it could be argued that a Virginia server is used, without even having to piss over the difference between server and workstation), but I don't like this precedent as it applies to the Internet in terms of long-arm jurisdiction and the like.
Also, would defamation online be constituted as libel or slander? It's "spoken" in a way, but it's also "printed" in a way (libel relates to untrue printed information, slander is to spoken information), but in reality it's neither. But if 'printing' is constituted by the act of publishing, well, technically anything you say on Usenet is eventually 'published' (on the web) thanks to Deja et al.
Complicated issue, not at all helped by the fact that our government and legal system are still set up in terms of late 1700s/early 1800s technology.
Perhaps the best way to look at this whole argument is how it would be on a national television broadcast. If someone in Texas is on CNN and broadcasts his message about how someone in Virginia is a stupidhead because he thinks such-and-such about the purported JFK conspiracy, where would a slander case be held? (Since in this case it's definitely speech at issue, it'd definitely be slander, not libel.) Woudln't it be a federal case? Take it one step further; if someone in Iraq slanders the American president and military and broadcasts it (and some of Saddam Hussein's speeches were, I might add, rebroadcast in the USA as part of the US media's anti-propaganda propaganda) and Clinton decided to sue that certain Iraqi for slander, where would a case, if any, exist?
Many things to consider.
---
"'Is not a quine' is not a quine" is a quine.
I'm split on this issue, given that I'm now a resident of Virginia. I'm unsure as to support this ruling (it's got both its good and bad points; good because it leaves lots of loopholes to get around shitty rulings, but bad because long-arm jurisdiction sucks, and puts in more shitty rulings to need to get around), but at least it means now I can threaten a lawsuit to AOL spammers with legal grounds to stand on. :) (I'm not sue-happy, but I'm sick and tired of spam. blah.)
---
"'Is not a quine' is not a quine" is a quine.
ExploreZip is an executable. Melissa was a macro. This worm will work on a system that doesn't have MS Office; Melissa won't.
Not to be overly pedantic, but isn't an executable file technically a macro for execution on an x86 Von Neumann virtual machine? It just has less levels of indirection. :) (I'm using 'virtual machine' in the pure CS sense, i.e. anything which can take in instructions and perform actions based on them.)
---
"'Is not a quine' is not a quine" is a quine.
All in all this is a very good article. I have one problem with it, though, based on a silly little thing which doesn't have anything to do with the topic of the article:
Last summer also witnessed the debut of "Back Orifice," which grants unauthorized users remote access to machines running either Windows 95 or Windows 98, the operating system of choice for most home computers.
Last I checked it wasn't the OS of choice, but of taciturn consent. Didn't work for Louis XIV... Ahwell. Otherwise I'd love to see more articles like this.
---
"'Is not a quine' is not a quine" is a quine.
One big partition makes it difficult to try different distributions of Linux, and in the case of some distributions (such as Slackware) it can be downright difficult to even upgrade if everything's on one big partition. By keeping /home and /usr/local on separate partitions, you can nuke the rest of the system and not have to worry about going through the pain of recovering user files, reinstalling non-packaged programs, etc. (though I generally just keep /home since changing distributions gives me a good reason to seek out later versions of /usr/local programs anyway).
---
"'Is not a quine' is not a quine" is a quine.
I wonder if people will go to the theatre for SW1 and, forgetting where they are, leave in disgust when they don't show the SW1 trailer before the film. :)
---
"'Is not a quine' is not a quine" is a quine.
Well, these days, 'class A' means /8, 'class B' means /16, and 'class C' means /24. It's a lot more convenient, IMO, to still refer to those common denominations as such.
---
"'Is not a quine' is not a quine" is a quine.
Well, it was my impression that the kernel itself was going to incorporate ALSA. Generally individual projects which have been included in the kernel don't make much mention of it if any (see the joystick driver homepage, the bttv (one of the Video for Linux drivers) homepage, etc. for good examples of this. That is, I'd be more inclined to use "what I've heard from people who read linux-kernel" as an indication that it is than to use the fact it's not mentioned on one particular page that it isn't.
---
"'Is not a quine' is not a quine" is a quine.
Well, the first and most important one I can think of is ALSA replacing OSS as the soundsystem. FINALLY some decent, modular multimedia and MIDI functionality in Linux! :)
---
"'Is not a quine' is not a quine" is a quine.
Another advantage of Gnome is that it uses CORBA, so it could theoretically be ported to systems which don't use X. I don't know what inter-client communication mechanism KDE uses; could someone please clarify that? I'm too lazy to find it on kde.org. :) Of course, this is just, as you said, another interface like a file. In the case of TCP/IP sockets, the interface doesn't even need to appear different except in the way the file is opened, at least in UNIX. Dunno how it is in Windows; I'd imagine that Windows, which is based on interfaces specific to the hardware instead of the nice consistent general file interface for everything, has its own set of APIs which would need to be wrapped.
Whatever. Anyway. Cooperation between the various usability projects is vital.
---
"'Is not a quine' is not a quine" is a quine.
Even moreso, it doesn't even matter if people use the same suite (KDE or Gnome) as long as they use open file formats and can be made compatible with one another. Cooperation, and not competition, is VERY important to both KDE and Gnome. Myself, I don't care much for either, and I don't want to have to run KOffice under KDE to read someone's document any more than anyone wants to have to run MS Office under Win'98.
UNIX is great in that file formats have always been *open* and relatively easy to deal with. I mean, there's several dozen commonly-used mail programs out there, and practically all of them use the Sendmail /var/spool/mail fileformat to store everything. PINE, ELM, Mutt, mail, etc. can all read and write to my various mail folders without a problem, and so switching programs isn't a daunting task. (Wish the same could be said for Netscape Communicator, but I've not used Netscape for reading mail ever since Netscape 3.0 consistently corrupted all my messages and did other unpleasant things.)
I mean, okay, I don't need to run KDE or Gnome to use KDE or Gnome applications, but that's not the point. The point I'm trying to make is that just because a program is available now and is free doesn't mean one should be locked into that single program. It'd suck if I were stuck using PINE forever and ever, for example. (I want to try Mutt someday.)
That said, perhaps some work should be done in deciding/drafting formal document format specs. We already have a perfectly good word processing format (namely TeX) and many word processors which can use it (such as LyX/KLyX, and even some lesser-known Windows word processors such as Scientific Workplace)... but what about spreadsheets? Are the formats of GNUmeric and (whatever KOffice's spreadsheet program is) compatible? What about StarOffice? Whatever happened to 'sc'?
What about groupware programs? It's hard enough to deal with KDE/Gnome flamewars on Slashdot. What about in the workplace where the manager wants everyone to use a KDE-based groupware app but all the Gnome zealots want to use a Gnome-based one, and then all the desktop-agnostics (which I currently am) want to just use one which works under fvwm2?
(Yes, I know there's better WMs than fvwm2. But I have it setup to work nicely with my Datahand and I don't feel like learning how to configure another WM right now.)
Many things to think about.
---
"'Is not a quine' is not a quine" is a quine.
I've somehow managed to avoid getting spam for a while. I use the RBL religiously, I don't put my email address online promiscuously in a way that harvesters catch, and I'm constantly adding things to my .procmail/rc.spam filter. Yaaay. Still the occasional piece of spam reaches me, at which point I get horribly medieval on the spammer, the spammer's upstream, all the relays used, InterNIC, etc. I don't even do it with a script, as I take some sort of insane pleasure in dealing with it by hand.
---
"'Is not a quine' is not a quine" is a quine.
Would an IBM 3270 terminal qualify for this? It had a pretty funky protocol for doing that sort of thing with respects to its window interface.
---
"'Is not a quine' is not a quine" is a quine.
I think you're on crack. :)
---
"'Is not a quine' is not a quine" is a quine.