Abso-fuckin-lutely. Where I work, telnet is a very bad word. It's SSH or nothing.
Do remember that SSH gives you other advantages than hiding your passwords, as important as that is.
Say you want to access a box behind a masquerading box. On the box behind the masquerader, you can run a reverse-tunnel (-R) to a box you already have access to. Then you can contact the box you normally have no access to through the box you tunnelled to. It comes in very handy when you have to support masqueraded boxes.
That kind of stuff, when you know how to do it, makes your bosses think you're some kind of god.:-)*
Yeah, I was annoyed by Katz as well, even though I used to stick up for him.
His story would have been *perfect* if he left out that technology angle he applies to *everything*.
If I ignore the technology part, I see a very well written and clear story that hits you right where it matters.
To be fair, it seems that he started the whole technology-in-every-story thing due to the gripes of Slashdotters. ("This is not news for nerds! Go away!")
I think Katz should just leave out the technology angle where it doesn't fit (most of the time) and concentrate on the articles themselves. I'd probably start reading him again (I used to read and enjoy all of his articles) if he did that...
That was exactly my first thought. Actually, it always is whenever this comes up.
If the US enforces these backdoors, that will mean that there *are* backdoors. Do you honestly think the US military will use the same encryption schemes?
Just as they have to protect themselves from their enemies, we have to protect ourselves from our own enemies.
The worst thing, of course, is that they will essentially force the rest of us (I'm Canadian) to use their garbage as well, if we want to communicate securely to Americans.
I've read stories about the NSA et al. doing intelligence on foreign companies and relaying the data to American ones. That scares the shit out of me.
It's good to know I helped someone. I found IBM DeveloperWorks to be a very good resource for both XML and Python info, among many other languages.
I probably should have mentioned a couple of things:
Using Expat in xml_objectify speeds up processing and decreases memory usage by a few orders of magnitude. I learned this while loading a 1.8MB file with no PCDATA larger than about 1k (TV listings)...
In xml_objectify, UTF-8 is the default encoding. I had to change it, in xml_objectify.py, to ISO-8859-1 for an app I was writing, that used data containing many French characters (I'm in Canada).
it would be nice if you reported the difference between that and your alternative.
I take it you've never used TeX.
I recommend trying out LyX, an advanced frontend. After going through the tutorial, I promise that you will be forever annoyed with all other "word processors."
Unfortunately, due to marketing forces, I had to stop using LyX at work some time ago. Every time I write a document now, I feel like I'm using a frigging hammer-and-chisel.
Use your fave language and load the XML into your own data structure.
I did a project using DOM and, while that's all well and good for C and Java (I recommend it highly for those languages), I was using Python and was spoiled by the way Python works with regards to large, complicated data structures, which is, quite well.
Later, I found this article about a module called xml_objectify, which transforms XML into a data structure that Python people (and probably LISP and even Perl people as well) would feel more comfortable with. Remember that we could care less about index numbers half the time.:-)*
Whether you use Python or not, I highly recommend the article for it's discussion on the topic of converting XML into complex data structures in your fave language.
There are still people hung up on the perceived XML == ++HTML thing.
I realised the importance as you have, and use it *constantly*. I use it for all stored files, data interchange, and I even stick XML-RPC into everything now.
While it still is a format, I realised it was better to think of it as a protocol. It, for some reason, made more sense to me.
<REPLY TYPE="response to troll">How can you say that Windows NT is better at running Broadcast 2000 than Linux? It doesn't even run under Windows! RTFM!</REPLY>
My Logitech Mouseman Optical is very nice and light. I find it very comfortable to use and it is symmetrical, so you'd likely have the same experience as me.
Where there's no ball or mechanical parts, the weight is very low, while not feeling 'cheap.' I also like the fact that I can use it on just about any surface, including the cushions on my couch or my pant leg. Not only are mouse pads not required, they are a hinderance as well.
Over at Spacedaily, there is an article about how a 3D printer was used to fabricate a replacement part in a production environment--the first known case.
I've personally seen steel parts created directly from CAD drawings years ago. The machine I saw was made by Mitsubishi. It's standard fare in an auto-parts factory.
This has been going on for a lot longer, too. You can see all kinds of articles about the technology at Don Lancaster's site.
There are also some custom auto shops that have deals with factories to get parts made for their customers, at a *very* high cost, of course.
For continued production, this is only cost effective for one-at-a-time parts. It is always significantly cheaper to retool your presses (one produced every few seconds) than to waste time with a 'Santa Claus machine' (one produced every few minutes) when you have to ship 1,000,000 units.
Not a happy day for fans of open source business models.
Ok, lessee here, an OS database company that few people have heard of and a company that many of us have heard of but have been taking a very self-destructive path over the past couple of years (if you don't know what I'm talking about, do a search for Caldera in the/. archives).
How, exactly, does this spell bad times for open source business models?
I'd submit the fact that most of the companies that have been fucked over the past year or so were not actually open source companies.
This has everything to do with bad management and worse decisions. Even before the downturn, only 1 in 10 new companies survived their first year.
Re:Beauty for beauty's sake makes crappy software
on
Software Aesthetics
·
· Score: 5, Informative
And with all due respect to engineers, most software is a lot more complicated than building a bridge.
...and you know this how?
Ever designed a bridge? Do you know how long it takes? Did you know you only get one chance to get it right during the implementation stage?
Yours is the attitude that spawns crappy software.
You might tolerate a Windows BSOD, but what if your TV started crapping out on a daily basis? Do you know how involved the TV design process is? There's a hell of a lot more to it than you think and there's a lot more thought put into it than 90% of the code out there. Then you have such things as life-support systems and critical automation, which are a *lot* harder to do than any software out there.
Oh, and I'm saying this as someone who does both, BTW.
The ISO's contain software packages (not SuSE's) that are not to be downloaded. You are allowed to install on multiple computers, share the discs, etc. as long as those packages (a couple in the 'pay' series) are not put up for public download.
Rather than maintain two sets of ISO's plus the full tree, they maintain one sellable set and allow users to install from the distribution tree directly from them or one of the mirrors. You only need one to three floppies (depending on your hardware) to install over 6GB of software. I find it much quicker than downloading 6 ISOs, burning them, and then installing them so the users don't usually mind.
You may mirror it yourself as well. I buy the discs (but I only use the DVD to install) *and* I keep a tree on a server at work so I don't have to get the DVD whenever I want a certain package.
You obviously know very little of either BSD or SuSE... or maybe both.
BSD, like the other poster said, has one giant script per runlevel. If you want to have something start, you add the necessary shell code, usually a check for the existence of the binary, a message echoed to the screen, and therunning of the program itself.
rc.config is *nothing* like that. It's just a huge list of shell variables (well commented, like Apache's httpd.conf) that gets sourced by every init script. That's useful because you have 'START_X' variables for various programs. If a program's start variable is set to "no" then the init script will exit. Also, any init script requiring an IP or some other piece of information, can get it from the variable in rc.config. This way, you *never* have to edit your init scripts and re-edit them when you upgrade the RPM's or the entire distro.
It's also used by SuSEconfig, which reads the variables and sets up a number of configuration files, such as resolv.conf or the sendmail configs, or sets up symlinks, if necessary to various things.
How you've concluded that rc.config is "BSD-style" is totally beyond me.
Ummm, you obviously didn't look very hard. Check out this directory.
Actually, how in hell did you miss it?
Oh wait, I forgot. This is Slashdot, you didn't even bother checking, did you? You simply believed someone else, who has a stupid grudge, without proof.
BTW, I know this will seem incredibly petty and shallow, but IMHO the real reason BeOS didn't take off was C++.
Not only petty and shallow, but totally wrong as well.
Are you trying to tell me that Microsoft doesn't use C++?
Lemme guess, you're a GTK/GNOME zealot, right?
Do you honestly think that Be would be in a better situation if they had used straight C? There are bigger things at work than the mere language of implementation.
Ok, it appears there's a lot more there than I thought there would be. The list of 1921 discs is longer than the 1922 list and I haven't even checked for older ones. I'll stop here, but I'll bet at least one of those discs are CSS encrypted.
I have Metropolis on DVD right in front of me. I'd consider that old, especially considering it's not even a 'talkie'.
You'd be surprised at what you can find in digital form. I live in Canada and I've found that http://www.cnl.com has a lot of old stuff that you won't find in your local Future Shop (even though I found Metropolis there).
However, the service would have to get popular before you'll see old movies trickle in but the infrastructure, even though it requires big-ass servers is still cheaper than pressing DVDs.
Slashdot Mirror
Abso-fuckin-lutely. Where I work, telnet is a very bad word. It's SSH or nothing.
:-)*
Do remember that SSH gives you other advantages than hiding your passwords, as important as that is.
Say you want to access a box behind a masquerading box. On the box behind the masquerader, you can run a reverse-tunnel (-R) to a box you already have access to. Then you can contact the box you normally have no access to through the box you tunnelled to. It comes in very handy when you have to support masqueraded boxes.
That kind of stuff, when you know how to do it, makes your bosses think you're some kind of god.
I can't telecommute. It's not like my company won't let me or anything, but it's simply that I won't get anything done with any efficiency.
When I'm home, I do home stuff, when I work I do work stuff. Trying to put both together will be a total failure, no matter how hard I try.
Besides, I *like* drinking beer at lunch with my coworkers...
Yeah, I was annoyed by Katz as well, even though I used to stick up for him.
His story would have been *perfect* if he left out that technology angle he applies to *everything*.
If I ignore the technology part, I see a very well written and clear story that hits you right where it matters.
To be fair, it seems that he started the whole technology-in-every-story thing due to the gripes of Slashdotters. ("This is not news for nerds! Go away!")
I think Katz should just leave out the technology angle where it doesn't fit (most of the time) and concentrate on the articles themselves. I'd probably start reading him again (I used to read and enjoy all of his articles) if he did that...
That was exactly my first thought. Actually, it always is whenever this comes up.
If the US enforces these backdoors, that will mean that there *are* backdoors. Do you honestly think the US military will use the same encryption schemes?
Just as they have to protect themselves from their enemies, we have to protect ourselves from our own enemies.
The worst thing, of course, is that they will essentially force the rest of us (I'm Canadian) to use their garbage as well, if we want to communicate securely to Americans.
I've read stories about the NSA et al. doing intelligence on foreign companies and relaying the data to American ones. That scares the shit out of me.
It's good to know I helped someone. I found IBM DeveloperWorks to be a very good resource for both XML and Python info, among many other languages.
I probably should have mentioned a couple of things:
Using Expat in xml_objectify speeds up processing and decreases memory usage by a few orders of magnitude. I learned this while loading a 1.8MB file with no PCDATA larger than about 1k (TV listings)...
In xml_objectify, UTF-8 is the default encoding. I had to change it, in xml_objectify.py, to ISO-8859-1 for an app I was writing, that used data containing many French characters (I'm in Canada).
Hope that helps.
I take it you've never used TeX.
I recommend trying out LyX, an advanced frontend. After going through the tutorial, I promise that you will be forever annoyed with all other "word processors."
Unfortunately, due to marketing forces, I had to stop using LyX at work some time ago. Every time I write a document now, I feel like I'm using a frigging hammer-and-chisel.
Sigh.
You're right. XSL sucks hard.
:-)*
Use your fave language and load the XML into your own data structure.
I did a project using DOM and, while that's all well and good for C and Java (I recommend it highly for those languages), I was using Python and was spoiled by the way Python works with regards to large, complicated data structures, which is, quite well.
Later, I found this article about a module called xml_objectify, which transforms XML into a data structure that Python people (and probably LISP and even Perl people as well) would feel more comfortable with. Remember that we could care less about index numbers half the time.
Whether you use Python or not, I highly recommend the article for it's discussion on the topic of converting XML into complex data structures in your fave language.
Right on!
There are still people hung up on the perceived XML == ++HTML thing.
I realised the importance as you have, and use it *constantly*. I use it for all stored files, data interchange, and I even stick XML-RPC into everything now.
While it still is a format, I realised it was better to think of it as a protocol. It, for some reason, made more sense to me.
<MODERATION SCORE="-1">troll</MODERATION>
<REPLY TYPE="response to troll">How can you say that Windows NT is better at running Broadcast 2000 than Linux? It doesn't even run under Windows! RTFM!</REPLY>
<EXPRESSION TYPE="angry">Damn lameness filter!</EXPRESSION>
Damn, that was too funny! I just laughed my "portal" off.
My Logitech Mouseman Optical is very nice and light. I find it very comfortable to use and it is symmetrical, so you'd likely have the same experience as me.
Where there's no ball or mechanical parts, the weight is very low, while not feeling 'cheap.' I also like the fact that I can use it on just about any surface, including the cushions on my couch or my pant leg. Not only are mouse pads not required, they are a hinderance as well.
I've personally seen steel parts created directly from CAD drawings years ago. The machine I saw was made by Mitsubishi. It's standard fare in an auto-parts factory.
This has been going on for a lot longer, too. You can see all kinds of articles about the technology at Don Lancaster's site.
There are also some custom auto shops that have deals with factories to get parts made for their customers, at a *very* high cost, of course.
For continued production, this is only cost effective for one-at-a-time parts. It is always significantly cheaper to retool your presses (one produced every few seconds) than to waste time with a 'Santa Claus machine' (one produced every few minutes) when you have to ship 1,000,000 units.
Ok, lessee here, an OS database company that few people have heard of and a company that many of us have heard of but have been taking a very self-destructive path over the past couple of years (if you don't know what I'm talking about, do a search for Caldera in the
How, exactly, does this spell bad times for open source business models?
I'd submit the fact that most of the companies that have been fucked over the past year or so were not actually open source companies.
This has everything to do with bad management and worse decisions. Even before the downturn, only 1 in 10 new companies survived their first year.
...and you know this how?
Ever designed a bridge? Do you know how long it takes? Did you know you only get one chance to get it right during the implementation stage?
Yours is the attitude that spawns crappy software.
You might tolerate a Windows BSOD, but what if your TV started crapping out on a daily basis? Do you know how involved the TV design process is? There's a hell of a lot more to it than you think and there's a lot more thought put into it than 90% of the code out there. Then you have such things as life-support systems and critical automation, which are a *lot* harder to do than any software out there.
Oh, and I'm saying this as someone who does both, BTW.
Yup, the GPL specifically forbids arbitrary restrictions to specific groups, even if they are spammers.
It looks like Bero is trying to use the DMCA for 'good' but I don't agree with it, for the same reason the GPL disallows it.
We shouldn't be acting like the Adobes or the MPAAs of the world. We don't want to stoop to their level, even if spam annoys us.
The ISO's contain software packages (not SuSE's) that are not to be downloaded. You are allowed to install on multiple computers, share the discs, etc. as long as those packages (a couple in the 'pay' series) are not put up for public download.
Rather than maintain two sets of ISO's plus the full tree, they maintain one sellable set and allow users to install from the distribution tree directly from them or one of the mirrors. You only need one to three floppies (depending on your hardware) to install over 6GB of software. I find it much quicker than downloading 6 ISOs, burning them, and then installing them so the users don't usually mind.
You may mirror it yourself as well. I buy the discs (but I only use the DVD to install) *and* I keep a tree on a server at work so I don't have to get the DVD whenever I want a certain package.
The confusion probably comes from the fact that the chameleon's name is 'Geeko.'
You obviously know very little of either BSD or SuSE... or maybe both.
BSD, like the other poster said, has one giant script per runlevel. If you want to have something start, you add the necessary shell code, usually a check for the existence of the binary, a message echoed to the screen, and therunning of the program itself.
rc.config is *nothing* like that. It's just a huge list of shell variables (well commented, like Apache's httpd.conf) that gets sourced by every init script. That's useful because you have 'START_X' variables for various programs. If a program's start variable is set to "no" then the init script will exit. Also, any init script requiring an IP or some other piece of information, can get it from the variable in rc.config. This way, you *never* have to edit your init scripts and re-edit them when you upgrade the RPM's or the entire distro.
It's also used by SuSEconfig, which reads the variables and sets up a number of configuration files, such as resolv.conf or the sendmail configs, or sets up symlinks, if necessary to various things.
How you've concluded that rc.config is "BSD-style" is totally beyond me.
Probably because of SuSE's dedication to s390.
Probably because SuSE resells IBM stuff pre-packaged and integrated.
Probably because SuSE tries to cover every little detail for major software like DB2 and Notes.
Probably because SuSE pushes big-iron solutions for banks and the like, which just happens to be IBM's focus as well.
Why SuSE, indeed.
Ummm, you obviously didn't look very hard. Check out this directory.
Actually, how in hell did you miss it?
Oh wait, I forgot. This is Slashdot, you didn't even bother checking, did you? You simply believed someone else, who has a stupid grudge, without proof.
...License text... seems fair...
Just thought I'd share that...
There's a lot more to YaST2 than the cool graphic/text UI abstraction.
All of those YaST2 modules are scripts. Yup, you can make your own YaST2 modules. It's Perl-based, too. Check out the devel packages sometime.
Also, for running X programs as root in SuSE, use 'sux -' instead of 'su -' so that the X stuff is handled for you. I always use sux now.
Not only petty and shallow, but totally wrong as well.
Are you trying to tell me that Microsoft doesn't use C++?
Lemme guess, you're a GTK/GNOME zealot, right?
Do you honestly think that Be would be in a better situation if they had used straight C? There are bigger things at work than the mere language of implementation.
Does anybody know if Criterion CSS-encrypt *all* of their discs?
They distribute the first disc in the following list of pre-1923 DVDs for sale at CNL (The links may ask you to login as guest):
Ok, it appears there's a lot more there than I thought there would be. The list of 1921 discs is longer than the 1922 list and I haven't even checked for older ones. I'll stop here, but I'll bet at least one of those discs are CSS encrypted.
It would be interesting to know for sure, though.
I have Metropolis on DVD right in front of me. I'd consider that old, especially considering it's not even a 'talkie'.
You'd be surprised at what you can find in digital form. I live in Canada and I've found that http://www.cnl.com has a lot of old stuff that you won't find in your local Future Shop (even though I found Metropolis there).
However, the service would have to get popular before you'll see old movies trickle in but the infrastructure, even though it requires big-ass servers is still cheaper than pressing DVDs.