It's worse than that. MS only appears to care about big customers, typically large corporations, institutes, and government departments. i.e places that are behind a firewall and have (nominally) competent IT staff to keep the network running smoothly. Just look at the number of TCP/UDP ports they keep open. That sort of behaviour is ok on a safe intranet, but it's sheer negligence for home users connected directly to the internet. I'm constantly seeing incoming requests to the "windows networking" ports (137, 135, 445) on my ADSL connection. Those ports just should not be open to the wider internet. And lastly, witness the number of error dialog boxes in windows that simply advise the user to seek help from "the network administrator".
It's the big customers that MS cares about, not the home users. And we're all worse off when the latest round of worms clog up the internet.
I think you'll find that Comcast is only blocking outgoing connections TO port 25, and only after your behaviour has been deemed suspicious. They're not blocking incoming ports, if I understand it correctly. Your ISP however, is probably acting more out of a no-server policy rather than a desire to block spam.
Now look at the IM universe. AIM, MSNM, Y!M, ICQ, Jabber, etc. Of all those systems, only two work together...
Woah there boy. Methinks your inclusion of Jabber in that list was a little hasty. Jabber has many components for connecting to other IM networks. I noticed that the Yahoo component has stopped working for me today. Oh well, no loss. I don't know anyone on the Yahoo network. I just added it for the hell of it because I already had a Yahoo email account. Bye bye, Yahoo!
Like the AC said, put Exchange behind a proper MTA. Keep your exchange server inside the firewall for the suits to fiddle with their calendars and crap. Setup Postfix, Qmail, Sendmail, Exmim or some other MTA as your internet-facing email server. I use Postfix with Amavis forming a nice interface to Clam-AV and SpamAssassin. I don't run exchange though. Can't help you there.
You have to provide Spamassassin with good samples to train its bayesian filter properly. Autolearning only kicks in when the message gets a suitably high score (along with other criteria) that it can be sure the message really is spam. If you're not training it and autolearning isn't kicking in, then it's no surprise that Spamassassin is not performing well.
Collect up all your false-positives and false-negatives and run these commands:
The -L option stops Spamassassin sending off reports to Razor. Razor doesn't seem to like it when you hammer it suddenly with a load of reports. I've had my Razor account revoked several times doing that:P
I can imagine that many spams have similar language to business email (e.g the standard Nigerian spam). I had a similar problem with my mother recieving financial emails (stocks and stuff). Spamassassin initially classed them as spam so for a while I had to drag the emails out of the Spam folder and specifically re-train Spamassassin's bayesian filter to understand the fine line between legitimate business/financial emails and spam. Now we're getting hit by spam emails with nothing but an image (or several) taken from an external server and a block of benign text. An extra honeypot account and custom perl code has helped (e.g re-train automatically when messages are dragged into/out of the spam folder). But these spams look like they'll be a challenge for bayesian filters. Hopefully the new heuristics coming in Spamassassin 3.0 will help.
The problem with pop-up warnings is that Windows users are already in the habit of clicking OK or hitting enter without reading them. They're already bombarded with so many that it's a pavlovian response. I know I've even had a few experiences on Windows machines of hitting enter too quickly and then thinking dang! what did that last dialog say? it looked different and might have been important.
If you're going to calculate that, then you should also calculate the time it takes to test every friggin' hotfix and update against all your mission-critical applications (both server-side and client-side). And calculate the cost of the alternative: having all or part of your business grind to a halt when an important app suddenly stops working because you blindly installed an MS update or hotfix.
E.g CodeRed was so sucessful because the hotfix (which, just like you say, was available a month or two beforehand) broke some important functionality of IIS. But most businesses decided that it was easier to wait for a proper fix or solution than to do without this feature (autoindex or something?) on their web site.
That would add $20 to $31 to the TCO for everyone that got infected.
Per virus/worm/trojan, which happens a few times a year. And throw in the cost of downtime to do regular re-installs, the rate of which does vary greatly (home user vs professional enterprise setup). Over a few years (WinXP is three years old) these costs can easily rival the original purchase price of Windows.
An interesting use of OpenVPN is to bridge the OpenVPN TUN/TAP interface with the local ethernet interface. This way you have all your broadcast packets going over the VPN and keeping Network Neighborhood and other b'cast protocols happy. Plus you only have one address space and don't have to stuff around with a seperate subnet. I haven't done much with IPsec, but I don't think it can do either of these things. Still, IPsec does have the whole Standrd thing going for it.
You seem to think that Ogg Vorbis and/or Theora must have massive mainstream success (like MP3) or it is worthless. They're not commercial products. As long as people are maintaining the code base and using the software, then it has "worth". I don't think we'll see mainstream success, at least not for a while. And we shouldn't depend on success in that one market, or the lack of it, to measure the general success of Ogg Vorbis or Theora.
As others have pointed out, Ogg Vorbis is being used by a number of games for the packaged music. I think we'll see the Ogg codecs (Vorbis, Flac, Speex, and Theora) showing up in more embedded packages/devices. Really anywhere there is the need to handle audio and/or video in an application. The application is responsible for playing (and encoding too perhaps) the material. All the user does is install the application and uses it like any other app. It could be VOIP, a training/teaching program, a demo, cutscenes for a game, a TiVO/MythTV/Freevo-like set-top box, etc. You get the idea.
Wow, you really seem to have miscronstrued a lot of what I've written. Sorry if you think I was insulting your intelligence. I simply meant that any reasonably computer-literate person should be able to explore and find the setting.
...you are full of contradictions. Though you say that you are a "die-hard linux fan", I don't see you saying many good things about free software.
I'm not a cheerleader. I've used Linux and OSS/FS since about early 1996, so about 8 years now. But I'm not going to be some mindless fanboy. I'd prefer to leave that to the MS sycophants.
As for the three posts you pointed out:
How did I say bad things about "cruft" in Debian packages? I even say "And no, it's not a failing of Debian.". I simply tried to explain that there are some old packages that have been orphaned. Debian's been around a while. These things happen. I've been using it since 1998. And I'm not sure how I complained about the time it took GNOME 2.6 to go into Debian. The only thing I can see is the work "finally". Is that all it takes? One word?
I don't like PHP or Perl? Well, you're half right. And half wrong. I'm a big Perl fan. I even had a rather nice contract job for most of 2000 working on a large Perl codebase for a legal publisher. No it wasn't CGI or web related. It really was a rather large chunk of well written perl. I learnt a lot from that job. So in that sense I would consider myself a perl professional. In that second link, perhaps you should take my advice: I try not to stray into zealot territory. Zealotry is about emotion and irrational cheerleading.
I don't even have nice things to say about Koalas! WTF has this got to do with anything? You must think that just because they look cute and fluffy that any attack is just mean. Here's some information from down under: Koalas are not all that nice. They have long claws and are very territorial. And they're not bears.
My little comment at the end was in reference to your recent journal entry about having a stalking troll. I see he's hit you again in this thread. I read a lot of the links in his post and I tend to agree with him, at least on most of the posts he links to. I saw your post that started this thread and it was too much for me.
I don't know how you missed that.
It's simple. I wasn't paying too much attention. I mostly use terminals to do almost all of my file operations anyway. So I wasn't on the lookout for every piece of Nautilus news. And I figured that I'd cross that bridge when I got to it. Gconf isn't quite as bad as the windows registry, but this setting is still obscure and not easy to find without knowing beforehand where to go/look. I think it would be much better to place the setting in the preferences dialog for nautilus. I don't like the idea of all these "hidden" settings in Gconf i.e settings that don't have a control in the application. Create an "advanced" control panel or something. It doesn't matter. There should be a nice interface, preferably complete with tooltips and a help page. Although, I just noticed that Gconf does have the ability to attach both a long and short description to each setting. I was about to suggest just such a thing!
The Windows UI might be easy for me? Hell no. It agrevates me to no end. No virtual desktops. No sloppy focus. No window shading (?). No paste on middle-click. The XP Luna theme looks like some sort of toy for the sight-impaired. And the Windows environment is no better. An aweful lot of windows apps still don't allow network paths in their file dialogs, requiring them to be mapped to drive letters (a single root and mounting filesystems on unix/linux works beautifully).
Look I see you're trying to be some sort of advocate, but you're going about it the wrong way. It's good t
I'm a die-hard Linux fan and I still know where to look for this in windows. For anyone with some intelligence it's not hard to find it in under a minute, certainly a lot easier to find than the single entry in the Gconf editor. I had to go through this just this weekend when the GNOME 2.6 packages finally made their way into the main Debian repository. I hadn't paid too much attention when this whole "spatial" controversy had started. Mainly because the term "spatial" didn't mean much to me in the context of a file browser. And any discriptions were long-winded and didn't quickly point out the biggest point: it opens a new f**king window for each folder! A quick google turned up a few pages with the simple instructions for turning it off. It wasn't hard, but it was certainly more trouble than simply going to the Preferences dialog box.
(To twitter: I'm starting to think you really are a Linux zealot troll. You're off my friends list for now)
Yes, yes. Choice. But there is a lot of "cruft" in the official Debian package respitory. Subscribe to DWN (Debian Weekly News) or check out the archives. They have a section at the end for orphaned packages. There's quite a few now. And no, it's not a failing of Debian. It's human nature. Debian is a community-run organisation. People become disinterested, or no longer have free time, and move on to other things. Simple.
Yes, I agree most of those examples are a bit out there. But I think questions #3 and #4 interesting. Just where is then of an associative array? When you add to the end, does it find the greatest (integer) key and add 1 to it? And what happens when you try to iterate over the elements in an array (treating it like a regular indexed array) and it has gaps?
These are very important questions. PHP is meant to be a simple and easy language for non-programmers to pick up. But it has many issues such as this one, where intended simplicity has in fact resulted in complexity and ambiguity.
You can do type casting in PHP? Wow, what a confused language PHP is. Is it loosely or strongly typed?
So to use examples that are a little more realistic -
In Perl ($a == 5) becomes ((int)$a == 5) in PHP.
And ($s eq 'name') becomes ((string)$s == 'name').
The type casting effectively forces the == (or !=) operator to do what you want it to do.
Would you need two casts when comparing two variables? Like ((int)$a == (int)$b). Or is it just necessary to cast one? I'm thinking you need both casts, unless you know the type of one (e.g a loop variable). In which case you only need to cast the unknown variable (e.g some input).
What index do we have to consult to get the value 11 out of $a1?
What's the iteration order for $a2? Is it numerically-indexed or hash-style?
What happens if you use PHP's $a2[] =... construct to add a new element to the 'end' of the array?
Can numerically-indexed arrays have elements missing? If so, can you still trivially iterate over the values in index order?
While these questions can be answered (by careful reading of the manual and not a little experimentation), the answers aren't entirely obvious. Programmers with a reasonable understanding of basic data structures would be well advised to program as if a given PHP array can be indexed by either contiguous integers or by strings, but not both.
Just a minor point - That's still not the same thing as Perl's split operators. These tripled operators (=== and !==) still focus on the supposed type of the variables, even though PHP is a loosely-typed language. With perl you can still do 3 == "3" to do a numerical comparison, or 3 eq "3" to do a string comparison. In PHP, 3 == "3" would always match, not matter which comparison it decided to use, and 3 === "3" would never match because they're not the same type. I still think Perl does it the right way, because of the whole loosely-typed language thing. As to whether it makes a big difference, I don't know. It's been several years since I've done much PHP coding.
Slashdot Mirror
It's worse than that. MS only appears to care about big customers, typically large corporations, institutes, and government departments. i.e places that are behind a firewall and have (nominally) competent IT staff to keep the network running smoothly. Just look at the number of TCP/UDP ports they keep open. That sort of behaviour is ok on a safe intranet, but it's sheer negligence for home users connected directly to the internet. I'm constantly seeing incoming requests to the "windows networking" ports (137, 135, 445) on my ADSL connection. Those ports just should not be open to the wider internet. And lastly, witness the number of error dialog boxes in windows that simply advise the user to seek help from "the network administrator".
It's the big customers that MS cares about, not the home users. And we're all worse off when the latest round of worms clog up the internet.
So would I, except for the really bad review it got just over a month ago.
I think you'll find that Comcast is only blocking outgoing connections TO port 25, and only after your behaviour has been deemed suspicious. They're not blocking incoming ports, if I understand it correctly. Your ISP however, is probably acting more out of a no-server policy rather than a desire to block spam.
Open1x
Haven't used it myself but I have looked at it. It uses FreeRADIUS, which authenticates against LDAP or various SQL databases.
I think we can add this title to the list of contradictory terms:
Woah there boy. Methinks your inclusion of Jabber in that list was a little hasty. Jabber has many components for connecting to other IM networks. I noticed that the Yahoo component has stopped working for me today. Oh well, no loss. I don't know anyone on the Yahoo network. I just added it for the hell of it because I already had a Yahoo email account. Bye bye, Yahoo!
er, right. What are we talking about again?
Like the AC said, put Exchange behind a proper MTA. Keep your exchange server inside the firewall for the suits to fiddle with their calendars and crap. Setup Postfix, Qmail, Sendmail, Exmim or some other MTA as your internet-facing email server. I use Postfix with Amavis forming a nice interface to Clam-AV and SpamAssassin. I don't run exchange though. Can't help you there.
You have to provide Spamassassin with good samples to train its bayesian filter properly. Autolearning only kicks in when the message gets a suitably high score (along with other criteria) that it can be sure the message really is spam. If you're not training it and autolearning isn't kicking in, then it's no surprise that Spamassassin is not performing well.
Collect up all your false-positives and false-negatives and run these commands:
The -L option stops Spamassassin sending off reports to Razor. Razor doesn't seem to like it when you hammer it suddenly with a load of reports. I've had my Razor account revoked several times doing that :P
I can imagine that many spams have similar language to business email (e.g the standard Nigerian spam). I had a similar problem with my mother recieving financial emails (stocks and stuff). Spamassassin initially classed them as spam so for a while I had to drag the emails out of the Spam folder and specifically re-train Spamassassin's bayesian filter to understand the fine line between legitimate business/financial emails and spam. Now we're getting hit by spam emails with nothing but an image (or several) taken from an external server and a block of benign text. An extra honeypot account and custom perl code has helped (e.g re-train automatically when messages are dragged into/out of the spam folder). But these spams look like they'll be a challenge for bayesian filters. Hopefully the new heuristics coming in Spamassassin 3.0 will help.
Um, is Munich significant enough?
The problem with pop-up warnings is that Windows users are already in the habit of clicking OK or hitting enter without reading them. They're already bombarded with so many that it's a pavlovian response. I know I've even had a few experiences on Windows machines of hitting enter too quickly and then thinking dang! what did that last dialog say? it looked different and might have been important.
If you're going to calculate that, then you should also calculate the time it takes to test every friggin' hotfix and update against all your mission-critical applications (both server-side and client-side). And calculate the cost of the alternative: having all or part of your business grind to a halt when an important app suddenly stops working because you blindly installed an MS update or hotfix.
E.g CodeRed was so sucessful because the hotfix (which, just like you say, was available a month or two beforehand) broke some important functionality of IIS. But most businesses decided that it was easier to wait for a proper fix or solution than to do without this feature (autoindex or something?) on their web site.
Per virus/worm/trojan, which happens a few times a year. And throw in the cost of downtime to do regular re-installs, the rate of which does vary greatly (home user vs professional enterprise setup). Over a few years (WinXP is three years old) these costs can easily rival the original purchase price of Windows.
An interesting use of OpenVPN is to bridge the OpenVPN TUN/TAP interface with the local ethernet interface. This way you have all your broadcast packets going over the VPN and keeping Network Neighborhood and other b'cast protocols happy. Plus you only have one address space and don't have to stuff around with a seperate subnet. I haven't done much with IPsec, but I don't think it can do either of these things. Still, IPsec does have the whole Standrd thing going for it.
Not anymore. It looks like EA/ACL's for ReiserFS are in 2.6.7.
You seem to think that Ogg Vorbis and/or Theora must have massive mainstream success (like MP3) or it is worthless. They're not commercial products. As long as people are maintaining the code base and using the software, then it has "worth". I don't think we'll see mainstream success, at least not for a while. And we shouldn't depend on success in that one market, or the lack of it, to measure the general success of Ogg Vorbis or Theora.
As others have pointed out, Ogg Vorbis is being used by a number of games for the packaged music. I think we'll see the Ogg codecs (Vorbis, Flac, Speex, and Theora) showing up in more embedded packages/devices. Really anywhere there is the need to handle audio and/or video in an application. The application is responsible for playing (and encoding too perhaps) the material. All the user does is install the application and uses it like any other app. It could be VOIP, a training/teaching program, a demo, cutscenes for a game, a TiVO/MythTV/Freevo-like set-top box, etc. You get the idea.
Wow, you really seem to have miscronstrued a lot of what I've written. Sorry if you think I was insulting your intelligence. I simply meant that any reasonably computer-literate person should be able to explore and find the setting.
I'm not a cheerleader. I've used Linux and OSS/FS since about early 1996, so about 8 years now. But I'm not going to be some mindless fanboy. I'd prefer to leave that to the MS sycophants.
As for the three posts you pointed out:
My little comment at the end was in reference to your recent journal entry about having a stalking troll. I see he's hit you again in this thread. I read a lot of the links in his post and I tend to agree with him, at least on most of the posts he links to. I saw your post that started this thread and it was too much for me.
It's simple. I wasn't paying too much attention. I mostly use terminals to do almost all of my file operations anyway. So I wasn't on the lookout for every piece of Nautilus news. And I figured that I'd cross that bridge when I got to it. Gconf isn't quite as bad as the windows registry, but this setting is still obscure and not easy to find without knowing beforehand where to go/look. I think it would be much better to place the setting in the preferences dialog for nautilus. I don't like the idea of all these "hidden" settings in Gconf i.e settings that don't have a control in the application. Create an "advanced" control panel or something. It doesn't matter. There should be a nice interface, preferably complete with tooltips and a help page. Although, I just noticed that Gconf does have the ability to attach both a long and short description to each setting. I was about to suggest just such a thing!
The Windows UI might be easy for me? Hell no. It agrevates me to no end. No virtual desktops. No sloppy focus. No window shading (?). No paste on middle-click. The XP Luna theme looks like some sort of toy for the sight-impaired. And the Windows environment is no better. An aweful lot of windows apps still don't allow network paths in their file dialogs, requiring them to be mapped to drive letters (a single root and mounting filesystems on unix/linux works beautifully).
Look I see you're trying to be some sort of advocate, but you're going about it the wrong way. It's good t
I'm a die-hard Linux fan and I still know where to look for this in windows. For anyone with some intelligence it's not hard to find it in under a minute, certainly a lot easier to find than the single entry in the Gconf editor. I had to go through this just this weekend when the GNOME 2.6 packages finally made their way into the main Debian repository. I hadn't paid too much attention when this whole "spatial" controversy had started. Mainly because the term "spatial" didn't mean much to me in the context of a file browser. And any discriptions were long-winded and didn't quickly point out the biggest point: it opens a new f**king window for each folder! A quick google turned up a few pages with the simple instructions for turning it off. It wasn't hard, but it was certainly more trouble than simply going to the Preferences dialog box.
(To twitter: I'm starting to think you really are a Linux zealot troll. You're off my friends list for now)
Yes, yes. Choice. But there is a lot of "cruft" in the official Debian package respitory. Subscribe to DWN (Debian Weekly News) or check out the archives. They have a section at the end for orphaned packages. There's quite a few now. And no, it's not a failing of Debian. It's human nature. Debian is a community-run organisation. People become disinterested, or no longer have free time, and move on to other things. Simple.
Yes, I agree most of those examples are a bit out there. But I think questions #3 and #4 interesting. Just where is then of an associative array? When you add to the end, does it find the greatest (integer) key and add 1 to it? And what happens when you try to iterate over the elements in an array (treating it like a regular indexed array) and it has gaps?
These are very important questions. PHP is meant to be a simple and easy language for non-programmers to pick up. But it has many issues such as this one, where intended simplicity has in fact resulted in complexity and ambiguity.
You can do type casting in PHP? Wow, what a confused language PHP is. Is it loosely or strongly typed?
So to use examples that are a little more realistic -
In Perl ($a == 5) becomes ((int)$a == 5) in PHP.
And ($s eq 'name') becomes ((string)$s == 'name').
The type casting effectively forces the == (or !=) operator to do what you want it to do.
Would you need two casts when comparing two variables? Like ((int)$a == (int)$b). Or is it just necessary to cast one? I'm thinking you need both casts, unless you know the type of one (e.g a loop variable). In which case you only need to cast the unknown variable (e.g some input).
Sorry, I was thinking of one of the sections this web page and got mixed up. I'll quote some of it here just to be clear:
Just a minor point - That's still not the same thing as Perl's split operators. These tripled operators (=== and !==) still focus on the supposed type of the variables, even though PHP is a loosely-typed language. With perl you can still do 3 == "3" to do a numerical comparison, or 3 eq "3" to do a string comparison. In PHP, 3 == "3" would always match, not matter which comparison it decided to use, and 3 === "3" would never match because they're not the same type. I still think Perl does it the right way, because of the whole loosely-typed language thing. As to whether it makes a big difference, I don't know. It's been several years since I've done much PHP coding.