That's interesting. As a test, I loaded C|Net's home page once with "Accept all cookies" and once with "Only accept cookies originating from the same server as the page being viewed.".
With the former, I was prompted for two cookies, one from hg1.hitbox.com and one from ad.doubleclick.net, which I didn't get with the latter option selected. So it seems that on Navigator 4.7 on Linux at least, my interpretation is correct.
If you try this test yourself, do let me know the results, I'm very curious. The fact that the wordings on the menu differ is weird enough; it'd be really bizarre if the functionality wasn't the same as well.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Anyone (still) using Netscape Navigator can avoid this problem completely. In the Edit menu, go to Preferences, and click on the "Advanced" category; you'll see the configuration for cookie handling. One of the options is "Only accept cookies originating from the same server as the page being viewed." Check that and you're set.
As far as I can tell IE5 doesn't do this: The cookie settings are under Tools/Internet options in the security tab, but it's either Enable, Disable or Prompt for cookies.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Move it out of the kernel and you have multiple maintenance points (X, SVGA, Y, and any other graphics program of library)
No, not necessarily. Although this is how it stands now, there's now reason there couldn't be a single graphics API outside the kernel. For example, you could build an X server on top of SVGALib. (I'm not saying this is a good idea or that SVGALib in particular is well suitd to the task -- just that it's possible.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Putting drivers in the kernel does not magically confer stability; the code is only so good not matter where it is. In fact, if it's in user space, a program (be it SVGALib, X, whatever) has fewer prvileges and therefore has less of a chance of screwing up the *really* important stuff. X crashing and dumping you to the console is not nearly as bad as a kernel panic!
That being said, there are tradeoffs. Because userspace graphics need direct access to the hardware, they need to run setuid root, which means security issues. SVGALib is especially problematic because it's actually "foreign" code, linked against the library, that runs as root, rather than just the server (or xdm) in the case of X. I think it deals with this fairly intelligently, though, dropping root privileges at the first possible juncture. It's still not perfect.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
The kernel hasn't got anything to do with the C library. It's completely self-contained.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Re:Download the thing, use it, and then b***h...
on
Whither Netscape 5.0?
·
· Score: 1
Given XUL, shouldn't this be pretty easy actually?
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Re:Download the thing, use it, and then b***h...
on
Whither Netscape 5.0?
·
· Score: 3
The fact that the latest version of the developpers' source is always available is a big plus that has been overlooked a lot, I think. In a closed-source environment, even if the developpers are very responsive to bug reports and feature requests, the customer only gets results based on the release schedule -- the code they wanted could have been written for months.
Speaking of developper responsiveness, I just noticed something uber-cool over at Mozilla HQ: Bugzilla, the bug-tracking database, will now allow you to vote on what you want fixed quickest! I think this is a great way to get the community involved. I'm beginning to believe Moz' greatest contribution won't be the browser itself (which should be awesome when it's eventually ready) but the ancillary stuff like this, and Tinderbox, LXR, Bonsai, etc. etc.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Under "strategic partnerships", one of the names listed is "HalfPrice Hosting". So:
glait:/opt/nephtes/src\>host www.armed.net www.armed.net has address 208.141.56.223 glait:/opt/nephtes/src\>host www.halfpricehosting.com www.halfpricehosting.com has address 208.141.56.249
Same subnet.
So the reason the server is IIS is because that's what their hosting service uses. Now, you could argue that they ought to be pickier about who hosts their site, but at least they didn't actually decide to set up an NT web server to sell Linux! (No personal interest, incidentally, just thought I'd shed some light.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
while there isn't always inital recourse for stupidity like that it does tend to keep away new customers. (Btw the school I went to is now starting to seriously loose students..
So really, in a "free market" school system the only way parents would have more choice was if there were more schools to choose from -- since "keeping away new customers" works in both systems. How likely is this? The people with little choice now are in locations where the population density is low enough that not choosing the nearest school is unfeasible. Could those be big enough "markets" to sustain competing schools in a "free market" system? I doubt it.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Both of your examples seemed to be of public schools, that took money from private corportations. In which I'd be willing to bet the parents at those schools had absolutly no resource to what the school did (except maby legally)
No recourse? Hmm. Well, I don't know how authority is divided up in the American school system, but here in Quebec (Canada -- it's a provincial ministry) school boards are comprised of "officials" elected by the schools' parents. It works much like political representation. If it works the same way in Georgia, had the school board officials' "constituents" objected to that unbelievable "Coke" stunt, they certainly would have had the authority to prevent it, and severely reprimand whoever was responsible for suspending the kid with the Pepsi shirt.
The problem, of course, is that voter turn-out to school board elections is miserably low. You could argue that in the "free market" system, people would be more likely to exercise their prerogative by changing schools, which may be true. But that wouldn't be very convienient in the middle of the school year, not to mention the psychological impact on the kid due to being shuffled between schools every time their parents deemed something sufficiently objectionable. And what about the more insidious influences, such as biased learning materials, consumerist propaganda, etc. ?
In any case, in the current system parents certainly have a say in what happens in their kids' school (at least here.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Realy now, do you *really* think that a government (especially the crappy, inneficient, US government) could possibly do anything *better* than a corporation would do it in a compeditive, free market, enviornment?
Define "better". If you mean could they graduate a kid from grade 12 having spent less money, then no, probably not.
But that's not what I'm addressing: it's what the kid has learned at the end of those years that's much, much more important (to me anyways). Would you trust the lowest bidder (especially one with other interests) to teach impartial, accurate information?
I for one would not.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
We need taxes to support things like the highway system, public transportation, and public education.
Yes but do we need these things to be publicly funded? Often highly inefficient IMVAO
Hell yes! Are you seriously suggesting privately funded education? ("Welcome to MSClassroom 1.0, I'm your teacher Mrs. Vendu. Please take your seats. crrreeeeeaaaakk... Seats and desks all collapse, maiming most of the 200-student class)
In all seriousness, there are some things that private interests should not be allowed to do; education is definitely one of them. It's bad enough as it is. See also here and here for a few more examples.
I don't know about money for highways, but I do like to be able to use public transit (relatively) cheaply. Both of these, however, pale in comparison to the importance of keeping advertising out of education. Is anyone else thinking of Huxley's Brave New World here?
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Appeasing the "This isn't newsworthy" crowd
on
Phrack 55 released
·
· Score: 3
I suggest a new article category, "New releases", for this kind of story. I myself am happy to see them, but recognize that some might not.
In that vein, the idea has been bouncing around in my head for a while that it should be possible to put articles in two categories. Many are the times I've seen an article and thought, "Funny, I would've put that under 'Linux', not under 'SGI'."
Clink, clink, . SNF.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
I tried adding some *.ttf files to the ttfonts directory and running/etc/rc.d/init.d/xfs restart, but the fonts still aren't available.
init.d/xfs is just a script, which starts the actual server. The server binary itself is typically/usr/X11R6/bin/xfsft. Make sure it's present. If you're running the script with "restart", make sure xfsft is actually running (ps waux | grep xfsft). If not, use/etc/rc.d/init.d/xfs start for initialization.
Assuming that's ok, possible problems are:
The directory with the TrueType fonts isn't being pointed out to xfsft. Look for the config file loaded on the commandline by the init.d/xfs script; it may be/usr/local/etc/xfsft.conf. Check that file for the catalogue= option, which should point to the directory your fonts are in.
The font server is not in your font path. Diagnose by checking xinitrc files (/usr/X11R6/lib/xinit/xinitrc and ~/.xinitrc) for xset +fp; you should see something like xset +fp tcp/127.0.0.1:7100; xset fp rehash. It needs to be added if not there.
No fonts.dir file in the TrueType font directory. To generate it if needed, run the ttinst Perl Script (see my original post).
I think that about covers it. You can e-mail me if your problem still isn't solved...
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Why does everyone keep pining for TrueType fonts? You can have them in X now; I certainly do. You can either rebuild your X server with support compiled in, or have a separate, dedicated font server. I prefer the latter solution because it's easier to upgrade.
Here's how I did it:
Grab a copy of the FreeType font server here (for linux/x86 w/glibc2), h ere (for solaris/SPARC) or here (patch to XFree sources -- not for the faint of heart).
Put the xfsft executable somewhere in your $PATH.
Get a directory full of TT fonts. I have a directory on my Linux partition full of symlinks to/dosc/windows/fonts/*.ttf, for example./usr/X11R6/lib/fonts/tt is not a bad place.
Run the ttinst script in that directory; this will create a fonts.scale file.
Actually you can usually get away without even rebooting. Alt-SysRq-K kills all the processes running on a given VT, so if SVGALib something or X pukes, just hit Alt-SySRq-K (offending application dies horribly), Alt-Fx (switch to a VT where you've got a login running -- blindly, since your display is hosed), then run SVGALib's textmode utility.
Ta-dah! Unhosed system. This is the main reason I upgraded to Linux 2.2.
n.b. You need to have previously run savetextmode. And don't run the dangerous app on VT 1, or Alt-SysRq-K will kill off important stuff like init...
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
First of all, let's make the distinction between X and XFree86. A lot of people seem to have the two confused, by complaining for example about X's lack of multihead support, or poor hardware support. This in fact has nothing to do with X, but rather with the features (or lack thereof) available in specific implementations.
Another misconception is that X can't support TrueType fonts, or can only do so kludgily. Not true; you can compile TrueType font support directly into the display server; it becomes no different than PostScript font rendering. Stock XFree86 4.0 is supposed to ship with TT support built-in. Antialiasing (and possibly font hinting, not sure), however, is a real problem.
That being said, I'm no great fan of X myself, in particular the API. Just what I've seen from reading GDK source is monstrous. It's very true that a lot of important features were not provided for in the initial design and had to be "bolted on" afterwards.
Finally, I don't think YAX has been mentioned yet. The idea is similar to Berlin, which has been mentioned, and it's a very interesting project. They've only released an API so far, but it sounds nice: From their description it's actually rather Be-like. An actual implementation is apparently in progress and they're also in the process of porting GTK to it, in order to immediately give it a set of available applications.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Linux looks pretty darn good now, and there's no reason this couldn't be done again once it's even better.
Looks pretty darn good to whom? You like it, I like it, but how will it look to someone who's never seen anything but Windows?
What I'm worried about is that people will try it, be disappointed (What do you mean I can't play Tribes? What do you mean I can't use Excel? Why is Netscape so slow and crashy?), and forevermore have Linux pegged as useless in their minds.
What bugs me most about this is that there are so many projects that are so close, collectively, to providing the kind of features that will really impress these people. It would be heartbreaking if what I fear does in fact happen, and could be avoided by waiting, say, 6 months.
I don't know how likely that is. But if the people behind this project really push the media's buttons, and in fact Linux is a fairly hot item in the news these days (as far as technology goes)... Well, I just hope I'm wrong.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Considering this is going to be aimed primarily at the masses, in other words desktop users, I think it would've been wise to wait a while until Linux' desktop offerings improve. I'm thinking in particular of KOffice and Mozilla, but there are a lot of other not-quite-there-yet projects which could be better used to wow the great unwashed, including games above all. Celebrating Linux' "birthday" is a nice idea, but perhaps not very practical.
If this stunt gets all the attention it's trying to, Linux had better put in a damned good showing, because there won't be a second chance; this is going to be a first-impression experience for most.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
It was my understanding that the SeaMonkey releases were branches off the main source tree, tweaked for stability and generally more release-frienly. Is this no longer true? . SNF .
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Slashdot Mirror
With the former, I was prompted for two cookies, one from hg1.hitbox.com and one from ad.doubleclick.net, which I didn't get with the latter option selected. So it seems that on Navigator 4.7 on Linux at least, my interpretation is correct.
If you try this test yourself, do let me know the results, I'm very curious. The fact that the wordings on the menu differ is weird enough; it'd be really bizarre if the functionality wasn't the same as well.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
As far as I can tell IE5 doesn't do this: The cookie settings are under Tools/Internet options in the security tab, but it's either Enable, Disable or Prompt for cookies.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
No, not necessarily. Although this is how it stands now, there's now reason there couldn't be a single graphics API outside the kernel. For example, you could build an X server on top of SVGALib. (I'm not saying this is a good idea or that SVGALib in particular is well suitd to the task -- just that it's possible.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
That being said, there are tradeoffs. Because userspace graphics need direct access to the hardware, they need to run setuid root, which means security issues. SVGALib is especially problematic because it's actually "foreign" code, linked against the library, that runs as root, rather than just the server (or xdm) in the case of X. I think it deals with this fairly intelligently, though, dropping root privileges at the first possible juncture. It's still not perfect.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Speaking of developper responsiveness, I just noticed something uber-cool over at Mozilla HQ: Bugzilla, the bug-tracking database, will now allow you to vote on what you want fixed quickest! I think this is a great way to get the community involved. I'm beginning to believe Moz' greatest contribution won't be the browser itself (which should be awesome when it's eventually ready) but the ancillary stuff like this, and Tinderbox, LXR, Bonsai, etc. etc.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
glait:/opt/nephtes/src\>host www.armed.net
www.armed.net has address 208.141.56.223
glait:/opt/nephtes/src\>host www.halfpricehosting.com
www.halfpricehosting.com has address 208.141.56.249
Same subnet.
So the reason the server is IIS is because that's what their hosting service uses. Now, you could argue that they ought to be pickier about who hosts their site, but at least they didn't actually decide to set up an NT web server to sell Linux! (No personal interest, incidentally, just thought I'd shed some light.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
So really, in a "free market" school system the only way parents would have more choice was if there were more schools to choose from -- since "keeping away new customers" works in both systems. How likely is this? The people with little choice now are in locations where the population density is low enough that not choosing the nearest school is unfeasible. Could those be big enough "markets" to sustain competing schools in a "free market" system? I doubt it.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
No recourse? Hmm. Well, I don't know how authority is divided up in the American school system, but here in Quebec (Canada -- it's a provincial ministry) school boards are comprised of "officials" elected by the schools' parents. It works much like political representation. If it works the same way in Georgia, had the school board officials' "constituents" objected to that unbelievable "Coke" stunt, they certainly would have had the authority to prevent it, and severely reprimand whoever was responsible for suspending the kid with the Pepsi shirt.
The problem, of course, is that voter turn-out to school board elections is miserably low. You could argue that in the "free market" system, people would be more likely to exercise their prerogative by changing schools, which may be true. But that wouldn't be very convienient in the middle of the school year, not to mention the psychological impact on the kid due to being shuffled between schools every time their parents deemed something sufficiently objectionable. And what about the more insidious influences, such as biased learning materials, consumerist propaganda, etc. ?
In any case, in the current system parents certainly have a say in what happens in their kids' school (at least here.)
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Define "better". If you mean could they graduate a kid from grade 12 having spent less money, then no, probably not.
But that's not what I'm addressing: it's what the kid has learned at the end of those years that's much, much more important (to me anyways). Would you trust the lowest bidder (especially one with other interests) to teach impartial, accurate information?
I for one would not.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Yes but do we need these things to be publicly funded? Often highly inefficient IMVAO
Hell yes! Are you seriously suggesting privately funded education? ("Welcome to MSClassroom 1.0, I'm your teacher Mrs. Vendu. Please take your seats. crrreeeeeaaaakk... Seats and desks all collapse, maiming most of the 200-student class)
In all seriousness, there are some things that private interests should not be allowed to do; education is definitely one of them. It's bad enough as it is. See also here and here for a few more examples.
I don't know about money for highways, but I do like to be able to use public transit (relatively) cheaply. Both of these, however, pale in comparison to the importance of keeping advertising out of education. Is anyone else thinking of Huxley's Brave New World here?
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
I suggest a new article category, "New releases", for this kind of story. I myself am happy to see them, but recognize that some might not.
In that vein, the idea has been bouncing around in my head for a while that it should be possible to put articles in two categories. Many are the times I've seen an article and thought, "Funny, I would've put that under 'Linux', not under 'SGI'."
Clink, clink, . SNF .
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
My morning so far:
And so it goes... . SNF .
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
init.d/xfs is just a script, which starts the actual server. The server binary itself is typically /usr/X11R6/bin/xfsft. Make sure it's present. If you're running the script with "restart", make sure xfsft is actually running (ps waux | grep xfsft). If not, use /etc/rc.d/init.d/xfs start for initialization.
Assuming that's ok, possible problems are:
- The directory with the TrueType fonts isn't being pointed out to xfsft. Look for the config file loaded on the commandline by the init.d/xfs script; it may be
/usr/local/etc/xfsft.conf. Check that file for the catalogue= option, which should point to the directory your fonts are in. - The font server is not in your font path. Diagnose by checking xinitrc files (/usr/X11R6/lib/xinit/xinitrc and ~/.xinitrc) for xset +fp; you should see something like xset +fp tcp/127.0.0.1:7100; xset fp rehash. It needs to be added if not there.
- No fonts.dir file in the TrueType font directory. To generate it if needed, run the ttinst Perl Script (see my original post).
I think that about covers it. You can e-mail me if your problem still isn't solved...Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
xfsft -port 7100 -config /usr/X11R6/lib/fonts/tt/xfsft.conf &
Without that ampersand, X would never get past that line to the rest of the script.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Here's how I did it:
- Grab a copy of the FreeType font server here (for linux/x86 w/glibc2), h ere (for solaris/SPARC) or here (patch to XFree sources -- not for the faint of heart).
- Put the xfsft executable somewhere in your $PATH.
- Get a directory full of TT fonts. I have a directory on my Linux partition full of symlinks to
/dosc/windows/fonts/*.ttf, for example. /usr/X11R6/lib/fonts/tt is not a bad place. - Run the ttinst script in that directory; this will create a fonts.scale file.
- ln -s fonts.scale fonts.dir
- echo "catalogue=/usr/X11R6/lib/fonts/tt" >
/usr/X11R6/lib/fonts/tt/xfsft.conf - Add the following to your
.xinitrc:
And you're set!xfsft -port 7100 -config /usr/X11R6/lib/fonts/tt/xfsft.conf
sleep 1 # Give xfsft a chance to start up
xset +fp tcp/127.0.0.1:7100
xset fp rehash
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Ta-dah! Unhosed system. This is the main reason I upgraded to Linux 2.2.
n.b. You need to have previously run savetextmode. And don't run the dangerous app on VT 1, or Alt-SysRq-K will kill off important stuff like init...
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Another misconception is that X can't support TrueType fonts, or can only do so kludgily. Not true; you can compile TrueType font support directly into the display server; it becomes no different than PostScript font rendering. Stock XFree86 4.0 is supposed to ship with TT support built-in. Antialiasing (and possibly font hinting, not sure), however, is a real problem.
That being said, I'm no great fan of X myself, in particular the API. Just what I've seen from reading GDK source is monstrous. It's very true that a lot of important features were not provided for in the initial design and had to be "bolted on" afterwards.
Finally, I don't think YAX has been mentioned yet. The idea is similar to Berlin, which has been mentioned, and it's a very interesting project. They've only released an API so far, but it sounds nice: From their description it's actually rather Be-like. An actual implementation is apparently in progress and they're also in the process of porting GTK to it, in order to immediately give it a set of available applications.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
Looks pretty darn good to whom? You like it, I like it, but how will it look to someone who's never seen anything but Windows?
What I'm worried about is that people will try it, be disappointed (What do you mean I can't play Tribes? What do you mean I can't use Excel? Why is Netscape so slow and crashy?), and forevermore have Linux pegged as useless in their minds.
What bugs me most about this is that there are so many projects that are so close, collectively, to providing the kind of features that will really impress these people. It would be heartbreaking if what I fear does in fact happen, and could be avoided by waiting, say, 6 months.
I don't know how likely that is. But if the people behind this project really push the media's buttons, and in fact Linux is a fairly hot item in the news these days (as far as technology goes)... Well, I just hope I'm wrong.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
If this stunt gets all the attention it's trying to, Linux had better put in a damned good showing, because there won't be a second chance; this is going to be a first-impression experience for most.
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty
It was my understanding that the SeaMonkey releases were branches off the main source tree, tweaked for stability and generally more release-frienly. Is this no longer true? . SNF .
Steve 'Nephtes' Freeland | Okay, so maybe I'm a tiny itty