No, slow down there. It was had (directly) from PDF, which was used because Adobe didn't want them to use DisplayPS (which is apparently too old to be useful technology:). Apple got access to DisplayPS with an alpha channel from NeXT, but MOSX doesn't use DisplayPS.
Whoops, I thought I mentioned that. It didn't bother me so I didn't bother getting a new CD...
Either way, it points out the reason why tracking FreeBSD-stable for 'QA' purposes -- as was recently suggested on "BSD Today" -- is pointless -- bugs happen even in releases.
Because it's in a seperate section, "Ask Slashdot." Only the 'big' Ask Slashdot questions get front-page status. All of them are available in the slashbox.
For/. sections you're interested in, I recommend adding the slashbox for that section and removing the stories from your default page, so you see all of them precisely once on the page.
Well, first of all, the extension would probably have to implement pre-emptive multitasking in order to provide threads...
However, classic MacOS uses extra processors differently, essentially (as I recall) letting software that knows about the extra processor(s) take advantage of them by offloading tasks. For algorithms which are not easily parallizable, or programmers that are lazy, this makes some amount of sense -- it allows you to essentially devote a single processor to a single process, and use the other CPU for inconsequentials like displaying the results of the calculations.
When the system was bought/built/ordered with a specific multiprocessor app in mind (like Photoshop), and a specific purpose, it can work OK. Of course, for end-to-end improvement and responsiveness, pre-emptive multitasking, threading, and SMP are all necessary.
Since I'm not too familiar with it, hopefully someone else will come along and correct/inform me:)
I thought it was about the same. My only problem was that the installer bombed out if you tried to add packages from within it, so you have to had packages afterwards.
Now OpenBSD, that was easy to install. Seriously, I mean that -- just because it's secure doesn't mean it's hard to install. The only difficulty is in the fdisk/disklabel setup, which is relatively straightforward (although very different!) if you've used Linux fdisk before.
The biggest difference between Linux and BSD installs is that BSD doesn't install the kitchen sink by default (hint: if you ever try a commercial UNIX system, they don't either). OpenBSD is moving towards 'install all the sysadmin utilities a person could want, and then let the sysadmin install the user utilities' -- nc, ssh, a decent ftp client, etc. -- but it doesn't install any editor but vi:)
(a): MacOS X was demonstrated on an x86 box at WWDC (reportedly). Sounds pretty portable to me, if they don't even have a beta out, but they ported it to another platform. (b) device drivers are written in C++ using IOKit, and interface with BSD devices seperately -- much easier that straight BSD drivers. (c) MacOS X makes use of Mach threads, which are available.
I wouldn't expect Adobe to port anything to Linux for a long time. The last year or two has seen Adobe not even make their Mac products have the same features as their Windows versions, at least upon release.
I assume you mean other than Distiller and FrameMaker? Distiller is ported, and a FrameMaker beta for Linux is available. FrameMaker is definitely worth it, and I wouldn't be surprised if Photoshop were ported in 2001 (since it already runs on a variety of UNIX, it's mostly minor porting issues).
First, it's a simple database file. Second, if NetInfo fails to load (or the db is corrupted), almost everything that uses it can default to a flat configuration file (I have a system here with system filesystems listed in/etc/fstab, and NFS user partitions in NetInfo).
Second, it's much more flexible than the registry, since remote servers can offer up NetInfo stuff as well -- no sick hacks to copy/remove registry entries from remote systems as people login or crap like that -- so there's better delineation between network-wide and local information.
So we all know CmdrTaco is horrible at spelling, my question is, is "Oddessey" a knowing and humorous play on that, or is it an error? Given that the correct (or accepted as correct in English) spelling would be odyssey, is there any plan to correct the spelling?
I'm genuinely curious... with some people, it's difficult to tell the intentional from the unintentional spelling errors.
I think that good math and science courses, coupled with plenty of social interaction, are the most important individual elements (although I would actually consider a more balanced curriculum better than one in which math/science is tops, and everything else is non-existant). The reason is that it teaches a way to think that is both abstract and directly verifiable, so it's more critical of the thoughts. Then again, I think all of it is important.
Math doesn't just teach abstract reasoning, but it teaches abstract reasoning with a starting point in reality. A calculus, after all, is simply a stone (and it's stil used that way in the medical community, I believe) -- and to calculate is to play with stones, counting and measuring. Addition and subtraction are pretty clear concepts, but you can abstract them into multiplication, and then you can abstract the "undoing" of multiplication as division, and so on, until you're dealing with things that are apparent nowhere in the real world (not really true, but true prima facie).
I don't see calculators as being at all useful for teaching until after the skill has been taught. Addition is easy -- especially if you do it sufficiently concretely or abstractly that you use the way different people think, rather than work against the way they think -- and other arithmetic follows from it. As in all things, computers are useful so long as they don't try to give skills, but merely use them faster. What's the point of hiring a human to do something if all he does is push buttons unknowingly?
Memorization of multiplication tables is, IMO, a good example of 'testing' gone wrong. I mean, clearly anyone who can do multiplication with relative ease can reproduce multiplication tables as requested, so it appears (to my mind, which isn't very familiar with the history of math education) to have been a goad to get kids to learn to multiply. But the skill itself is not tested, and hence not learned. As for long division -- is there something wrong with learning it? How is "long" division different from "regular" multiplication when lots of digits are involved?
And finally, first get rid of the idea that the kids who get good grades in math are "gifted" or "inclined" to math; the two sets intersect, but are not the same. The reason they are not, of course, is because different people learn different ways, and in a lot of schools, everything is taught one way. The reason children who do well in math are considered gifted is because a lot of people have trouble with math, precisely because it's taught the wrong way to different kids! So, by teaching it only one way, they are creating an artificial elite out of the children who are basically competent and happen to learn things the "right" way. Bah.
Obviously this person has violated the click-license agreement. But what happens if Jeremy picks up the stripped specs and implements them? Jeremy has agreed to nothing. So is he accountable? Is there something in the DMCA or other law which would place him in jeopardy?
You mean, like copyright? Owning copyrighted without ever having been licensed to do so? Modifying copyrighted works?
Just because the information contained in the doc is a trade secret, doesn't mean that the doc itself has no protection. It's no different from grabbing a Word doc of Ender's Game and throwing it on the web, sans copyright information.
Take a look at http://www.deepspacetech.com/Hard ware/systems.htm. They offer a dual P90 with 32MB RAM, 10BaseT, and narrow SCSI for $65, and that should have plenty of serving power for a home network.
Then all you need is a better SCSI card (for Ultra2, LVD, and the like), your hard drives, and possibly an ethernet card (if you want 100BaseT). Everything on it's supported by FreeBSD and Linux.
Then your cost is, say, $75 for a decent SCSI-3 controller, $225 for a 20GB SCSI-3 drive, and $65 for the system. If you prefer to go with an IDE system (which is unfortunately how I'm using one at home), your costs are even less.
I know you said you don't want to turn a crappy box into a server (an opinion I strongly agree with, after having two PeeCee systems die doing 24x7 IPMasq), but these were originally intended to be servers and workstations (a system built when P90's were fast with 32MB of RAM!?), and it shows in the quality of the components.
No, I'm not affiliated with the guys at DeepSpace, but I've bought a lot of PC and NeXT gear from them and been extremely happy.
OS/2 (and possibly KDE or Gnome?) provides a very nice facility to store bookmarks in the filesystem. The name of the file is the 'bookmark' and the contents, in plain ASCII, is the URL. With lots of subdirectories this makes it very easy for me to locate bookmarks based on subject, and if that fails, any standard file searching tool (locate for bookmark names, grep for when you know part of the URL) works.
It also works across a LAN with shared filesystems, but for for access from anywhere on the 'Net, you'll have to use either one of these weird bookmark sites (sorry, I just can't get used to it), or else try your own LDAP server.
I'm not up on the technical details, SciTech being non-free software, but AFAIR they have an OS-specific driver which runs a small virtual machine, mapping system calls in a theoretical 'nice' OS to system calls on the current OS, and then a binary 'driver' which is loaded, and interacts.
May sound questionable (performance wise), but their drivers work in OS/2 and Windows at least (I remember reading that they also had a Linux version for X displays), and I've heard good things about both stability and performance.
1. Darwin and FreeBSD share kernel code (above the lowest level, where Darwin is Mach-based), from what I understand, and Darwin and NetBSD share userland code (i.e., I believe a lot of the userland was pulled over from NetBSD). So, it should be relatively easy to pull out userland from Darwin and try to run it on a NetBSD system. As for kernel code... I don't know. Yet.
2. It's under the APSL.
Some of the Darwin code has already been brought over to other systems, notably (IMO) the NetInfo stuff.
For those that don't know, NetInfo is a configuration system like NIS, in that a particular system serves up all the machine-inspecific information (like hostnames, network printers, users, etc.). Unlike NIS, it provides for multi-level domains, so (most basically) a single server provides generic information to several computers, and then each computer provides through NetInfo information for itself.
It gets better, though, when you add more machines into the mix -- you can have users (like the CTO, maybe?) that can login to any machine, and then users that can login to any of the machines in a particular dept. (like if you have cubicles on a first-come, first-serve basis -- ick;), and you can also have users that can only login to one machine. You can carry this up to as many levels as you want, and in large settings could be quite useful.
With MacOS X coming out soon, it's possible that for the first time since OS/2 was popular there will be another consumer PC operating system able to work along with or replace NT, but it's also UNIX that supports storing the information samba uses in network databases (NetInfo, NIS), and it also supports providing access to older Macs through Appletalk.
My understanding of, for instance, Mac Services for Windows NT and UNIX Services for Windows NT is that it provides services from the same databases, just with different protocols.
So if you can see where this is going, is there any work on making samba able to make use of network-wide databases for user authentication, share specification (I know it can already use the autohome map, but more than that!), etc.?
In particular, I'm interested in things like:
Being able to authenticate netatalk, samba, and UNIX users all the same way (i.e., not having smbpasswd, NIS, and/etc/passwd all need to be updated every time a user changes his password or is added)
Being able to specify at the same time what my file server serve up, via netatalk, samba, and NFS (so I don't edit three configuration files every time I add a share, or move a share)
Being able to specify from one system what each and every file server serves up, without having to connect to the machine in question and edit the smb.conf by hand (or by web)
Clearly this depends on more than just the samba team, but are there plans to add NIS authentication (i.e., instead of or in addition to smbpasswd), NetInfo authentication, and/or smb.conf NetInfo or NIS databases?
32-bit architecture -- even if the numbers you're working with have the same number of bits, FP usually (I believe) calculates with extra bits and a 64-bit arch may have more extra bits
C vs. FORTRAN -- I can't say FORTRAN would be more accurate, but almost guaranteed it would be faster (that is, without f2c)
Dunno that it's a solution, but you may want to look at it.
First let me say this: I use Red Hat at work, my room mates use Debian, but there's always been a special place in my heart for Slackware. I stopped using it when I decided to upgrade my system wholesale (for glibc 2.0 betas, as I recall) and I didn't want to fuss with upgrading everything by hand (now I use OpenBSD and FreeBSD at home). If Slackware upgrades are now (or were with 3.4) quite safe and as easy as I describe, discount this:)
Have you considered looking at, e.g., OpenBSD's method for upgrading -- installing all the base packages except etc.tar.gz, which replaces everything in the tarballs but leaves/usr/local,/etc,/home,/root,/mnt, and possibly/var alone, and then manually upgrading/etc -- as a way to add better upgradability to Slackware?
I guess the only problem with this is that, as I recall, each individual package puts its own stuff in/etc. Would it be possible to add scripts in the packages so that they check for possible overwrites into protected directories and if so, install the new ones as e.g./etc/foo.conf.new -- and throw a message to the console saying either "new config file written, it's the same file as in the last distribution" or "new config file written, it's in a new format so you'll need to migrate before using this package at all"?
Slashdot Mirror
No, slow down there. It was had (directly) from PDF, which was used because Adobe didn't want them to use DisplayPS (which is apparently too old to be useful technology :). Apple got access to DisplayPS with an alpha channel from NeXT, but MOSX doesn't use DisplayPS.
MacOS X didn't get transparency from Berlin, they got it from PDF.
Whoops, I thought I mentioned that. It didn't bother me so I didn't bother getting a new CD...
Either way, it points out the reason why tracking FreeBSD-stable for 'QA' purposes -- as was recently suggested on "BSD Today" -- is pointless -- bugs happen even in releases.
Because it's in a seperate section, "Ask Slashdot." Only the 'big' Ask Slashdot questions get front-page status. All of them are available in the slashbox.
/. sections you're interested in, I recommend adding the slashbox for that section and removing the stories from your default page, so you see all of them precisely once on the page.
For
Well, first of all, the extension would probably have to implement pre-emptive multitasking in order to provide threads...
:)
However, classic MacOS uses extra processors differently, essentially (as I recall) letting software that knows about the extra processor(s) take advantage of them by offloading tasks. For algorithms which are not easily parallizable, or programmers that are lazy, this makes some amount of sense -- it allows you to essentially devote a single processor to a single process, and use the other CPU for inconsequentials like displaying the results of the calculations.
When the system was bought/built/ordered with a specific multiprocessor app in mind (like Photoshop), and a specific purpose, it can work OK. Of course, for end-to-end improvement and responsiveness, pre-emptive multitasking, threading, and SMP are all necessary.
Since I'm not too familiar with it, hopefully someone else will come along and correct/inform me
I thought it was about the same. My only problem was that the installer bombed out if you tried to add packages from within it, so you have to had packages afterwards.
:)
Now OpenBSD, that was easy to install. Seriously, I mean that -- just because it's secure doesn't mean it's hard to install. The only difficulty is in the fdisk/disklabel setup, which is relatively straightforward (although very different!) if you've used Linux fdisk before.
The biggest difference between Linux and BSD installs is that BSD doesn't install the kitchen sink by default (hint: if you ever try a commercial UNIX system, they don't either). OpenBSD is moving towards 'install all the sysadmin utilities a person could want, and then let the sysadmin install the user utilities' -- nc, ssh, a decent ftp client, etc. -- but it doesn't install any editor but vi
(a): MacOS X was demonstrated on an x86 box at WWDC (reportedly). Sounds pretty portable to me, if they don't even have a beta out, but they ported it to another platform. .
(b) device drivers are written in C++ using IOKit, and interface with BSD devices seperately -- much easier that straight BSD drivers.
(c) MacOS X makes use of Mach threads, which are available
First, it's a simple database file. Second, if NetInfo fails to load (or the db is corrupted), almost everything that uses it can default to a flat configuration file (I have a system here with system filesystems listed in /etc/fstab, and NFS user partitions in NetInfo).
Second, it's much more flexible than the registry, since remote servers can offer up NetInfo stuff as well -- no sick hacks to copy/remove registry entries from remote systems as people login or crap like that -- so there's better delineation between network-wide and local information.
But I was just curious. Sometimes people do pull jokes like that :)
So we all know CmdrTaco is horrible at spelling, my question is, is "Oddessey" a knowing and humorous play on that, or is it an error? Given that the correct (or accepted as correct in English) spelling would be odyssey, is there any plan to correct the spelling?
I'm genuinely curious... with some people, it's difficult to tell the intentional from the unintentional spelling errors.
Just because the information contained in the doc is a trade secret, doesn't mean that the doc itself has no protection. It's no different from grabbing a Word doc of Ender's Game and throwing it on the web, sans copyright information.
I would have sent this via email, but... no address.
I've bought two monitors from them, I've been very happy with both (although the screen needed some cleaning).
Take a look at http://www.deepspacetech.com/Hard ware/systems.htm. They offer a dual P90 with 32MB RAM, 10BaseT, and narrow SCSI for $65, and that should have plenty of serving power for a home network.
Then all you need is a better SCSI card (for Ultra2, LVD, and the like), your hard drives, and possibly an ethernet card (if you want 100BaseT). Everything on it's supported by FreeBSD and Linux.
Then your cost is, say, $75 for a decent SCSI-3 controller, $225 for a 20GB SCSI-3 drive, and $65 for the system. If you prefer to go with an IDE system (which is unfortunately how I'm using one at home), your costs are even less.
I know you said you don't want to turn a crappy box into a server (an opinion I strongly agree with, after having two PeeCee systems die doing 24x7 IPMasq), but these were originally intended to be servers and workstations (a system built when P90's were fast with 32MB of RAM!?), and it shows in the quality of the components.
No, I'm not affiliated with the guys at DeepSpace, but I've bought a lot of PC and NeXT gear from them and been extremely happy.
OS/2 (and possibly KDE or Gnome?) provides a very nice facility to store bookmarks in the filesystem. The name of the file is the 'bookmark' and the contents, in plain ASCII, is the URL. With lots of subdirectories this makes it very easy for me to locate bookmarks based on subject, and if that fails, any standard file searching tool (locate for bookmark names, grep for when you know part of the URL) works.
It also works across a LAN with shared filesystems, but for for access from anywhere on the 'Net, you'll have to use either one of these weird bookmark sites (sorry, I just can't get used to it), or else try your own LDAP server.
I'm not up on the technical details, SciTech being non-free software, but AFAIR they have an OS-specific driver which runs a small virtual machine, mapping system calls in a theoretical 'nice' OS to system calls on the current OS, and then a binary 'driver' which is loaded, and interacts.
May sound questionable (performance wise), but their drivers work in OS/2 and Windows at least (I remember reading that they also had a Linux version for X displays), and I've heard good things about both stability and performance.
1. Darwin and FreeBSD share kernel code (above the lowest level, where Darwin is Mach-based), from what I understand, and Darwin and NetBSD share userland code (i.e., I believe a lot of the userland was pulled over from NetBSD). So, it should be relatively easy to pull out userland from Darwin and try to run it on a NetBSD system. As for kernel code... I don't know. Yet.
;), and you can also have users that can only login to one machine. You can carry this up to as many levels as you want, and in large settings could be quite useful.
2. It's under the APSL.
Some of the Darwin code has already been brought over to other systems, notably (IMO) the NetInfo stuff.
For those that don't know, NetInfo is a configuration system like NIS, in that a particular system serves up all the machine-inspecific information (like hostnames, network printers, users, etc.). Unlike NIS, it provides for multi-level domains, so (most basically) a single server provides generic information to several computers, and then each computer provides through NetInfo information for itself.
It gets better, though, when you add more machines into the mix -- you can have users (like the CTO, maybe?) that can login to any machine, and then users that can login to any of the machines in a particular dept. (like if you have cubicles on a first-come, first-serve basis -- ick
Or Essential System Administration, from O'Reilly. Frankly, having looked through both, I think the O'Reilly book teaches concepts better.
However, I don't think it's necessarily appropriate as an "Introduction to Unix via Linux" course. I'm not sure what is.
My understanding of, for instance, Mac Services for Windows NT and UNIX Services for Windows NT is that it provides services from the same databases, just with different protocols.
So if you can see where this is going, is there any work on making samba able to make use of network-wide databases for user authentication, share specification (I know it can already use the autohome map, but more than that!), etc.?
In particular, I'm interested in things like:
- Being able to authenticate netatalk, samba, and UNIX users all the same way (i.e., not having smbpasswd, NIS, and
/etc/passwd all need to be updated every time a user changes his password or is added) - Being able to specify at the same time what my file server serve up, via netatalk, samba, and NFS (so I don't edit three configuration files every time I add a share, or move a share)
- Being able to specify from one system what each and every file server serves up, without having to connect to the machine in question and edit the smb.conf by hand (or by web)
Clearly this depends on more than just the samba team, but are there plans to add NIS authentication (i.e., instead of or in addition to smbpasswd), NetInfo authentication, and/or smb.conf NetInfo or NIS databases?- 32-bit architecture -- even if the numbers you're working with have the same number of bits, FP usually (I believe) calculates with extra bits and a 64-bit arch may have more extra bits
- C vs. FORTRAN -- I can't say FORTRAN would be more accurate, but almost guaranteed it would be faster (that is, without f2c)
Dunno that it's a solution, but you may want to look at it.First let me say this: I use Red Hat at work, my room mates use Debian, but there's always been a special place in my heart for Slackware. I stopped using it when I decided to upgrade my system wholesale (for glibc 2.0 betas, as I recall) and I didn't want to fuss with upgrading everything by hand (now I use OpenBSD and FreeBSD at home). If Slackware upgrades are now (or were with 3.4) quite safe and as easy as I describe, discount this :)
/usr/local, /etc, /home, /root, /mnt, and possibly /var alone, and then manually upgrading /etc -- as a way to add better upgradability to Slackware?
/etc. Would it be possible to add scripts in the packages so that they check for possible overwrites into protected directories and if so, install the new ones as e.g. /etc/foo.conf.new -- and throw a message to the console saying either "new config file written, it's the same file as in the last distribution" or "new config file written, it's in a new format so you'll need to migrate before using this package at all"?
Have you considered looking at, e.g., OpenBSD's method for upgrading -- installing all the base packages except etc.tar.gz, which replaces everything in the tarballs but leaves
I guess the only problem with this is that, as I recall, each individual package puts its own stuff in
They will learn by reading the README. It's not difficult in the least.
They're going to have to read the thing first to figure out how to report bugs, anyways.
Do you mean Red Hat is forgetting about script updates, and generally paying no attention to detail?
Who would have thought... that's why I like Slackware.