Dual head with two cards?
Is it technically possible to construct a dual head machine by using two video cards (say a PCI and AGP card)? Any tips where to look? -- William Burrow -- New Brunswick, Canada
Re: Galeon for testing?
On Sat, Jan 05, 2002 at 11:52:02AM -0500, Stan Brown wrote: I've been playing around withe Galeon browser on another OS and I decided I really like it. ... Anyone know where I can get a .deb? From a response to a message I posted a while back: Hmmm... Relevant sites, in /etc/apt/sources.list: deb http://galeon.euber.net/galeon sid/ deb http://http.us.debian.org/debian unstable main contrib non-free And the library that satisfies libz-dev is zlib1g-dev. Not sure where that comes from... HTH. -- William Burrow -- New Brunswick, Canada
Re: removal of XFree3
On Fri, Jan 04, 2002 at 02:05:26PM -0500, Rick Pasotto wrote: I have upgraded my system to use XFree-4 and have been sucessfully using it fora couple of weeks. Time to free up some disk space. What's the best way of finding the no-longer-needed version 3 packages? Is there still any need for xfstt? xfs? Probably something like: dpkg -l | grep ' 3\.3' | less Then hack away at the old X packages that remain. You might want to peruse the entire list to make sure (take out the grep). -- William Burrow -- New Brunswick, Canada GUIs normally make it simple to accomplish simple actions and impossible to accomplish complex actions. --Doug Gwyn (22/Jun/91 in comp.unix.wizards)
Re: [OT] Open Portable Document Format?
On Thu, Jan 03, 2002 at 12:46:00PM +0200, Sunil Thomas Thonikuzhiyil wrote: On Wed, Jan 02, 2002 at 06:29:31PM -0800, Karsten M. Self wrote: I'd strongly encourage you to use tools other than Adobe's proprietary products, fortunately there are many which can be used to produce PDFs. Can you point us a few ones (in debian) To produce pdfs, use pdflatex or create a PostScript file somehow (usually a standard output format for Unix programs) then use ghostscript to create pdf output from that. -- William Burrow -- New Brunswick, Canada
Signatures
On Wed, Jan 02, 2002 at 08:58:32PM -0800, Karsten M. Self wrote: stuff Peace. OK, so here is one of the few places that I get signed email fairly often, so it is the first time it is getting a real workout. For some reason, gnupg is unable to verify the signatures however (for example, Karsten's). I am beginning to wonder what it is I am doing wrong? Here is the result from mutt running gnupg on Karsten's sig: gpg: Signature made Thu Jan 3 00:58:31 2002 AST using DSA key ID 55F2B9B0 gpg: requesting key 55F2B9B0 from search.keyserver.net ... gpg: [fd 9]: read error: Connection reset by peer gpg: no valid OpenPGP data found. I have setup search.keyserver.net as the default keyserver to fetch keys, but it seems to not give a response. When I go to www.keyserver.net and manually enter the key ID, after a long pause I get a blank screen. So, what is being done wrong here? -- William Burrow -- New Brunswick, Canada
Re: how to reset the video mode?
On Wed, Jan 02, 2002 at 07:29:23PM -0800, Karsten M. Self wrote: on Mon, Dec 31, 2001 at 10:46:46PM -0500, Michael P. Soulier ([EMAIL PROTECTED]) wrote: Is there a way to reset the video after an occasion like this? In my experience, no, generally. Sometimes it's possible to restart X successfully, but consoles stay hosed. I generally reboot in this case. If you are very lucky, then SVGATextMode might be able to reset the consoles. I don't think this works with any modern cards, however. Wouldn't hurt to try, though. -- William Burrow -- New Brunswick, Canada
Re: Upgrade potato - woody
On Wed, Jan 02, 2002 at 01:35:39AM -0800, Brian Nelson wrote: Colin Watson [EMAIL PROTECTED] writes: What package did you end up needing to install? Was that all you had to do? Maybe xserver-xfree86, which is only Suggested: by xserver-common? I sent a reply, but I don't see it on the list. The package xfree-clients was missing. I don't see this mentioned as a bug for this package. Should I file a bug report or what? -- William Burrow -- New Brunswick, Canada
Re: Upgrade potato - woody
On Thu, Jan 03, 2002 at 11:41:04AM -0600, Colin Watson wrote: (xbase-clients, rather - your reply just went to me, sorry I didn't get round to responding.) Oops, hit the wrong key. I *think* the problem is in apt, not xbase-clients. If you had startx before the upgrade, you had xbase-clients - which means that something in the upgrade caused xbase-clients to be removed. A number of packages, including things like xfonts-base and xlibs, conflict with old versions of xbase-clients, but this should have caused xbase-clients to be upgraded rather than removed. You could be right. Next upgrade I will try the debugging output. However, it is unlikely I will be upgrading another Debian machine soon. Perhaps someone else on the list will try it, eventually. Should the bug system still get an informational note, perhaps? -- William Burrow -- New Brunswick, Canada
Re: My xfree died when upgrading some packages...
On Wed, Jan 02, 2002 at 04:14:11PM -0600, [EMAIL PROTECTED] wrote: login prompt (this is nice), but when i try to startx, heres what i get: camilux#:startx X:cannot stat /etc/X11/X (no such file or directory), aborting giving up The file /etc/X11/X is a symlink to your X server. If you know the name of the server that you use, just recreate the symlink again (use ln). If you are lost, do you happen to recall the name of the server when last configuring X? -- William Burrow -- New Brunswick, Canada
Re: Bring up ppp link on shell
On Wed, Jan 02, 2002 at 07:46:44PM -0500, Jeremy L. Gaddis wrote: In a shell script, the easiest way to find out if the link is up or down is to check for the presence of /var/run/ppp0.pid. I just put scripts in ip-up.d and ip-down.d that sent a one line message to a tty indicating the status of the link. You only need to switch to that tty to see the status of the link. e.g. echo Local $PPP_LOCAL connected to remote $PPP_REMOTE /dev/ttyX or echo Disconnecting PPP link on `date`. /dev/ttyX -- William Burrow -- New Brunswick, Canada
Upgrade potato - woody
It seems that after completing the upgrade from potato to woody, that X was completely unusable. After a few hours searching around for the appropriate package, downloading and installing, I got X working again. I am wondering if anybody has had this problem and filed a bug report on it yet? Debian is great for the install once and upgrade scheme, but something always seems to break on upgrade. Having no working GUI after an upgrade where one was installed previously seems like a pretty serious bug. -- William Burrow -- New Brunswick, Canada
Re: upgrading xfree86
On Tue, Jan 01, 2002 at 08:46:23PM -0600, Deva Seetharam wrote: i just wanted to see what would happen if i do a apt-get dist upgrade i get the message that xbase-clients, xdm, xf86setup etc will be removed. i dont understand what is going on. Could anyone pls. help? All the old packages are removed, and new ones are installed. I'd like to know how your upgrade goes, I had to install xbase-clients manually after the update and wonder if it is something I did. -- William Burrow -- New Brunswick, Canada
Re: Potato/Unstable Issues
On Fri, Dec 10, 1999 at 09:47:17PM +0100, Gergely Madarasz wrote: On Fri, 10 Dec 1999, Joey Hess wrote: Craig Coles wrote: 1: On a couple of boxes, while doing an update with dselect, I keep seeing the message Cannot find termcap: Can't find a valid termcap file at /usr/lib/perl5/5.005/Term/ReadLine.pm line 305 There is a ReadLine.pmm file there. Install libreadline-gnu-perl. This is a known and many-times reported perl bug. libterm-readline-gnu-perl actually I am now having the same problem, unfortunately installing libterm-readline... didn't fix the problem. Am I overlooking something, or is something else broken on my system? -- William Burrow -- New Brunswick, Canada Real programmers don't comment their code. It was hard to write, it should be hard to understand. potato: 9:46pm up 7 days, 21:59, 8 users, load average: 1.04, 1.07, 1.12 OpenBSD: 9:46PM up 54 days, 10:03, 2 users, load averages: 1.08, 1.08, 1.08
Upgrade to potato left Perl 5.004 behind??
I upgraded to potato, hoping to get access to all the fine stuff in the distro, but find I am stuck with Perl 5.004. Packages I want to install want 5.005. I want Perl 5.005. Why is it that apt left 5.004 behind? Is the reason logged somewhere? Is there something I missed that might do the upgrade? Do I have to ditch some packages that depend on 5.004? Why would dist-upgrade miss this? Thanks for info! -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 9:06pm up 21:17, 8 users, load average: 1.02, 1.03, 1.00
Re: Upgrade to potato left Perl 5.004 behind??
On Tue, Dec 28, 1999 at 09:01:32PM -0500, Brian Servis wrote: Perl 5.005 and Perl 5.004 are significantly different in their compatibility that it was decided to make two independent packages of the two that can coexist on the system. You can install the perl-5.005 I tried using dpkg to install perl 5.005, but it complained there was a conflict with the already installed 5.004 package. I am attempting to set up enough to get dselect to work, maybe it will do something. -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 1:24am up 1 day, 1:35, 8 users, load average: 1.00, 1.00, 1.00
Re: slink -- potatoe?
On Wed, Dec 29, 1999 at 11:40:16AM -0500, Brian Servis wrote: *- On 29 Dec, Robert L. Harris wrote about slink -- potatoe? Ok, I have a box on a 33.6K modem I want to up grade to potatoe, either before or after it goes final. Is therre an apt-get command I can run before I go to bed that will do it cleanly for me? Edit /etc/apt/sources.list to point to potato (no e unless you are Dan Quayle, =) ). apt-get update apt-get -d dist-upgrade (notice the -d option) ... The -d option tells apt-get to only download and not install the packages. With a modem it is possible that you will loose the connection and some files not get completely downloaded. Once all the The -d option does not seem to be required. apt-get will automatically detect that it has not downloaded all the packages and asks you if you want to continue downloading when you restart it. With my 28k8 modem, it took nearly 24 hours to download 195MB of updates. This depends on how chock-a-block you made your system. I have most of the development stuff installed, because that is an interest area for me. apt-get dist-upgrade (without the -d option) You have to sit by during this process and hand-hold apt-get through the process because it stops to ask you questions. This is somewhat annoying. Also, there are two or three changes to config files you must make manually, depending on the packages you have installed (modules, svgatext and one other I've forgotten to update, obviously). -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 12:54pm up 1 day, 13:05, 8 users, load average: 1.00, 1.14, 1.14
Re: slink -- potatoe?
On Wed, Dec 29, 1999 at 09:52:28AM -0700, Robert L. Harris wrote: When the final CD is released I buy em. Until then, I'd like to try to upgrade one of my boxes. The CD would really be nice to have, but you still don't need it. After the upgrade, you will still be stuck with Perl 5.004, and unable to install a number of packages including the previously mentioned debconf. The way I had to go was download all the perl 5.005 stuff in the interpreters and base directories from the unstable distro manually and get dselect to install them. Once it finally was convinced (I've avoided dselect like plague so far), I did another apt update and dist-upgrade which will install the Perl 5.005 dependent stuff. This is going on as I type. -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 2:01pm up 1 day, 14:12, 8 users, load average: 1.00, 1.00, 1.00
Re: Timeline for potato
On Thu, Dec 30, 1999 at 02:08:17AM +0800, Rainer Dorsch wrote: The timeline is great! But I think it should be continued. As experience shows, shortly after releasing potato (or even shortly after freezing potato), kernel 2.4 and XFree 4.0 will appear. Should there be a potato update in this case or is Debian seriously wanting to shorten release cycles? Not that anyone asked my opinion, but OpenBSD's six month cycles seem to be just about right. Of course, they have more control over what goes into their distro and when. -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 2:53pm up 1 day, 15:04, 8 users, load average: 1.00, 1.02, 1.03
Re: debian-user mangles mails
On Thu, Dec 30, 1999 at 03:16:00AM +0800, Kai Henningsen wrote: Note the second line. I assume something processing debian-user mails mishandles (that is, drops) continuation lines. Probably the same stuff that rearranges them (something that also should not happen). I noticed some multipart messages that had missing separators. I wasn't sure if this was a problem my end, so commented out all formail processing in my procmailrc. Noone has written a multipart message since, so I had no comment. Perhaps the list processor is indeed to blame. -- William Burrow -- New Brunswick, Canada Niklaus Wirth has lamented that, whereas Europeans pronounce his name correctly (Ni-klows Virt), Americans invariably mangle it into (Nick-les Worth). Which is to say that Europeans call him by name, but Americans call him by value. 5:22pm up 1 day, 17:33, 8 users, load average: 1.00, 1.00, 1.00
Re: Netscape printing:diagnosis needed
On Thu, Dec 16, 1999 at 06:57:16PM +, Howard Mann wrote: I have just configured my HP Laserjet 5L printer with magicfilterconfig, stipulating the ljet4 filter. When I print an ASCII file from KFM ( KDE file manager) using lpr, the result is fine. When I print an e-mail message in Netscape Messenger ( or a page in Navigator) , the printed fonts are huge and offset on the page. Changing the font sizes in the Edit...Preferences...Fonts menu item does not cure this ailment. The screen font size seems to not make one whit of a difference to Netscape. Save the print job to disk instead (click file instead of printer in the dialog box). Try feeding the result through ghostscript to your printer using various filters available to gs. See what works and use that instead of ljet4. -- William Burrow -- New Brunswick, Canada How the Internet explodes myths...
Re: cvs behind a firewall?
On Thu, Dec 16, 1999 at 04:32:14PM +, [EMAIL PROTECTED] wrote: I'm trying to access a cvs server from behind a firewall. I don't think I'm making it through our firewall. Can anyone give me any tips or pointers to how I can do this. I can't seem to track down the ports/protocols that cvs uses, so I don't know how to change our ipchains script. If you manage to get this to work, I'd like to know -- I've got a machine behind IP Masqerading (NAT) and cvs doesn't work properly. -- William Burrow -- New Brunswick, Canada
Re: running two webservers on the same machine
On Thu, Dec 16, 1999 at 01:50:06PM +1100, Marc-Adrian Napoli wrote: I work with Shao here and we've been asked to basically get zeus and apache running on the same machine, both on port 80 but both binded to different ip addresses... (That way we could have the functionality of both). 1. I'm not sure how to bind Zeus/Apache to certain IP's and The webserver is told which IP to listen to. Read the Apache docs, they spell it out more or less. 2. I thought that a port was physical for a computer.. ie. port 80 on a machine is port 80 for all the IP's on that machine. Our boss thinks differently... so we're not sure. :) The Apache docs suggest running multiple servers on port 80, one for each IP, if necessary. So it can be done, but the servers must be told to bind to a particular IP address, otherwise they will monitor for all IPs. Interesting! (Apache docs on: http://www.apache.org or installed on your machine if you are lucky.) -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: Summary: logout/halt/reboot as ordinary user, gnome logout button?
On Tue, Dec 14, 1999 at 07:58:17PM -0500, Joe Bouchard wrote: On Tue, Dec 14, 1999 at 10:01:22AM +0100, Svante Signell wrote: - I want to enable my family to use Linux instead of the other OS. Therefore it is important that they can start the computer, run it and shut down in a CONTROLLED way. Restart/shutdown are menu entries in the other OS!! Can you teach your kids to type sudo reboot? (I don't mean that sarcastically, I have kids too . . .) You may even be able to put it in a menu item. I believe Debian supports the ctrl-alt-del key sequence for shutdown. You might have to enable it in /etc/inittab. The line looks like: # What to do when CTRL-ALT-DEL is pressed. ca:12345:ctrlaltdel:/sbin/shutdown -t1 -a -r now -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: running two webservers on the same machine
On Wed, Dec 15, 1999 at 06:35:58AM -0800, Pann McCuaig wrote: On Wed, Dec 15, 1999 at 16:35, Shao Zhang wrote: I would like to know if it is possible: to run two different web servers on the same physical machine binded to two different ip addresses, both of them using port 80 Try it. (This is always my first answer to this sort of question.) It should fail. If so, you can probably set up a redirector in Apache that sends the request to the other web server on another port. Maybe squid can be setup to do something like this. -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: bogomips
On Thu, Dec 16, 1999 at 10:59:18AM +1300, Tim Nicholas wrote: Alberto Maurizi wrote: Is this normal? I mean, that so low bogomips value. model name : Pentium 60/66 cpu MHz : 59.999660 bogomips: 23.91 I've got an Intel P133. Approximating pro rata using your data gives 23.91 / 60 * 133 = 53. Actual reported value is 53.25 So I'd say it looks exactly right. So mine doesn't make much sence then does it??? this is my /proc/cpuinfo vendor_id : AuthenticAMD model name: AMD-K5(tm) Processor cpu MHz : 100.230957 bogomips : 199.88 It makes perfect sense, for an AMD-K5. Bogomips are as their name implies, bogus. Bogomips only make sense with respect to one particular CPU family. There is a bogomips explanation somewhere out there, in a FAQ somewhere else, too, no doubt. -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: running two webservers on the same machine
On Wed, 15 Dec 1999, Alvin Oga wrote: I have multiple web servers on the same machine... - same ip# for various servers... - different ip# for various web servers sales.foo.com rnd.foo.com www.foo.com www.bar.com www..com To be clear, the previous author was not talking about virtual hosts or virtual servers, he was talking about running two separate web server programs on the same port. One was Apache and the other Zeus, if I recall. This is different than setting up multiple domains in Apache. -- William Burrow -- New Brunswick, Canada o Copyright 1999 William Burrow ~ /\ ~ ()()
Re: Problem forwarding X over ssh
On Mon, Dec 13, 1999 at 01:49:55PM -0700, Gary Hennigan wrote: what network it was attached to. This local name doesn't necessarily correspond to its real name on the network and thus ssh was setting the display to this local name instead of the laptops real network name. Obviously this disallows a functional forward of X11. In my case I need to figure out how to set the host name to the real name of the laptop depending on what network it resides (maybe PCMCIA has some provision for this?). For now I can set it's host name manually via the hostname command to it's real network name and that fixes the problem. You might be able to get around it by generating a cookie for the host name as it is seen from the network: xauth generate outside.name:0 . Dunno if that will work or not. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: disabling remount ro on errors
On Tue, Dec 14, 1999 at 02:55:11PM -0600, [EMAIL PROTECTED] wrote: Did you run fsck on the filesystem? The cron job triggering the problem might be something like updatedb that touches everything on the disk. Fix the filesystem and the problem will go away. oh, yes. I fsck, and I fsck. This throws off bad sector errors. I'm convinced the underlying problem is hardware, both from the funny sounds the drive makes and the patterns in which the errors occur. Bad blocks would do it. Reformatting the partitions affected is one way to fix the problem (when you overwrite the disk block, the drive will substitute spare blocks instead of using the bad blocks again). I successfully installed onto a replacement drive, but for whatever reason, I can't get the network functioning--even after copying /etc and /lib from the old system (freebsd can't reach the network, either). It's a tulip card, and we have a single incoming ip with switches rather than real subnets (ie, I talk directly to *.*.1.1 as my router). Is the kernel and all its modules properly copied over? they should be. I did a clean install onto this disk. The old disk ran for months with the stock kernel. When it wouldn't talk to the network, I cp -R'd /etc and /lib from the old drive. A puzzler, something else might be involved. Does the network work with the old drive? -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: xserver for 3dlabs Permedia
On Tue, Dec 14, 1999 at 11:31:54AM -0500, David Teague wrote: One of my students has a 3dlabs permedia AGP 8Meg video card. He has managed to get Xwindows up with SVGA, but wants to better use the cards features. We found xf86 server for this card in Unstable. Use the 3DLabs server, included with Slink in main/binary/i386/x11. Am I right to assume that Potato binaries won't work directly with Slink? Seems so. Does he have to recompile from source to use this Potato Xserver with Slink? No, use the 3DLabs server. This is what I use. -- William Burrow -- New Brunswick, Canada How the Internet explodes myths... - the GOOD TIMES email virus hoax, brought to life courtesy Microsoft - MAKE MONEY FAST brought to reality on Wall Street by dot-coms and Linux
Re: Digital cameras and Linux
On Sun, Dec 12, 1999 at 06:19:25PM -0800, Wade Curry wrote: Good starter cameras ($300 range) check out the Olympus D-340R, HP C-200, Fuji MX-1200. All of the cameras in this range are a little slow between pics. They are 1MegaPixel cameras, so the image quality is just OK for web and 4X6 prints. The Olympus can save an uncompressed TIFF, but that means a 4-5MB file gets generated. Adds to the image quality, but you'll want more storage. All good points. People might also consider consulting the rec.photo.digital and rec.photo.film+labs newsgroups for info. The dejanews site might be handy for this. Also, prices are much lower on the Net than in the retail stores. Check out www.computers.com for price lists. If you buy it retail, you're nuts. Interesting point. :) -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: xntpd not functioning anymore...
On Mon, Dec 13, 1999 at 08:45:04AM -0200, Henrique M Holschuh wrote: Don't kill tcp or udp packets from/to the ntp service port, nor delay them. When in doubt, try ntpq -p host to ping the servers. This will work even if the ntp ports are blocked by ipchains. Be sure port 123 is accessable through ipchains. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Slackware to Debian
On Mon, Dec 13, 1999 at 12:46:49PM +0100, Petru NOTINGHER wrote: I have a machine running under an old version of Slackware, and I would like to change to Debian. Is there any possibility to do it without a complete re-installation ? That is the easy way. Just save /usr/local (Debian shouldn't touch it anyway). Unfortunately, /usr/lib gets filled up with loads of rubbish over time, but say goodbye to it anyway. Lots of it is outdated libc5 stuff. After a few weeks with Debian packages, most things will be back in order again. Personally, I created a /usr/local-libc5 directory to put all the old /usr/local stuff in. This way, I can migrate to newer glibc versions as I upgrade, or delete the old stuff in favour of newer Debian packages. BTW, I upgraded from a much customized Slack 2.1, how old is your system? :) -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Where is fromdos command pkg in slink or potato?
On Mon, Dec 13, 1999 at 02:47:49PM -0600, John Foster wrote: I have a working command called fromdos on my test system which is a box that has been upgraded from bo to hamm to slink to potato. It also uses the dos2unix symlink. Does anyone know where the current .deb package for this can be found. I am building a new server and this command is NECESSARY for it to do the intended job. There may be some other pkg that will do the trick but I have not found one. The purpose is to use it to convert database files from DOS/pc to Unix/Linux format. I tried copying the old file from my old box to the new one, but it does not work, it segfaults. Thanks! Not sure if I found it on my Debian system or not, but I cobbled this up for someone else... should work on most Unix systems: - #!/bin/sh # a script to strip carriage returns from DOS text files # WARNING: may modify and delete files, may cause data loss # Handle stdin if [ $# -eq 0 ] then tr -d '\r' fi for file in $* do if [ -f $file ] then tr -d '\r' $file $file.tmp \ ( cp $file.tmp $file; rm $file.tmp ) fi done - -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore Composed: 5:12pm
Re: disabling remount ro on errors
On Mon, Dec 13, 1999 at 12:06:12PM -0600, [EMAIL PROTECTED] wrote: OK, I know I'm playing with fire here, but I'm backed into a corner. ... Yesterday was the first sunday in a month that my hardrive didn't switch to ro due to errors on the weekly cron run. The remount as ro both stops incoming mail, and makes it impossible to telnet in to fix it. Did you run fsck on the filesystem? The cron job triggering the problem might be something like updatedb that touches everything on the disk. Fix the filesystem and the problem will go away. I successfully installed onto a replacement drive, but for whatever reason, I can't get the network functioning--even after copying /etc and /lib from the old system (freebsd can't reach the network, either). It's a tulip card, and we have a single incoming ip with switches rather than real subnets (ie, I talk directly to *.*.1.1 as my router). Is the kernel and all its modules properly copied over? If I edit out the remount=ro from /etc/fstab, and remove the weekly cron file, is my system likely to keep running? I know I'm playing with fire, but the remount is catastrophic. So is continuing with the errors. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Mail servers for large numbers of users
On Sun, Dec 12, 1999 at 05:11:54AM +0800, Ronald Tin wrote: How could I deliver mails to accounts that don't really exist? (I can't allocate 10 uids on a single machine, right?) I have only read the FAQ and anatomy for postfix Shall I play with the mailbox transport option for local delivery, or do I create local users with same uid (and disable their login) ? Do the 2nd solution really work? But still it doesn't seem very efficient to store 10 files in a single mail spool directory? ... The MTA won't give you any problems - they typically don't care about the password database. However, here's another reason why you don't want to But the MTA have to lookup the local user from the password database.. and mails are stored under owner's uid. That's the case for sendmail and exim at least. qmail can handle an unlimited number of users in an unlimited number of domains using the likes of vchkpw. Check out: http://www.qmail.org qmail is highly efficient and designed to be secure. It does have its quirks though -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Mail servers for large numbers of users
On Sat, Dec 11, 1999 at 06:27:58PM -0800, George Bonser wrote: On Sat, 11 Dec 1999, William Burrow wrote: qmail can handle an unlimited number of users in an unlimited number of domains using the likes of vchkpw. Check out: So can exim. The problem is not delivering email via smtp ... the problem is accessing it via IMAP. If you have 64K users, how do you set permissions so that one user can not deduce the path to another user's directory and potentially read their mail? Oops, missed the IMAP bit, was looking at POP3. Anyway, it sounds like some kind of problem with the IMAP server. Perhaps you can elaborate in private email the particular problem that an IMAP server has in this regard. Then it will become possible to search out a solution. The problem will be corrected soon with the raising of UIDs from 16 bits to 32 bits but until then, how do you make sure that each user can only access their own directory from anywhere on the planet if you have more users than UIDs? You design your server to separate the paths that users are permitted to access in a consistent, logical manner. The fact that a path exists to the user does not mean it maps directly to any shared path on the server. Think of virtual domains and web sites. UIDs are irrelevant. For reading email, I cannot see any reason to allow an external user to peruse the entire system directory structure. It doesn't make sense. Excuse my ignorance on how this service is implemented, but I can't see this being a problem in a properly designed system. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Debian 2.2
On Sun, Dec 12, 1999 at 02:02:52PM +1100, Damon Muller wrote: I think one issue is that, if you need to do a long, manual download, it might be worth waiting. I have potato running on my machine at work, and do an apt-get ugrade about once a week, and it's always a large download. I think, if you are planning on doing it by modem, you're going to have to do it all over again in a few months when it eventually goes stable. It's a long (though not terribly painful) process to do more than once. Is the upgrade like the install, full of annoying questions that stop the upgrade process until a response is entered? I find this aspect of Debian somewhat annoying. RedHat installs by asking questions before and after, but none during the install. Very quick feeling, as one does not have to monitor the process as it proceeds. ~/.gnome-desktop), it works fine. If you only want a few key packages updated, and you're willing to risk a few minor problems, just update the stuff you need and see how you go. I want to do the upgrade because of all the nifty packages available, that require new libraries. :( -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Mail servers for large numbers of users
On Sat, Dec 11, 1999 at 06:27:58PM -0800, George Bonser wrote: So can exim. The problem is not delivering email via smtp ... the problem is accessing it via IMAP. If you have 64K users, how do you set permissions so that one user can not deduce the path to another user's directory and potentially read their mail? Whew, I did some investigation on the issue. Looked at Courier-IMAP, an alternative IMAP server to the UW-IMAP server. It clearly explains in its documentation how you can setup thousands of virtual accounts without requiring any more than just *one* UID. Courier-IMAP has a link on the qmail web page, and seems to be written independent of any particular mailer, so long as Maildir format is supported (see safecat for adding Maildir format mailboxes to any mailer). In the meantime, the EASIEST method is to simply use more than one server for READING email and using something to auto-direct the user to the proper server to read their mail. The EASIEST method is to understand your IMAP server or get one that works properly. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Mail servers for large numbers of users
On Sat, Dec 11, 1999 at 07:21:49PM -0800, George Bonser wrote: On Sat, 11 Dec 1999, William Burrow wrote: You design your server to separate the paths that users are permitted to access in a consistent, logical manner. The fact that a path exists to the user does not mean it maps directly to any shared path on the server. Think of virtual domains and web sites. UIDs are irrelevant. For reading email, I cannot see any reason to allow an external user to peruse the entire system directory structure. It doesn't make sense. Excuse my ignorance on how this service is implemented, but I can't see this being a problem in a properly designed system. You can not equate http and imap in this manner. http serves files. You Why not. An email resides in a file. In a Maildir setup, exactly one message resides in one file. With Unix mailbox format, several messages exist in one file, big deal. can create a different path for each virtual domain. If I type in the full path for a different domain, I see a different site. Not a big deal. You can't get from http://www.virtualOne.com to http://www.virtualTwo.com from from http://www.virtualOne.com by typing paths (other than a link directly to virtualTwo's site). It is impossible. Same deal with IMAP servers. IMAP does not transfer mail to the user. It allows a remote user to access their mail on a local filesystem. That is the point of IMAP. You can check I am aware of the purpose of IMAP, I have not seen the implementation. The point is that other users on the system MUST be prevented from reading my mail files. This is done with ownership permissions on the directory. Do that with ONE UID. Courier-IMAP does this. All users must access their mail through IMAP. It makes sense. It works. It is the way it is done. You don't own that directory according to the database in the IMAP server, you aren't allowed to enter it. Your request is turned away for entering a bad path. Just like entering a path that doesn't exist for a particular virtual domain on a web server. Otherwise, I could log in as me but tell IMAP to use someone elses directory and read their mail. This is why it breaks when you have more users than you have bits to assign unique user ID's. You can't do this because the path you specify is not associated with your IMAP login ID. The database tells the server what is the acceptable base path. The path is logically constructed, so it is easy to tell apart illegal paths from legal paths. The scenario you present is all in your head. POP3 is no big deal, SMTP is no big deal. IMAP is a big deal because it is a direct read/write file access to a file that remains persistant on the server. Persistency is not a big deal, email can be made to persist on POP3 servers as well, it is just not often done (and seems to be often discouraged with small quotas). I am just a little bothered by statements that implementations of Internet services are broken and useless by design. Particularly when RTFM shows otherwise. I hope my outburst is not taken the wrong way. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: need help with modem
On Sun, Dec 12, 1999 at 11:11:05AM -0500, Joe Bouchard wrote: On Fri, Dec 10, 1999 at 04:19:59PM +, Robert Helmer wrote: You can change the permissions on the device ( /dev/ttyS0 or whatever ) make it writable by the dialout group, and put the users who can use the modem in the dialout group. Some of us use the dip group instead of dialout. What's the difference? dip used to be the program ages ago that you setup a SLIP or PPP dial up link with. SLIP fell into disuse and so did dip. I am supposing that you are still using some legacy group name related to the dip program. Makes no difference in the end. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: I have my own MTA, thank you very much
On Sun, Dec 12, 1999 at 01:33:41PM -0500, Russell Nelson wrote: How do I convince Debian that I have my own MTA, thank you very much, and it shouldn't attempt to install one of its own choosing? Is there a way to permanently fulfill a dependency by hand without installing a package? I'm not interested in having to tell dselect, dpkg, or apt to ignore dependencies. I want to tell them that I have fulfilled a dependency via non-package means, and never be asked about that dependency again. Install the equivs package. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: gateway.net
On Thu, Dec 09, 1999 at 03:46:21PM -0800, Rick Dunnivan wrote: I am trying to connect through gateway.net. When I get a Login: prompt, I type my name and then I get a Password: prompt. I type it and get **BAD PASSWD** as a response. I have already ensured I am using the Probably just a terminal server, not configured for general logins. The help desk is rather dumb about such things. correct password. When I contacted gateway, they said they do not support this operating system environment. Ask them if they support Windows 3.1. They will say words that can be mapped to something similar on Linux, you have to know what they are saying. PAP and CHAP are key words. Has anyone else out there run into this problem and if so is there a fix other than switching my ISP. Are you going to find one that supports Linux? Where? -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: [Fwd: C-Kermit 7.0 Beta.11 ready for testing]
On Thu, Dec 09, 1999 at 04:10:00PM -0500, Tom Allard wrote: [EMAIL PROTECTED] said: Is Debian going to include Kermit in its next iteration, or is it too late to include now? Notice Frank de Cruz' recent comment on licensing... It's still not DFSG compliant: According to the license: The source code may not be changed beyond any patches required for successful building and operation on the target platform. While this might make it palatable to something like RedHat, it doesn't make it Free (speech). Is this identical to the problem Qt has (had)? Surely, you can make a package for the non-free part now. Before, no package at all could be made. -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
[Fwd: C-Kermit 7.0 Beta.11 ready for testing]
Is Debian going to include Kermit in its next iteration, or is it too late to include now? Notice Frank de Cruz' recent comment on licensing... -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore ---BeginMessage--- C-Kermit 7.0 Beta 11 was announced yesterday; this should be the last Beta before the final release. As always, a fair number of builds were done for Linux platforms, including various Red Hat and Slackware versions. The announcement is on: comp.protocols.kermit.misc or you can read it here: ftp://kermit.columbia.edu/kermit/test/text/ck70b11.txt The C-Kermit 7.0 web page contains all the info and download links: http://www.columbia.edu/kermit/ck70.html This is your last chance to give C-Kermit 7.0 a spin and report any problems before it's too late. As you can see from the web page, we have a lot of Linux binaries but still not nearly enough. If anybody can make builds not listed on the web page (the binaries list is at the end), please let me know; I'll be glad to help. For example, Debian, Mandrake, SuSE, Caldera, Corel, plus just about anything on non-Intel hardware. Hopefully make linux will work for all Linux varieties (it looks around for libc/glibc differences, curses/ncurses, etc, and tries to adapt automatically). (Note that we're not at the install package stage yet -- we're mainly interested in getting compilation, linking, and runtime operation right before we begin with the packaging.) Also note the change in the license, which is specifically designed to allow C-Kermit to be included in Linux distributions. Thanks! Frank da Cruz The Kermit Project Columbia University 612 West 115th Street New York NY 10025-7799 USA Email: [EMAIL PROTECTED] ---End Message---
Re: startx as non-root
On Thu, Dec 09, 1999 at 03:24:38PM -0800, pplaw wrote: my startx only works as root. but i need to run it as non-root. (error message: you are not authorised to run the x server.) ... any suggestions? Try editing /etc/X11/Xserver -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: Newbie rudely asks for directional pointers
On Thu, Dec 09, 1999 at 01:35:49PM -0800, Darryl Röthering wrote: MVS!). I just got an old CD-less laptop loaded up with the basic Debian. Can anyone give me some profitable pointers for determining what packages I need to install under dselect? I think I should put in the admin and shell and vi and development stuff first, but am unsure what the best way to proceed is. First off, there is _no_ such thing as rudely asking for help. There is such a thing as rudely answering requests for help. Perhaps you've encountered a few of the latter. :) We may assume that this laptop has a relatively small drive? I would suggest picking one of the smaller default setups to install with. Then, peruse the package tree for packages you would like installed, as you need them. On the CD or FTP site, one of the useful directories is: /dists/stable/main/binary-i386 I also had a major problem getting the Debian linux installed via the floppy. I was told that this was due to an inverted DCL in the floppy drive, which required a different boot parameter. This took care of the problem, but I am wondering if I will continue to have this problem with each package I attempt to install. The reason why I wonder this, is that I don't seem to I have no idea what an inverted DCL is... VMS backwards? :) -- William Burrow -- New Brunswick, Canada A 'box' is something that accomplishes a task -- you feed in input and out comes the output, just as God and Larry Wall intended. -- brian moore
Re: apt-get
On Thu, Nov 11, 1999 at 11:07:17PM +0200, Martin Fluch wrote: My question is, when using apt-get to update the system, does it download all the files to local disk, then attempt to install them (not enough space on most home systems like mine) or does it work with one file at a time? first it downloads all files to /var/cache/apt/archives and then installs from there. It doesn't remove the downloaded files automaticaly, therefore an apt-get clean is needed. OK, so I guess I have to bite the bullet. How is the general opinion on potato at the moment? Generally, I don't like going bleeding edge, but if it is overall stable as it stands, then I might go for it. I'm not quite familiar with how Debian progresses in its development - -- Linux, because I'd like to *get there* today :) -- William Burrow -- New Brunswick, Canada Your education begins where what is called your education is over.
apt-get
I have yet to receive messages from this list, so I'm not sure if I am really on it or not. My question is, when using apt-get to update the system, does it download all the files to local disk, then attempt to install them (not enough space on most home systems like mine) or does it work with one file at a time? -- William Burrow -- New Brunswick, Canada Your education begins where what is called your education is over.
Re: Live filesystem on CDs
On Tue, 11 Feb 1997, Christoph wrote: On Tue, 11 Feb 1997, William Burrow wrote: aa126 Debian CDs seem to lack this feature. Even a small filesystem on the CD aa126 would be useful, and when the ext2 compressed filesystem is incorporated aa126 in the main kernel, a few more packages could be added for little more aa126 space on the CD. Can you come up with an idea which packages to install? We have much more than fit on a CD. Having such a live Cd would certainly be useful. Sometime within the next few weeks, I will consider constructing a CD with a live filesystem. The hardware I have to work with will likely only allow me to put only 400-500 megs onto the disk (it will depend on actual conditions). The disk will be constructed to be fairly usable, allowing demos of graphics programs (GIMP and/or ImageMagick for example), Internet (server and client stuff), development stuff, X and so on. The idea being that one can actually use it in place if desired, or just demo Linux to someone who does not want their harddrive touched. If anyone else has already constructed such a CD, I'd be interested in more info. Anyone interested in such a CD might want to drop me a line with comments. No guarantees, but once it is done, I might post the image on my web page, GPL like. -- William Burrow, VE9WIL (Adv, 5wpm) -- New Brunswick, Canada Copyright 1997 William Burrow -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Shell scripts (first line -- what is that called?)
On Wed, 27 Nov 1996, Johnie Ingram wrote: This is the first 2.1 kernel I've tried, so I don't know if its a bug or a feature. (Its startup messages mentioned being POSIX-certified, mabye things like #!/usr/bin/perl -w and #!/usr/bin/make -f are unsupported now?) Is this not the realm of the shell? I know that the kernel looks at the start of an executable to support Java, but it should not be changing it. -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow This line left intentionally blank. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Posting rejected - please read and agree to mailing list rules. (fwd)
On Mon, 25 Nov 1996, Tom Julien wrote: IMHO, a license like Qt's is long overdue. It makes a fine commercial product available to both X11 and Win32, yet it provides a great mechanism to promote freeware/open standards like Unix/X11 *over* propriety ones. Troll's reasoning for not allowing modified versions may not include this rationale, but I am certainly tickled pink to see it for this very reason. Yes, make the commercials pay when it takes commercial development to create something. Seems fair to me. Please work with Troll on this -- you may find that there's no need to argue at all. Y'know, maybe it does not really matter all that much right now. Slackware and RedHat may be more than happy to include KDE and Qt bundled and ready to go. This is a substantial chunk of the market, and surely captures the market KDE is aimed for: newcomers. In time, the license may be changed to suit Debian, and so Debian can join in the fun. What I would like to know is the Debian mission statement, its meaning and purpose. Something is lurking in the depths of my memory, but it is not at all clear. Is there a Debian FAQ or other info available? I'll check if a web site is around, if I ever get around to it. :) -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow This line left intentionally blank. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: VFAT32 support?
On Tue, 26 Nov 1996, Shaya Potter wrote: On Wed, 27 Nov 1996, Lawrence Chim wrote: Is linux kernel vfat support compatible with the vfat32 from the microsoft recent OEM WIN95 release? I would strongly doubt it. vfat used to be just a superset of the fat filesystem so it was too difficult to add support. vfat32, on the otherhand, is a big jump considering the fact that most dos utilities won't work with it either. No, vfat does not give FAT32 capability. It is a superset of the FAT filesystem (having nothing to do with the FAT itself really: it fiddles the directory entries). FAT32 is being worked on, I did see a message to this effect in c.o.l.dev. -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow This line left intentionally blank. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Please do not use Qt (fwd)
On Sun, 24 Nov 1996, Herbert Xu wrote: This is not the point. They can release it under any license as long as they allow other people to release modified versions, even if it has to be under a different name. This is so that the applications won't get caught when Troll stops releasing free Qt versions. Think of what would happen if you weren't allowed to release modified versions of X and that company which has taken over X decided to make X non-free. Well, you see, once the source code has been released, it is somewhat like Pandora's box. Programs relying on the old Qt will function as always. Free programs relying on the new, unreleased, pay-for-it Qt... well, there just won't be many of those around. As for not being able modify Qt, not really. Qt supports inheritance, so can be modified at a higher abstraction. Your argument is based on FUD. Meanwhile the freeware community diddles around with various uncompleted toolkits. flame Personally, I expected much more from the X development model than has appeared. The X world seems to have grasped a commercial library, Motif. Where is the plethora of excellent concepts and toolkits? Seems the big commercial OSes get those (MacOS, Windows, OS/2 -- not neglecting NeXT, it just didn't get to the big status). So, what went wrong? /flame I don't support Qt as much as I support a free toolkit that is being developed on a continual basis. It would not be a first to put a wrapper around Qt calls, if you so desired, either. -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow This line left intentionally blank. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Please do not use Qt (fwd)
On Sun, 24 Nov 1996, Herbert Xu wrote: like Pandora's box. Programs relying on the old Qt will function as always. Free programs relying on the new, unreleased, pay-for-it Qt... well, there just won't be many of those around. If there were no bugs in that free version. I already said, irrelevant. As for not being able modify Qt, not really. Qt supports inheritance, so can be modified at a higher abstraction. Your argument is based on FUD. Meanwhile the freeware community diddles around with various uncompleted toolkits. Again, this assumes the parent classes are bug free. But remember that Why. And as far as I am concerned there is no point in using Qt for developping kde when there are other free alternatives around, such as V. At last, an argument that might actually have substance. flame Where is the plethora of excellent concepts and toolkits? Seems the big commercial OSes get those (MacOS, Windows, OS/2 -- not neglecting NeXT, it just didn't get to the big status). So, what went wrong? /flame There is a plethora of toolkits; concepts? Don't know. The word excellent is missing from your sentence. I don't mind individual programs such as nethack using Qt. What I dislike is people putting in a lot of time and effort developping things like kde on top of a non-free toolkit like Qt when there are viable free alternatives around. Your opinion. Personally, I'd like KDE to be as library independent as possible, however, not much chance of that (at least until somebody assembles a wrapper library). Debian GNU/Linux 1.1 is out! { http://www.debian.org/ } Look, I'm not trying to get Qt included in Debian, Debian is the work of its developers. If it does not contain something, that something can be added later by the sysadmin. -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow This line left intentionally blank. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Please do not use Qt (fwd)
On Fri, 22 Nov 1996, Bruce Perens wrote: Yes. However, we still might look askance at Qt due to the other licensing terms, which are more restrictive than the GPL, especially since V (another C++ GUI) is under the GPL. Please understand I'm not making a technical criticisim. I just wish they would GPL the darned thing and leave it at that. They may be afraid of the situation that some other authors run into with freely distributable packages. For example, Knuth's TeX package had some of the fonts changed by someone other than the author. While the Knuth does not care if the fonts are modified, he really cares that the modified fonts were of the same name as his fonts. The result of the changed fonts were documents that looked different on different installations of TeX (depending on whether the site got the original fonts or not). This is very undesirable from the author's point of view, as he put considerable effort into ensuring that documents produced with TeX look the same regardless of the hardware used (within physical limits). While the Qt authors may have different concerns than Knuth does over TeX, the idea may be the same: modified versions may reflect badly on Troll Tech. -- William Burrow -- Fredericton Area Network, New Brunswick, Canada Copyright 1996 William Burrow The above is pure speculation. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]