- main - earth.li

Anyone in ALUG using Kylix ?
by bsamuels＠datamansys.co.uk 06 Jul '01

06 Jul '01

As per subject. Barry Samuels

2 1

Re: [Alug] Encrypting disks...
by MJ Ray 06 Jul '01

06 Jul '01

Andrew J Glover <flexifoil(a)btinternet.com> writes: > As to debian, Ive never used their kernel packages, and looking in > dselect, the kerneli patch in potato is for 2.2.14 but theres a bug in > < 2.2.19 isnt there? I'm not saying use the debian kernel packages, I'm saying build your own so that your system is still in a consistent state when you've done with it ;-) You can build your own packages of whatever kernel you so wish. Try "man make-kpkg" if you have the kernel-package package installed. I know there's a nasty bug in .16-18, but I don't know whether it's just performance and driver enhancements otherwise. -- MJR

2 2

RE:More C coding Woes
by Earl Brannigan 06 Jul '01

06 Jul '01

Hi Again : >Tried this, it didn't work, it transfered the errors to the linker not >the compiler. Which means it has compiled. Now, what exactly are you using ? is there a static/dynamic library or a .c file for it? Have you written it yourself ? Essentially (without seeing the linker error...please send this if still unresolved) this is probably because of unresolved function declared in the header file but not found when you link. in as plain English as I can Master : When a function prototype is encountered eg. void func(int param); //no function body the compiler generates a 'lookup' (an 'entry point' reference to a function) to be filled in later (when it actually generates(compiles) the function in another .c file). Your output file(s) will contain calls to functions (obtained via prototypes) which need to be resolved in order to make a whole executable. The Linker will resolve by either a) matching the calls to a function defined in another output file or b) through a library that it is aware of. Therefore you need either a .c file which defines the function to compile (to make an output file) and link together with your main.c's output file. Or your linker needs to know where to find the library which contains the pre-compiled function. Add the directory containing the library to your linker options. If you have a .c file add it to your project as a new node (if using an IDE) or make sure it is in your source directory (if using a makefile which processes all files within the project tree) If still stuck, or this makes absolutely no sense then email me over the stuff you have, and/or your linker error messages and I'll have a look. HTH Cheers Earl

1 0

Re: Arrrrggg! Mr Earl Brannigan Sir!
by David Howe 06 Jul '01

06 Jul '01

Jeez you guys are sooooooo bitchy on this mailing list.... >Neill has a serious problem expressing his dislike of certain >things...especially where programming is concerned. >To humour him I entertain his little expressions, Argggghhh being the >prime one. >Eventually when evolution has completed its course future generations of >Newmans will be able to use language much like the rest of us do today, >Until then Argggghhhh!!!!! is about as close as he can get to "I don't >like it that way..." >The speech therapist says he is making good progress and that next week >they will be studying two-syllabled words!! ...... so I thought I'd join in a bit...... (tongue firmly in cheek) Earl it is a moot point that your email domain is @lindenhouse - I was just wondering if this bore any relation to the Linden Centre at Broomfield Hospital in Essex where people go to "relax" a little and calm their nerves.... some people aren't very good at relaxing but eh nice people at the Linden Centre let them stay for ages free of charge - nice innit Dave

1 0

Re: [Alug] IPchains problem
by MJ Ray 06 Jul '01

06 Jul '01

"Martyn Drake" <martyn(a)alug.org.uk> writes: > ipchains -A input -j REJECT -p tcp -s 0.0.0.0/0 -d 0.0.0.0/0 3128 Maybe try using "-I input 1" instead of "-A input"? Pints of beers? You mixing your drinks again, Martyn? -- MJR

2 1

IPchains problem
by Martyn Drake 06 Jul '01

06 Jul '01

Okay dokey - my brain has switched off for the weekend, and this problemette is beginning to drive me nuts. I want to close a range of ports including 3128 (which is Squid) to the outside world. I would have thought that to do so would involve issuing the command:- ipchains -A input -j REJECT -p tcp -s 0.0.0.0/0 -d 0.0.0.0/0 3128 but it don't work. The machine itself is NATted through a Cisco 1600 series routed (the router itself is on a 192.168.1.x address). Does anybody have any suggestions on how to go about doing so? The current ipchains table has been reproduced below as a guide. Chain input (policy DENY): target prot opt source destination ports ACCEPT all ------ 192.168.0.0/24 anywhere n/a ACCEPT tcp ------ anywhere anywhere any -> 1025:65535 ACCEPT udp ------ anywhere anywhere any -> 1025:65535 ACCEPT tcp ------ anywhere anywhere any -> ssh ACCEPT icmp ------ anywhere anywhere any -> any ACCEPT all ------ 127.0.0.0/8 anywhere n/a ACCEPT tcp ------ anywhere anywhere any -> authChain forward (policy ACCEPT): target prot opt source destination ports MASQ all ------ 192.168.0.0/24 anywhere n/a Free pints of beers (note the plural) to anybody who manages to resolve this one :) Regards, Martyn

1 0

Re: [Alug] Encrypting disks...
by MJ Ray 06 Jul '01

06 Jul '01

Andrew J Glover <flexifoil(a)btinternet.com> writes: > This was after rebooting with the new kernel. The ciphers were built > into the kernel, and are there, dmesg: wrt the LO ones, are you compiling against the correct kernel-headers package? On Debian, you get this when you do make-kpkg, but you still have to remember to install it. -- MJR

2 1

Re: [Alug] Encrypting disks...
by MJ Ray 06 Jul '01

06 Jul '01

xsprite(a)bigfoot.com writes: > Something like this worked for me: > > # dd if=/dev/urandom of=/mycfs bs=1m count=$size This worked OK. > # modprobe $cipher > # losetup -e $cipher /dev/loop0 /mycfs > (prompts for key) > (add more layers here if needed - by incrementing the loop device and using > the previous loop device as the second parameter..) I failed here. I was trying to do /usr. Why are you looking at me funny? ;-) Remember kids, root, dd and your valuable file systems may not be a happy mix. Only go there if you really know what you're doing. -- MJR

2 1

Arrrrggg!
by Earl Brannigan 06 Jul '01

06 Jul '01

> > David Freeman wrote: > > > > > > > > > > > speaking of Earl... aarrgghhhhhh!!!!! > > > > > > ?????????????? Why Arrggg? Earl din't strike me as bing anymore > > strange > > > than the rest of us! > > > > Earl will understand ;).. > > Oh good, any chance you'll share the joke? > Hi Dave... Its No Joke... its very serious! Neill has a serious problem expressing his dislike of certain things...especially where programming is concerned. To humour him I entertain his little expressions, Argggghhh being the prime one. Eventually when evolution has completed its course future generations of Newmans will be able to use language much like the rest of us do today, Until then Argggghhhh!!!!! is about as close as he can get to "I don't like it that way..." The speech therapist says he is making good progress and that next week they will be studying two-syllabled words!! Cheers ;op Earl

2 1

Re: [Alug] Delaying tactics
by MJ Ray 06 Jul '01

06 Jul '01

Alexis Lee <alexis(a)turton.com> writes: > Lander. As I recall, Acorn did a bit of an Apple and created > something technically rather good, but then made a bit of a hash of > the marketing/sales. 10 times faster than the PC of its day. RISCOS was quite good, so its ideas have been cannibalised: see http://rox.sourceforge.net/ There's a Linux port of Zarch, the game Lander was cut down from, but it's binary-only against old libs, I think. Someone should email the author and ask them to release the source so that the buffer overflow root exploit against it can be closed... -- MJR

1 0