Re: [dev] [st] 0.1 Feedback - Was: A few small patches
I've had a pop at tidying it up a bit, I'm not getting as many silly characters filtered through, would you mind giving it a try with what you were doing and seeing if it's any better? Your new version of the patch seems to work *MUCH* better. In the couple tests I've run, the random line noise I was getting before is much improved. There's still some issues with curses / ncurses / slang applications (I've yet to identify which specifically), which leave some noise at the end of the lines, but - like I said - the amount isn't as bad. This patch could feasibly be useful. I think targeting URLs specifically would be too specific for the behaviour to be included in a terminal With regards to my ideas being too specific: The way I'm looking at doing this is to create a regular expression that matches the text present in the current buffer and then presents an indexed list of what you want to copy. If you want to have this work for an entire line, [.]?[\n]$ (or similar - my regex-fu is weak). You could also match a url, a name, a phone number. With proper planning, you could assign keys to different patterns and yank them all via separate keypresses. I'm planning on being flexible enough to get any data that one could feasibly want to yank, but specific enough that you get only what you want. I just need to find the time to implement it. Also - let me know how this email goes through. Gmail seems to be pretty stupid when it comes to email handling - specifically line width. I think this will come through as plain-text, wrapped @ 80 characters, but it's allowing me to input it at a flowed width, in a non-monospaced font, and is offering for me to 'switch to plain text'...
Re: [dev] [surf] [PATCHES] (1) GConf URL schema handlers (2) delete _SURF_GO xprop (3) close stdout sending XID
Security isn't kept. This seems like more of a prevention of accidental disclosure than real security. (And therefore pointless...?) As an example, with this patch applied, run the following: The output in the spy terminal is: _SURF_GO: not found _SURF_GO(UTF8_STRING) = 0x61, 0x73, 0x64, 0x66 == asdf _SURF_GO: not found You're absolutely right, thanks for noticing that flaw in my patch. Do you have any idea if it is possible to set up some write-only atom in X11 or we would need something else like unix socket, to prevent some other party spying what URLs are sent to *surf*? Cheers, -- Adam Strzelecki
Re: [dev] [ANN] sabotage 2011-04-09, a musl+busybox based distribution
Yo On 10/04/2011, at 15:53, Christian Neukirchen chneukirc...@gmail.com wrote: Hi, this is the first public release of sabotage, a distribution based on musl and busybox. Provided software is: 9base-6 binutils-2.21 busybox-1.18.4 curl-7.21.4 e2fsprogs-1.41.14 gawk-3.1.8 git-1.7.4 gmp-5.0.1 linux-2.6.38.2 m4-1.4.16 make-3.82 mpc-0.9 mpfr-3.0.1 musl-2011-04-09 openssh-5.8p1 openssl-1.0.0d perl-5.12.3 pkg-config-0.25 sed-4.2.1 syslinux-4.03 zlib-1.2.5 Looks cool. Did you had any problem compiling any of those packages against musl? Did you tried slpm? Everything is statically linked. Finally an usable statically linked distro to play with.. Will try asap. Kernel source is included in /src/linux, so you can rebuild the kernel--the shipped kernel should work in most virtual machines. (You need to include devtmpfs and have it automatically mounted.) I think the base system should contain kernel sauces. and probably i will not put git, perl or gmp. You can bootstrap your own build from the scripts at Cool :) will give it a try https://github.com/chneukirchen/sabotage or use a ready-to-boot disk image (186M, expands to 4GB sparse file, ~1GB filled) from As an idea.. If you add stuff to compile or develop i will suggest you to add tcc from git and Go. Both have a good codebase (simple and complete) to develop. About graphical support.. Im expecting to see the first release of wayland that should be public during this year. I think is the best alternative to x11 atm. It relies on modern hw. But its codebase is about 5000LOC. Not suckless, but at least smart enought and supported by all modern widget kits out there (multitouch support, opengl...) I would like to see somebody interested in swk. In order to have a decent suckless widget toolkit that works on sdl, x11 and other backends are welcome. http://xmw.de/mirror/sabotage/sabotage-2011-04-09.img.xz This image can be booted for example with: qemu-system-x86_64 -curses -hda sabotage.img -net nic,model=e1000 -net user Known issues: - first run of getty on tty1 doesn't work, use a different vc. - sshd won't let you log in, initgroups doesn't work yet. - man is broken, use nroff directly. Enjoy, -- Christian Neukirchen chneukirc...@gmail.com http://chneukirchen.org e58bfd50d56b6d4399679b9f55850a49c8219463 sabotage-2011-04-09.img.xz
Re: [dev] @bleidl, 26/03/11 19:41
You can use twitter from api. Bitlbee have a irc gateway for twitter too. And there's ttytter. Which is a cmdline client. Web sucks, and its not the only way to access it. It offers several stuff the irc doesnt like indexing contents, clear timelines (not noisy irc channels) and bulk dumps. So you can fetch data fom different clients and go back in time. You can probably do this with irc bots. But nobody did it yet. In fact IM can be replaced by irc. But still theres no irc client that can be used as an IM ( because of the user interface) this is the only reason why i use xmpp. It shouldnt be hard to write a simple ui to handle mail, im and twitter using the irc protocol. Which is imho pretty clean and nice. But you end up depending on bots who do the storage, indexing, security (ciphering) for you.. In fact moat irc clients are quite bloated and are more designed to follow the protocol specs than to focus on usability or most common use cases. A part from all the exposed above. The microblogging in general have a public nature that allows people from outside the network or not strictly following you to e able to read your messages at any time. For all those reasons you cant replace microblogging with irc. On 09/04/2011, at 21:10, Bjartur Thorlacius svartma...@gmail.com wrote: Seriously, do these sites provide anything that IRC doesn't? And no, groups have no use, and thus don't count. IRC is more distributed than StatusNet twitter-clones, and has way more implementations. Most of them are terrible backdoors that are best kept locked away, but some are almost usable. Really, I'd prefer an IRC bot/logger to these μblogging *websites*. HTML is a terribly overcomplicated overkill.
Re: [dev] [ANN] sabotage 2011-04-09, a musl+busybox based distribution
Is there a reason to not use dropbear in place of openssh?
Re: [dev] [ANN] sabotage 2011-04-09, a musl+busybox based distribution
On Mon, Apr 11, 2011 at 2:53 PM, pancake panc...@youterm.com wrote: About graphical support.. Im expecting to see the first release of wayland that should be public during this year. I think is the best alternative to x11 atm. It relies on modern hw. But its codebase is about 5000LOC. Not suckless, but at least smart enought and supported by all modern widget kits out there (multitouch support, opengl...) It also depends on Mesa. So you can add a few zeros to the sloccount. -- # Kurt H Maier
Re: [musl] Re: [dev] [ANN] sabotage 2011-04-09, a musl+busybox based distribution
pancake panc...@youterm.com writes: 9base-6 binutils-2.21 busybox-1.18.4 curl-7.21.4 e2fsprogs-1.41.14 gawk-3.1.8 git-1.7.4 gmp-5.0.1 linux-2.6.38.2 m4-1.4.16 make-3.82 mpc-0.9 mpfr-3.0.1 musl-2011-04-09 openssh-5.8p1 openssl-1.0.0d perl-5.12.3 pkg-config-0.25 sed-4.2.1 syslinux-4.03 zlib-1.2.5 Looks cool. Did you had any problem compiling any of those packages against musl? Yes :) Look into the build scripts to see the hacks. Did you tried slpm? No, I will provide OpenBSD-style file sets for the future, and leave more complex packaging to uses. You probably can use slpm then. Kernel source is included in /src/linux, so you can rebuild the kernel--the shipped kernel should work in most virtual machines. (You need to include devtmpfs and have it automatically mounted.) I think the base system should contain kernel sauces. and probably i will not put git, perl or gmp. The system so far contains (almost) everything too bootstrap itself again. About graphical support.. Im expecting to see the first release of wayland that should be public during this year. I think is the best alternative to x11 atm. It relies on modern hw. But its codebase is about 5000LOC. Not suckless, but at least smart enought and supported by all modern widget kits out there (multitouch support, opengl...) I have started working on Xorg, and I think I can at least get Kdrive to run. Not yet, though. -- Christian Neukirchen chneukirc...@gmail.com http://chneukirchen.org
[dev] Re: [ANN] sabotage 2011-04-09, a musl+busybox based distribution
Jacob Todd jaketodd...@gmail.com writes: Is there a reason to not use dropbear in place of openssh? Personal preference. Dropbear has been built with musl as well. -- Christian Neukirchen chneukirc...@gmail.com http://chneukirchen.org
Re: [musl] Re: [dev] [ANN] sabotage 2011-04-09, a musl+busybox based distribution
Slpm is probably much simpler than any ports system out there. It still needs some love..But it works for my use cases. I recommend you to take a look on it. :) I already packaged musl, tcc and other stuff in slpm Gnite. Tomorrow ill give a try to the hacks you talk about. Having x11 and gtk3 compiled statically can be a great benchmarking point to proove that statically compiled distros can beat dynamic ones. On 12/04/2011, at 0:25, Christian Neukirchen chneukirc...@gmail.com wrote: pancake panc...@youterm.com writes: 9base-6 binutils-2.21 busybox-1.18.4 curl-7.21.4 e2fsprogs-1.41.14 gawk-3.1.8 git-1.7.4 gmp-5.0.1 linux-2.6.38.2 m4-1.4.16 make-3.82 mpc-0.9 mpfr-3.0.1 musl-2011-04-09 openssh-5.8p1 openssl-1.0.0d perl-5.12.3 pkg-config-0.25 sed-4.2.1 syslinux-4.03 zlib-1.2.5 Looks cool. Did you had any problem compiling any of those packages against musl? Yes :) Look into the build scripts to see the hacks. Did you tried slpm? No, I will provide OpenBSD-style file sets for the future, and leave more complex packaging to uses. You probably can use slpm then. Kernel source is included in /src/linux, so you can rebuild the kernel--the shipped kernel should work in most virtual machines. (You need to include devtmpfs and have it automatically mounted.) I think the base system should contain kernel sauces. and probably i will not put git, perl or gmp. The system so far contains (almost) everything too bootstrap itself again. About graphical support.. Im expecting to see the first release of wayland that should be public during this year. I think is the best alternative to x11 atm. It relies on modern hw. But its codebase is about 5000LOC. Not suckless, but at least smart enought and supported by all modern widget kits out there (multitouch support, opengl...) I have started working on Xorg, and I think I can at least get Kdrive to run. Not yet, though. -- Christian Neukirchen chneukirc...@gmail.com http://chneukirchen.org
