AltME groups: search

It is missing important apps like CVS and SSH for instance, and I'm 
not a typical Linux user who can spend hours to find and download 
the GNU sources for an app, configure the build, and compile it.

 <--- don't know how Mandrake handles packages other than it's RPM, 
 but Debian handles these things in seconds with a single CLI command, 
 and it surely isn't necessary to hunt down sources manually. :-)

I always regarded Debian as a bit harder to start with but many times 
easier to maintain once it runs.

Mandrake isn't missing SSH AFAIK. I am using Mandrake and SSH regularly.

It depends on which Mandrake bundle you're using, and whether you 
opt for the default settings.

I've used Mandrake, and I have had the best luck with it over other 
distros.  urpmi worked snazzily for me.  I also use SSH constantly

I set it up here under VMWare for basic use (not a *nix guy :-) and 
it worked; SUSE was also OK; but RedHat solved a couple issues the 
others had under VMWare, which is really just a versioning/support 
issue in that environment. I had to find an old RedHat at that.

I'm sitting on a debian and Ubuntu box here. Fonts are a tad larger 
in Ubuntu than in debian. debian runs XFree86 and Ubuntu, I believe 
uses X.org.

and they are both X.org

well, it's the latest update. just switched out warty with hoary 
in the sources.list and so I run Hoary :-) I think I need to check 
the DPI on both machines. I don't have any warty machines.

http://hmkdesign.dk/debian.pngand http://hmkdesign.dk/ubuntu.png

my debian runs gnome 2.8.2 from experimental, but ubuntu runs 2.10 
and it's a lot more stable.

I had a number of problems with Warty, but it's just the first release, 
and it still compares favorably with the alternatives

No, Warty doesn't have it yet. Yes, there are nice improvements in 
Hoary, but I couldn't use the preview yet. I did add a number of 
repositories and applications, so I got halfway towards Hoary after 
all

No, font-size-problems. before the default was to large, and some 
things like h1 did not work.

So what's needed:

1. We need a way to continually parse the SSHD log-file, something 
like tial

2. We extract the IP address and add it to some firewall, to block 
it for some time
3. We need to remove the IP from the firewall

Anyone interested in such a project?

seems iptables can be changed on the fly from the command line. i 
guess you can switch things on and of, or just change them to what 
they are.

Yes, I dumped the rules and see the same IP several times.

there is a tailor.r in my folder on developer. it polls the tail 
of a file and calls a callback which each new line. uses /skip on 
file. may be a base for a daemon

I further investigated all this. Yes, the tables are used by the 
kernel (if told to do so but what I see it's mostly so by default) 
and can be changed dynamically.

I can run the script starting rebol and in console using do %script.r

have you tried using a shebang line at the top of the script and 
run the script directly as an executable script, rather than through 
the rebol executable?

...and the server is down

and which rebol version : my is: REBOL/Core 2.5.58.4.2

that rebol works here to. also when i start it from real console 
and logout. did you try really try the redirects of stdin and stderr?

I start a job and than use disown -h

than i exit and the process is still working

nice :-) works well and fast

But the flash client is now only 3kB and it's working in any browser 
not just in IE

and IOS is running within the VPS

Can anyone tell me where is rebol/cmd 2.5.125 on rebol.org/builds/ 
for linux??? On 15-Jun there are only Base, Pro and Core. But no 
command!

I see that all platform have since 6-jun Base, Pro, Core and Command, 
but linux is missing Commandµ

Carl told that the command build for linux was done on 15-Jun, with 
Pro, Base and Core, but I cannot find it!!

I'm currently trying to filter IP addresses that try to log into 
my SSH server. For this I use a tool "swatch" and "iptables" to filter 
them.

I see it, but it's related to background jobs, not at startup. I 
found some general documentation on init.d, but it's very convoluted 
and unclear....

and distro startup documentation is frequently convoluted and unclear. 
Good luck! ;-)

Dont know that eally well myself. what i firued out: usually (debian, 
suse) startup-scripts are in /etc/init.d. on debian there is a /etc/init.d/skeleton 
as base for own script. then there are the runlevels in /etc/rc?d/. 
links there go to /etc/init.d/ an tell the system what to start/stop. 
the numbers in the filenames are the priority, lowest run first. 
Usually there is a gui-tool like yast or ksysv which scans /etc/init.d/ 
for scripts and makes the appropriate links. tricky things are to 
write the startup-script, to figure out which runlevel to use and 
how the os figures out what to stop. although if you dont need that 
runlevel smartness, because you call it on boot and let it be killed 
by shutdown, you can just use a normal bash-script.

(runlevels tell which servers etc to run. so there is one with everything 
off for maintenance, one for console-use, console with network, console+network+gui-login, 
these with and without network, maybe some others)

Is there any reason that this group is private and not published 
to the web?  I know that there is a lot of good information in this 
group and everyone should have access to it.

I think there would be plenty of material for creating help docs 
about rebol driven daemons and services for either windows or linux 
servers, right? I'm going to run into this problem very soon and 
it would be nice to allow conform installation of rebol daemons this 
way. Has anyone worked on providing debian packages of rebol?

Festival is a software made in Scheme ( a Lisp clone) that allow 
voice synthésis (debian pakage are available on bedian.org website 
and can be acceded using Synapse GUI debian DLand Install software).

To run my solution I simply make a bash that tests if festival is 
yet running or not and then runs it (I'm using 1.2.1 REbol version 
for linux ...) and then runs my GUI software ;)

yes ;) But I allways fool my self betwin call and run functions

why the build use libstdc++ v5 and not libstdc++ v 6 ?

This is weird... rebol (1.2 and 1.3) can not connect to the internet 
with Fedora Core 4, but i have no problem with Debian 3.1... I think 
i will get rid off Fedora Core and keep Debian!!

thats what my link was about. bitdefender had the same problem, and:
SOLUTION:

Fedora Core 4 has gcc 4.0 that provides libstdc++.so.6 while BitDefender 
requires another version of this library, libstdc++.so.5. In order 
to satisfy this dependency, the compat-libstdc++-33-3.2.3-47.fc4.i386 
package must be installed. (you can find it on the distribution CDs)

Hi everyone,
I have a problem running the new view beta on suse 9.3.

I have libstdc++ 5, I can run rebol scripts (that need 1.3), but 
I can't start the desktop (I double-click the rebol icon, it just 
flashes -starts and stops immediatky-).
I'm new to Linux so this might be a known issue. 

I tried moving the VM to various places but same result at every 
try.
Maybe someone can help ?

debian is not packaged the same as ubuntu and ubuntu as some more 
configuration I intsalled and administrated debian for dec Alpha 
Workstation for nearbie 4 years ;)

so if I say to you that ubuntu is very good then it's vry good 

For example my mom was tired to get lot of problems with her windows 
in installed 2 month ago ubuntu  I use synapse to get some  missing 
packages and now all works well she is very happy with her pc ... 
My mom is 62 years old and dislike coputing but ubuntu have her preference 
...

They use Debian. I have Fedora at home. What would I like to ask 
is - where to put rebol interpreter on Debian? Where to put things 
like user.r or mysql-protocol.r or whatever other script I produce 
to them? The path is almost always trouble for me. I e.g. start console 
and I have to define set-net once again, as it does not use my user.r 
settings etc. Any insights?

All SDKs have shell, library and encryption. Command adds ODBC/Oracle 
and SSL (client side)

about special settings (user.r) i would write a short script dumping 
system/options and system/script. running from cgi. That should show 
the pathes.

call: i have 2.6.0.4.2 and call is there. IIRC thats the real release, 
not even beta.

when i upload things, i use a builder-script which takes my local 
installation and replaces some things, like the shebang. then i copy 
that result to the server. you could make a config-script, 
 exe: %/usr/local/bin/rebol
 data: %/svr/www/my-dir/
  ..
let them adjust that and run the builder.

Also it makes sense to not hard-wire most pathes, but put them in 
a config too and load them at runtime. but does not work with the 
shebang, it must be hardcoded in the script.

self updating java client. Some folks found it slow and eating too 
much CPU cycles, but no problem here at all ...

but that was reported some xyz versions before. They do regular updates, 
it is really a nice app, having auto-updates for itself and its plug-ins, 
simply a joy to use - that is how apps should work - install just 
once and then have it up-to-date for ages ...

I tried playing the presentation on Mandrake and all I got was the 
vid/no sound ? Which player are you guys using that will play a mp4 
?

tried mplayer and all I got was the vid/no sound-yet it does play 
a mp3 ?

i have a /usr/lib/win32 full of such codes. and there where one or 
two other dirs mentioned too. IIRC xine suggested to download the 
mplayer-codes too, althought i installed from distro.

(IMHO you should donwload Ubuntu, it will keep itself updated and 
it is easy to use - though gnome based)

problem with most Linux distro is that I have a dual cpu and most 
don't support dual cpus :(

about dual cpu: i guess thats wrong. usually distros have multiple 
kernels and you need one with smp in the name. should be as simple 
as telling the package-manager what you want. (in theorie, never 
tested smp, but on debian changed kernel for other reasons that way.

Does Cheyenne have an option to limit which network adapters it will 
serve to? For example, I have 5 adapters on this computer (ethernet, 
wireless, 2 VMware, Hamachi), maybe 6 counting localhost, and I would 
like the web server to only be visible to the virtual adapters, not 
even opening a port on the physical ones.


I remember this being difficult in REBOL, that REBOL would always 
open server ports on all available adapters, so you would have to 
inspect the incoming connections and filter. Have you come up with 
a better workaround for this, or a way to do it properly?

IIRC via set-modes/get-modes, you can know what adapter is involved. 
I am afraid there is no other way how to prevent this. Opening a 
socket is OS thing imo, and I wonder if C socket functions allow 
you to choose your device?

Can RSP pages retreive content using async http and return them to 
the client?

is is parse base, each rule consist of a match block!, a transform 
block! and a logic! to continue thru next rule or break

but at least under rest of the browsers it would looks better (and 
with overflow: scroll)

This is working for IE and FF and should be added to Carl's blog 
as well:

pre {font: bold 10pt "courier new", monospace, console; width: 630px; 
overflow: auto; background-color: #f0f0f0; padding: 16px; border: 
solid #c0c0c0 1px;}

The problem with the code above is, that it's fixed with for all 
pre tags, but in the blog there are at least 2 widths - one in the 
article and one in the comments

Doc,  I appreciate your time / work on this.  Just so I understand.... 
are you saying that the solution you would implement would only run 
on a Win machine and that solutions that run on other machines are 
not feasible  ... because they are basically cumbersome and unreliable? 
 e.g. http://drupal.org/node/44718And are schemes like NTLM Authorization 
Proxy Server unworkable?  refer http://www.geocities.com/rozmanov/ntlm/
    I can't see using Cheyenne in a work environment  unless it supports 
NTLM.

Terry: UniServe hasn't been rewritten, there are just a few features 
and new events added. The service API is exactly the same as before. 
Remember that Cheyenne and UniServe are two different products. Cheyenne 
is just using UniServe as low level networking layer.

Nenad, I dropped my service into Cheyenne's uniserve / service folder 
and no go? Works fine with the older Uniserve.

Erratum: cheyenne loads 2 services : HTTPd and Task-master.

I wonder if all of that can't be short circuited by just downloading 
binaries of openssl, creating the certificates, and then downloading 
the stunnel binaries.

I've been cranking hard on Framewerks, and a beta is coming soon. 
We've completely blurred the line between desktop apps, and web apps, 
and I'm looking to EASILY sell 100,000+ copies of Cheyenne as soon 
as it's ready.

Anyone interested in making Cheyenne (and Uniserve) an Apache killer?

Doc is rewritting sessions currently and back on Cheyenne, at least 
that is what he told to me a week ago ...

I sometimes escape to hunt for some food and drink some water, and...connect 
to AltMe ;-)

ah, not rpc, but rsp :-) well, i am not interested in anything rsp, 
unless it works with templates not mixing html and rebol source code 
:-)

and put a few pints in the letter  ;-)

Doc - all I need is my gfx man to work separately - that is way most 
templating systems don't suit me well. They are ugly mixture of code 
and presentation. I want only tags, and then parsing and replacing 
stuff.

pekr, remark is EXACTLY that but templates are not only for pages, 
but tags are templates... all code is outside of page and very easy 
to make your own tags.

Pekr: There's a lot of competing templating solutions, and AFAIK, 
XML+XSL is the most used one. You can also look at Enhydra XMLC here 
: http://www.enhydra.org/tech/xmlc/index.html(It's done with JSP, 
but the concept can be easily ported to any other language).

the only thing it needs is someone to adapt it for cgi use... I have 
too little CGI practice and no real need, atm... so its hard to put 
time on this... but its a rebol script, so its easy to adapt.  all 
it would need is to check if its been started as cgi and call a different 
startup, which only prints out one file.  for static pages (which 
CAN include dynamic conent like sql queries) then its a very good 
solution which already supports site magamenent and ftp xfer.

can I use http://sourceforge.net/projects/sslexplorerto set up a 
ssl vpn so that the remote user logins via the vpn, and then accesses 
the intranet's web server?

Oldes: thanks for the offer of help cgi mode with remark.  Anytime 
you want, just ping me and we can see what needs too be done to make 
it happen.

maxim: why do you think nobody's reading it? because your web's bandwith 
wasn't wasted in ten minutes and there are no ecstatic reactions? 
;)

hahaha, I don't expect to change the world... just that my announce 
didn't even generate one sign or having been read... so I wonder 
sometimes, if most of you use the custom dividers or are like me 
and just the "all groups" master didivers and clear them as they 
 are filled.

actually I've read it, but have no time right now to try it. I'm 
still looking for some web managment so I can rework my web, but 
the time...that's the problem :) I'm using just "all groups" and 
check red groups for anything interesting :)

that is probably true. however I may change the architecture a bit 
since the parser is limited to two levels and is not recursive. the 
pages are generated from a limited list. all I really wanted to do 
was to make it easy to combine blogger.r, makedoc documents, dir 
listings and customized html output in a pipeline. it does most of 
that now, but the parser is only about 80% elegant. :-)

I put 0% of time on site layout, once its done.  I just edit content 
and make the page.  its like editing make doc, but in html format... 
also, the tags are not standard html type tags  they nest directely 
 <p! text <b! is bold <i! and italic>>>

easier to edit, since I don't have to track the damned end tag all 
the time... this being said, normal html is still valid within the 
files... so its a 1/2 way between html and rebol code.

you install and setup an apache server which has ssl, reverse proxy, 
and redirection modules enabled.

then you accept the connections from the ssl port (apache silently 
handles the decryption) and redirect all traffic to your cheyenne 
server.

the setups where very basic though... 4-5 lines, basically setting 
up the ports, and urls ... not much wizardry.

yes... must agree with Doc. It's very easy, just download the latest 
precompiled version (I have 4.20) install, uncomment the https as 
doc said and than do:
      stunnel.exe -install
to install it as service.. and:
      stunnel.exe -start
to start it.

which is what I tried last year and could not get working.

Just download and install the precompiled version : http://www.stunnel.org/download/stunnel/win32/stunnel-4.20-installer.exe

Then STunnel => Run STunnel ...and that's all. Takes 2 mn, including 
the download time.

They did a good job with stunnel's installer, fast and easy to use.

Specifications are written, I should be able to implement it this 
week. Then I'll need at least a week to test and tune.

Sure, Cheyenne is the foundation tool for all our web projects (customer 
or internal projects). So I have to make as good (means here fast 
and reliable) as possible.