Don't know if this helps, but if you can access the apt file, you should
be able to use

$ sudo apt-get update
$ sudo apt-get dist-upgrade
$ sudo apt-get build-dep <package>

the previous command assumes that apt is already aware of WHAT the
package is, in other words, something with that name that has the list
of dependencies is already installed through apt.

At least on debian-based systems such as Ubuntu, Peppermint, LMDE or,
you know, actual Debian,

$ dpkg --build <tgz archive> <output-deb>
$ sudo dpkg --install <output-deb-from-previous>
$ sudo apt-get update

And if you have felt froggy enough to setup a lighttpd server as an apt
repository (a lot easier than it sounds, despite the utter lack of
documentation), then you can add it with, for example

$ sudo add-apt-repository 'deb https://dustgard.sx/apt/debian/

If you haven't done that yet, but want to, you can do so (via SSH into
the server in question)

$ sudo apt-get update
$ sudo apt-get dist-upgrade

If you do not already have a webserver installed, I highly recommend
using one you are familiar with. I personally prefer nginx, and for less
complicated things like this, lighttpd. Use what you are most
comfortable with. For most people, this is apache httpd.

$ sudo apt-get install <webserver>

Follow the instructions, man pages, and --help documentation, as well as
the docs and manuals at http://httpd.apache.org/ or http://nginx.org/ or
whichever, and configure your server so that it works.

$ sudo apt-get install reprepro

$ mkdir -p /home/www/dustgard.sx/htdocs/apt
$ mkdir -p /home/www/dustgard.sx/htdocs/apt/conf
$ mkdir -p /home/www/dustgard.sx/htdocs/apt/incoming

It should go without saying that placing a web server's document root in
/home is somewhat risky, even if you isolate the httpd's user in its own
user, in this case, www, unless you know how to restrict users to what
it needs to be able to access and nothing more. This is pretty much an
entire discipline by itself. If in doubt, ask a server admin friend to
do it for you, or risk your server getting hacked. Admins hate the idea
of letting their friends get hurt because they didn't do anything.
Admins, generally speaking, are helpful people. Let them help.

Any Linux server admin worth his salt is familiar with the principle of
least privilege. It's a security principle that says that users and
programs should have access to files, resources and other things it
needs to access to do what it needs to do, and nothing more. If a user
needs access to something new, such as a program or file that he or she
didn't need access to before, any server admin worth his salt will
always be available to add these access controls to your account. If you
have to wake him up at 2AM in the morning, his time, so be it. Such is
an admin's life.

Make sure to consult the documentation and manuals for reprepro and the
webserver you chose, double check it, and if there's anything you are
unclear about, ask your admin friends, and if they don't have the
specific information you want, and you have spent at least 1 hour trying
to find it, then join the relevant FreeNode or OFTC, and ask politely
for help as a question in a single line, and wait. This may take as long
as 12 hours in the case of the core developers being on the other side
of the planet, which is more likely than you'd think, if you are not
already very familiar with FOSS, or the programming and hacker community
in general already.

There are instructions on how to configure reprepro and import either
compiled source packages or prepackaged deb files to reprepro in the
required format, so I won't bloat this email further with unnecessary
redundancy, but it is entirely possible and highly recommended that
anyone capable of this should go ahead and do it.

It is also highly recommended to run a bug ticket system, such as Trac,
which is my favorite, and not just because it has a paw for its logo
(though that is a perk, admittedly).

I also highly recommend, for anyone maintaining nightly builds from
whatever git, svn or mercurial development code revision system is used
by the core developers, especially if the core developers are you, and
if nightly builds are not maintained, it is highly recommended that you
do, especially if the project's code is highly experimental or unstable
to begin with.

I'd highly recommend Jeeves, an automated build system you can configure
to run automatically every "night" which is after most of the core
developers have finished their coding and hackery for the day, and are
about to go to sleep. If something goes wrong with Jeeves itself, they
can worry about that when they wake up.

Jeeves allows you to automatically build source code on many different
architectures, and they are not necessarily (though possibly) virtual
machines on the same server. They can be remote devices that actually
have the target architectures.

If there is a project you are a huge fan of, a very helpful donation to
a blossoming project is to donate a new architecture to build with, or,
if they core developers are very overworked and cant do that themselves,
to buy a computer with the target architecture, work with them to
configure the computer so Jeeves can remotely login and automatically
compile and test the code before and after compilation, and report back
whether it works, passes tests, and seems to execute with the expected
responses, if applicable. This is extremely helpful for a hundred
different reasons, so the vast majority of coders will greatly
appreciate the hard work and time saved for them by doing this for your
idols.

Thank you for your time and patience reading this tutorial, and I hope
you find it helpful.

On 09/09/2013 07:03 AM, Carlos Ferreira wrote:
> Maybe this can help you
> Python3 for OpenWRT
>
> https://forum.openwrt.org/viewtopic.php?id=36958
>
>
> On 31 August 2013 08:59, Miss Valeska <[email protected]
> <mailto:[email protected]>> wrote:
>
>     Alright, So it has gcc, I tried installing a lot of other packages
>     based on what apt-file told me. I installed autoconf-archive, And
>     many other things. However, It STILL gives me the same error. Am I
>     doing something wrong? Can I just use the .deb version somehow?
>     Because that actually installs properly, I'd have to move or link
>     it's files to /opt/stackless somehow, And I'm not sure what would
>     need to be linked or moved.
>
>     I want to see if I can just do it with the installed version, If I
>     can, I'll just do that, And link/move the files, However, I will
>     need help in order to know what I need to move. Once that is done,
>     I can set up the script to do it all automatically, and no one
>     will ever have to bother with it again!
>

_______________________________________________
Netsukuku mailing list
[email protected]
https://mailinglists.dyne.org/cgi-bin/mailman/listinfo/netsukuku

Reply via email to