org-attach was the feature I needed.

Thank you.

On Mon, Jun 20, 2016 at 4:10 PM, Eric Abrahamsen <eric@ericabrahamsen.net> wrote:

Roland Everaert <reveatwork@gmail.com> writes:

> Hi,
>
> I am working as a sysadmin, In the organization, we use 2 tools to
> keep track of requests sent by the customers/users. As you can expect,
> those tools are not meant to be used to track all the gritty details
> of a sysadmin's job.
>
> So I am turning to org-mode (that I used for years) and its community
> to find a way to organize my job and being able to track what I have
> done and store the hundreds of lines of output from a command-line.
>
> I was wondering if it was possible to create directories and files
> with org-capture, based on data given interactively by the user.

You can use org-attach in conjunction with org-capture, to create a
directory connected to an Org heading.

> My intention would be to work this way:
>
> 1. Create a directory for a request or a group of tasks or a project,
> in short, an aspect of my job.
> 2. Create a file that will contain the information related to the
> request in addition to a journal allowing me to keep track of what I
> have done and store all the data that are useful to me.

I'd say you don't need a separate file for this, simply the subtree of
the heading you're using to track this job.

> 3. Store anything that is related to that request or aspect of my job
> into the related directory.

That's org-attach again.

> 4. Being able to search for a particular aspect or getting a list of
> them and access it.

I actually don't think there's any built-in way of searching files in an
org-attach directory.

> 5. When the job is done for an aspect, archive the directory.

I think that would happen automatically with org-attach.

> So far, I was using the configuration of norang, but I don't thing it
> is really adapted to my work-flow anymore.
>
> I know that org-mode is capable of a lot of things, but I was
> wondering if this is not a little bit to broad for org-mode to be an
> efficient tool. I was even thinking that all of this should be done
> through a server, with emacs being the interface to communicate with
> it.

Beats me!