When I started developing SocNetV back in 2005 the term “social network” had a simple sociological meaning; any group of somehow “connected” actors, which might interest a sociologist to study and analyse in terms of their social properties and patterns (centralities, triads, etc). The actors can be of any type (humans, animals, organisations, companies etc) and so the cause of them being connected: working in the same place, belong to the same hive/group, mentioning each other, having commercial relations etc. Thus the main reason behind developing SocNetV was to create a simple “point and click” application that would enable the researcher to load his real-life gathered data and visualize/analyze the network properties. Or, in case the network was small enough, perhaps recreate it with some clicks on a canvas before analysing it. Back then it had never occurred to me that the same “social network” would be used now-days to describe online communities of million of users. As a matter of fact, I always thought that this strand of sociology is more meaningful in analysing relatively small groups rather than thousands of interconnected actors. Nevertheless, the term is used now this way too and this means that Social Network Analysis software such as SocNetV face a new problem: people not knowing anything about Social Network Analysis as a hybrid sociology-mathematics discipline think that all the software does is to “automatically” analyse any given online virtual community; who is connected to whom and how. Which is certainly not true.
Author: Dimitris Kalamaras (Page 3 of 13)
If you use Git for version controlling some projects of yours and you are a number-crunch fanboy, you might be interested to know which hours you (or some fellows in your team) are active the most. I mean which hours during the day you do your commits. This is onliner will do this for you. Run it from inside your project dir:
Virtual mailboxes is a clear choice if you want to setup your mail server with multiple domains and non-Linux accounts. Of course, there are other and easier ways to setup Postfix for hosting multiple domains.
The most simple is to host additional (shared) domains by adding them to the ‘mydestination’ variable in postfix’s main.cf. But that solution lacks distinction between canonical and hosted domains ([email protected] is delivered to the same unix account as [email protected]).
A better approach is through “virtual aliases”, which allow separate domains and email addresses. For that, you only need to configure two parameters: ‘virtual_alias_domains = domain1.com domain2.com’ and ‘virtual_alias_maps = hash:/etc/postfix/virtual’. You declare the list of your hosted domains in the first and alias each one to a local linux account with the second (i.e. in /etc/postfix/virtual add “[email protected] root” to have mails sent to domain1.com’s postmaster delivered to root). However, this is not an optimal solution if you plan to host more and more domains or if you just don’t want to give shell access to your server.
The really best solution is to use the Postfix virtual mailbox delivery agent. With this approach, every email address has its own virtual mailbox and you need to setup only one system account (to be the owner of all mailboxes). And each domain’s mailboxes are properly arranged in the file system under the home dir of the owner account:
In this how-to, I describe the process that I followed to setup virtual mailboxes in my postfix server for a pet project of mine.
VirtualBox is a great piece of virtualization software. It not only allows us to run and test different Operating Systems inside Virtual Machines (VMs) from the comfort of our main desktop computer with point-and-click, but it also enables us to setup and run VMs on remote headless servers from the command line. And we can control these VMs remotely via a plain RDP client. This is great because when you host your VMs on server hardware (usually without any graphical UI) you don’t really need the full fledged GUI of VirtualBox (and its Qt/SDL dependencies) nor to display the VM output locally. Instead you just need some commands to setup and start your VM. Then all you want is to connect to it remotely and install an OS. With VirtualBox, this is accomplished via the commands VBoxManage and VBoxHeadless. This is how it can be done.