Reviews of Linux distributions: I've read a lot of them, and I've write some too, I've been thinking about how a good review should be, what it should include, and what it should aim to.
In the comments section of my friend's blog, started a nice discussion. In the review the writer stated that he has only installed the distribution under test in a virtual machine. One commenter said: "… To write a good review, you need to really install the distribution in you computer, and use it for some fair amount of time, not just install it in a virtual environment". Another commenter replied: "If the method is disclosed, it is OK to test in any possible environment".
I've been thinking about this discussion. Is it valid to just install a Linux distribution in a virtual machine, take a few screenshots of it and then write its review?. Or you really need to be some kind of expert in order to have the right to write a review.
The answer to that question is the same I gave to almost all subjective questions, It depends. Yes sir, it depends on what you are pursuing with your review and the audience you are writing for. Let's clarify this a little.
At least in my opinion, a review of something, being it a movie, a Linux distribution, or even a Hotel, the review should be written in such a way, that a reader can have a good idea of what he is going to find, once he decided to test the reviewed object by himself.
If the above statement is true, then what a Linux distribution review should contain?
It needs to have enough information for the reader, so that he can decide to test it or not. And that information, talking about Linux distributions, should be at least:
- The package included by default (I know you can then install almost anything)
- The versions of those packages
- Release cycle, so the user can have an idea about how frequent he will have to upgrade his system.
- Method of upgrades.
- Stability of the distribution.
- Philosophy of the maintainers. (If they are purist like in Debian, or they aim to make the distribution as easy as possible like Mint).
- Orientation of the distribution (server or desktop oriented).
- How easy is its administration.
- Method to install new software, and package manager tool.
Also the reviewer may need to compare it with one of the most known distributions, like Ubuntu, Debian or Fedora, not saying which one is better or worse, as usually, there is no such a thing as a "better or worse Linux distribution". No, the comparison should focus on things like:
"Gentoo, uses the system V init scripts, just like Debian, so if you are familiar with Debian you will find the same init scripts under Gentoo" (This is just an example).
Or maybe a better one: "In Arch Linux Apache is called httpd, while in Debian / Ubuntu it is called Apache2", And the like.
The reviewer needs to put some light over the community, is it newbie friendly or not? In some distributions the community expects more from the user before helping them.
Finally, if the distribution is Desktop oriented, there should be a focus on hardware support, and how to install non-free drivers and codecs. But if it is server oriented, there should be a focus on how stable it is, how easy or difficult an upgrade can be done, when a new release is out. What your options are if you need the latest version of a given software, like back ports was for Debian.
Of course with some distribution the reviewer can focus on both Desktop and server, this is the case for Debian, Ubuntu maybe Fedora and I'm sure there are others like these. Just to be fair here, I should said that any distribution can be used as server or desktop, and if the review is really a good one it should cover both aspects, just explaining how "not-easy" will be to have Slackware as your Desktop, or how "risky" is to have Arch Linux as your server.
Summarizing all the above: A Linux distribution review should have enough information, so that the reader after reading some of them, may decide if that distribution worth a test for his project or not.
Finally, and because I said above it depends, it is OK to make Lazy Linux distribution reviews, but in this case, clearly disclose your methodology, clearly state that you have not really tested the distribution, and better yet, do not call it a review, maybe something like "Linux Mint 11 screenshots and bird's eye view" is a better title if you do not have the time of the mood to actually go any deeper than just a simple USB, LiveCD or virtual machine test.