Journalism 101: How not to review software

I was unfortunate enough to come across this "review" of KDE 4.2 (release candidate) today. Written by "Jacob W. B. ", it is perhaps the worst attempt at journalism I have seen in quite some time, and I'd like to explain why I think so.

First, a disclaimer: I like KDE. I prefer it to Gnome. I'm using KDE 4 quite happily at home and soon at work too. Even so, I feel entirely justified in the remarks I'm about to make.

Let's leave aside the deficient grammar, punctuation and coherence of the article and start examining the content:

we think that lately the KDE team version by version have been taking large steps in the wrong direction. They have been doing a good job at making a Desktop Environment that feels almost invisible, but the general design of the Environment itself seems a little too much like a Vista (i.e. Widgets) and Mac (i.e. the theme) clone and this isn't what GNU/Linux was originally designed to be, a clone.

Version by version? "Lately", there has only been one new major version of KDE: that is the KDE 4.x series. Its predecessor KDE 3.5 has been around years. The minor releases after KDE 4.0 - surprisingly, 4.1 and 4.2 (due soon) - haven't exactly been "large steps" but continuing on the path started by 4.0.

The first sentence, then, can be summarised as: the author doesn't like KDE4 and that is the approach they will be taking for the rest of the review. The second sentence seems to contradict itself: either the KDE team have been doing a good job with KDE4's design, or they haven't. And, incidentally, I'd say the theme was more Vista-ish; the "widgets" idea has been on Mac for years, and was adopted by Vista. What's novel in KDE4 is that a widget can live on the desktop or on the panel or both - it's a new take on an existing paradigm.

As for "what GNU/Linux was designed to be" - providing a choice, and providing an open-source alternative to proprietary software - I see no features of KDE that prevent this from being the case. Just because something's been used in proprietary software, does not make it a bad idea, and does not mean it should be avoided if it's a good idea.

Personally, I wasn't sold on the whole KDE4 theme until I used it. My first impressions were "it looks too much like Vista, and what's the point of the Plasmoids (widgets)?" Since the 4.0 launch, there are plenty more themes available, as well as some genuinely useful Plasmoids - from the usuals like the Wastebin and Pager applets to RSS feeds and the Folder View applet, of which I use three.

Back to the review, which continues:

"The second problem we have with KDE is the fact that they are somewhat breaking away from the easy usability and user friendly design of the GNU operating system...for example KDE 4.2 would not run on a computer with an 800MHz processor, less than 64MB of video, and 128MB of memory, where GNOME easily would."

I'm sorry - since when did "user friendly design" equal "low system requirements"? I think perhaps the author is conflating two quite different issues. If you're trying to run KDE4 on a machine with those specs then you're doing it wrong. Might I suggest that, say, Xfce would run much more smoothly on that hardware - because it's supposed to? If my Eee PC can (just about) run KDE4 then I see no reason why the system requirements should be a cause for concern.

Now we get to the real killer - the reviewer doesn't like the letter K. Indeed, the letter K is evil and therefore KDE must be bad! I mean, at least he doesn't drag in a mention of the Ku Klux Klan at this point, right?

Ah, crap, he did.

The "K Desktop Environment" when native programs starting with the letter 'K', are listed read "KKK" KKKKK ... they sometimes even designed artwork for the system that makes it look intentional to say "KKK", as seen in the image below. And we know that's not the case. It just seems as if they take pride in the letter 'K'... Though GNOME does the same thing i.e. "GGG", but this way is better because anything to avoid the "Ku Klux Klan" name similarity and reference.

Aside from being written in something barely approximating English, that such a substantial section of a software review would be taken up with such drivel is quite astonishing. Not only that, but it's accompanied with an image labelled "KKK KDE artwork - KDE 4.2" that's actually taken from an earlier version of KDE. The image is irrelevant - the prose even more so!

Now somewhat incredulous, I continued reading.

"And the KDE Desktop Environment has always made programs look more integrated with the Desktop Environment than as separate programs, and this kind of behavior is more confusing and takes away more from the usability than it adds. Additionally, the KDE team have designed their own replacements for programs, so that they integrate better, nice, but when there is something already out there, it would probably be better-- particularly for compatibility reasons --to use it instead, and just modify, add and remove features."

I'm sorry, but I honestly can't decide what point the author is trying to make here. A consistent "look & feel" is supposed to be confusing?Maybe he just doesn't want to use programs so obviously endorsed by radical pro-white groups? Come to think of it, don't Gnome applications aim for a consistent look across them all - shared icon sets and so on? I'd be surprised if they didn't.

As for "their own replacements for programs" - if you don't like a particular program, don't use it. That's the choice that Linux gives you. I'd always found the Kubuntu default package manager, Adept, more intuitive to use than Ubuntu's, Synaptic - mostly because of it's quick-search bar. In Ubuntu 8.10, the version of Adept that shipped is, honestly, pretty dire. So I switched to Synaptic (the Gnome default, remember), which now has a quick-search feature. So what? If you don't like a program don't use it, use something else. The author is clearly used to his Gnome applications and wouldn't want to switch. The "compatibility reasons" mentioned seem to only exist in the author's own opinion - fine, but not in a review of this sort.

"But over all, it's not our favorite Desktop Environment, mainly because we believe more strongly in the philosophy of the GNU project, and the stability."

"But"? Wow, didn't see that one coming. That's a huge shock. Now let's translate the reasoning: "We don't like things that use Qt, because at one point in the past it wasn't released under an open-source license and that makes it inferior." As for stability - remember, he's reviewing a Release Candidate, not a final release - it might have been nice to back up the suggestion of KDE instability with, you know, some evidence? Even something anecdotal would have been nice.

Then, halfway down the page, we find this.

"Here is our review of the new KDE 4.2 release candidate..."

So what have I just wasted my life reading? I thought you'd already started that. Or maybe, in your opinion, it's a good plan to prepend any review with a rant on just how much you're going to hate the thing you're about to review.

OK, then. So, what's the headline story you can bring me from KDE4?

"First, now when you start any file operation, the progress will be displayed in a notification "bubble" that appears above the panel. "

Well, that isn't exactly how I would open my review of KDE 4.2. Not exactly going for the "Ooh, that's quite nice" here, are we?

The next few paragraphs read like someone who's read the KDE press release once or twice, detailing a couple of the things that were looked at between 4.1 and 4.2. Then, another inexplicable comment, when talking about the Folder View plasmoid:

"Users who prefer the conventional desktop icon paradigm can get equivalent functionality in KDE 4.2 by setting the Desktop Activity Type to "Folder View" ... But this doesn't seem to reduce processor or memory usage."

Should it? I see no reason to expect that switching between the two would have much effect on performance at all. To add that throwaway comment at the end just seems like a poorly-veiled attempt to suggest that KDE4 is a resource hog. I would like to know why the author doesn't just come out and say it, rather than drop unsubtle hints like this. If the reviewer isn't being honest about the reviewee, what's the review worth?

"Lastly" - as if an exhaustive list had been given! - "Lastly, the 24 hour clock is still a big problem, there isn't an easy way to set it to 12 hour mode, if at all."
System Settings -> Regional & Language -> Time & Dates. If you were that stuck, Jacob, you could have used Google.

And that's the end of the review. I came away still not quite believing the drivel of the first half of the article, and wondering if the author had actually spent more than five minutes in KDE to write it or not. Sadly, I have no evidence to suggest anything but "not".

Well done for making it to the end of this post - I realise it's a little longer than most of my posts are. I was disappointed that the original article was the second hit on Google for "fedora 10 kde 4.2" - I can only hope that people read a few other reviews before deciding if they want to give KDE a try or not.

Speaking of which - I did promise a KDE review here at some point. I think I've just found my excuse to write one.