FEB
25
2004

KDE Branding

The following post is mostly a response to Aaron's comments in his latest blog entry. It's basically about the reason why I like KDE and why I do the contributions I do, so it might be an interesting albeit highly theoretical read.



I think Aaron is approaching the whole issue from a very traditional (that is imo outdated) angle. While I understand what Aaron means I think many of his differentiations are artificial and, while they may reflect their commercial usage today, do not really fit to KDE but rather reflect the classical route how commercial branding is done.



I'm talking about computers as a commoditized tool where software is ultimately a tool offering specialized features adapted to the very specific situation in which a computer is used. Defaults do not matter here, the flexibility to adapt to any specific situation is what is important. Wrt graphical environments KDE is currently imo an unrivaled unified solution.



Vendors are users as well for me, they are service companies ideally offering specialized desktop configurations using KDE adapted to the very specific situation for which a given consumer want to use a computer. They, just like any user, need to be educated about the possibilities there are in KDE. Defaults do not matter here, the flexibility to adapt to any specific situation is what is important.



Branding is for me a mix of public awareness combined with (often pseudo-)education to make a brand stick in people's head as worthy to remember and consider or even superior looking compared to anything else less "branded". KDE imo does not need to resort to pseudo-education (a.k.a. propaganda imo), it already has much to offer that we can and need to educate people (and companies) about.



KDE as a project should not try to limit itself to a specific audience, may it be new, intermediate or expert users (which are all highly unspecific audiences which still differ highly depending on technical and cultural background anyway etc.) or even service vendors which resell service solutions based on KDE. Defaults are nice for showcasing features, but don't hold your breath for a single set of defaults to be magically able to suit all audiences and specific situations possible, this is not going to happen. Here again flexibility is the key, and education about the flexibility empowers users to make use of computer in new ways (which again opens room for commercial services for others where there is a market demand).



Defaults are ultimately nothing but a limited showcase of features. KDE does not have a specific audience to stick to, therefore it should not stick to specialized defaults either. KPersonalizer already shows a start of how KDE can quickly adapted to different special cases depending on a users personal expectations and needs. So instead tweaking controversial defaults future work would ideally be done to

  • Make specific features easier to edit for advanced users (think eg. right click menus)
  • Educate users and companies more than ever about the possibilities KDE offers, and encourage them to make active use of them
  • Extend KPersonalizer to offer more specialized adaptions (I'm personally considering to create a locked down KDE setting scheme mimicking the feature range of GNOME)

Please note that while I consider usability as very important I didn't mention it so far due to the fact that usability is not a universal feature but depends on the actual use cases. Something which is highly useful in one case might be complete overkill in another, and vice versa. So usability improvements would mostly be needed within and introduced to specific schemes available from KPersonalizer. (This of course doesn't exclude the possibility that there actually are areas which need usability improvements regardless of the use cases, but those will need to be solved by consent finding and not by adapting a special use case solution for all other use cases.)



Thanks for taking the time reading this all. =)

Comments

actually, i don't think you understand what i'm talking about, unfortunately. you're missign the boat completely by coming back time and again to configurability. here's the simple way of looking at it:

KDE has to ship with a certain set of defaults, by definition. KPersonalizer does not change that, we simply offer a choice of default settings, but even between those choices there are consistencies.

Every OS vendor that ships KDE also has to ship, by definition, KDE in a certain default set up.

Unfortunately right now most every OS vendor decides to ship with defaults that are different than the others. Not better, just different.

Imagine if we had 5 versions of KDE tarballs all configured completely differently and randomly cycled them as different people downloaded them. That's essentially what happens now. It's not good for the users (consistency) and it's not good for public awareness of this thing called "KDE" (which is what is called a "brand"). realize that this isn't just about widget sets (something i'm not really overly concerned with) but things like using the same K Menu layout and editting mechanism, to name just ONE example of customized functionality. KControl is another one, as more than one distribution "works around" it's brokenness. or will KPersonalizer install one of 4 different control center apps in the future? pfft.

and please stop thinking of "Nike" et al when i say "brand". think of the _actual_ meaning of the word and realize how we are completely missing that boat and how, at least in North America, that hurts us immensely.

perhaps many don't care because they are from a different cultural perspective where such things don't matter and people truly don't care that every desktop works differently depending on where you downloaded it from, and that's cool. but if so, don't tell me that because of your culture's biases that my culture's needs are obviously wrong.


By Aaron J. Seigo at Thu, 02/26/2004 - 00:13

Ok, then I completely missed the board... I still don't understand why you are putting such importance in inflexible defaults when it's apparent that a certain other project only got there by rigorously enforcing the view of a very specific kind of user while pretty much abandoning the rest and that KDE's strength is not in the (non-adapted) defaults anyway. Last time I checked packages are usually packaged with different distributions in mind and should not be mixed anyway. And regarding the "actual meaning of brand being missed, and that hurting KDE in North America" I would assume stuff like the recent Ars Technica article would be a step in the right direction, that is until you dragged everything into doubt. After a rather sleepless night I decided not to care since I can't make sense of it anyway and wish you good luck with whatever flies your boat.


By Na at Thu, 02/26/2004 - 13:45

"inflexible defaults": defaults are not inflexible, they are simply how things begin. we already have defaults. we are already flexible. unfortunately that flexibility has led to different groups who distribute KDE to have completely different defaults, including basic technologies such as application menu systems!

"a certain other project got there by rigorously enforcing the view of a very specific kind of user": why is it that whenever i speak of usability or branding, some people get concerned that i'm envisioning GNOME's approaches to things? if i thought GNOME was on the right track, i'd go join that project. obviously, i haven't. as a side note, GNOME hasn't achieved a level of effective branding either: Bluecurve, Ximian Desktop and Java Desktop all look, behave and feel different and none even mentions GNOME in their name. each of the three have a unique corporate-driven "brand", and in a way that's what i'm advocating against. in fact, i think that form of branding (which is more "traditional") hurts the source project more than it helps it.

"packages aer usually packaged with different distributions in mind and should not be mixed anyway": that's only half-true. people will and do come into contact with KDE-Mandrake in one place then KDE-SUSE in another and KDE-Red Hat in yet another (use any distro you want). while they may not mix and match packages (today) between OSes, people definitely do get exposed to more than one set of KDE packages. with nothing to counteract that, such as a "go to company" that "represents" KDE, we have achieved a rather nebulous market awareness.

and in the future it would be nice if people could mix and match packages from different OSes. in fact, i've done it myself between Red Hat and SUSE and it's rather nice. the future is compatibility, and it would be nice if KDE was a leader there.

Ars Technica article: that thing rocks! i loved it, and i think it's a brilliant addition to KDE's PR efforts! making KDE known as a brand does not obviate such articles, nor does all PR have to be rigidly lock-step with some greater vision. because that isn't the point. i'm not sure what makes you think that the great Ars Technica article and branding are at odds with each other. could you expand on that for me?

the point is to get people to be aware of and recognize the name and software that is KDE, to associate it with reliability and quality.


By Aaron J. Seigo at Thu, 02/26/2004 - 15:18

To be honest, I can't really say I'm sure what either of you are talking about, but I'll try and make some sense of it, and add my two cents as well.

Regarding defaults, I thought that it was always the policy of KDE to have "sensible defaults", which would obviously change when sensible does. For example, I have my wife use KDE, and when I upgraded the system to 3.2, the VERY FIRST thing she noticed was the bouncing feedback cursor (she loved it, BTW). In this instance, when something more 'sensible' came along, it was made the default. Users could still change it back, but it this case it was a smart change.

As far as the Konsole button goes, I rather miss it myself, but I can see the justification for its removal. My wife would practically die if she somehow ended up at a command prompt, after all. ;-)

I can tell you that I wished all of the KDE RPM packages had been compatible back when I used SuSE and Mandrake. "RPM Hell" was the major thing which made me switch to a different distro, in fact. Although packages such as kdelibs and kdebase are one thing, I don't think it would be too much to ask that a simple application such as Kimdaba or K3B could release a single RPM, and have it work everywhere. If this is the 'branding' that Aaron is talking about, I'm all for it.

Speaking of which, Americans (on average) like their products stylish, even at the expense of reliability or functionality. This doesn't always apply, but it's a good rule to consider in general. When styling is equal, obviously Americans go with the product they perceive as more reliable. There have been several companies that have improved sales of a product by doing nothing more than changing the logo, even with just a slight brush-up.

I'm not seriously suggesting changing the logo, but if there's something we can do to associate KDE with reliability and quality, then it's IMHO A Good Thing, especially as regards the North American area of the world.


By mpyne at Thu, 02/26/2004 - 20:27

unfortunately that flexibility has led to different groups who distribute KDE to have completely different defaults, including basic technologies such as application menu systems!



Why unfortunately? I think KDE should welcome it when service providers offer adapted solutions using existing KDE features designed to target the specific audience they target, it's the same what KDE should do when such middlemen wouldn't exist. Where it gets unfortunate is when there is no cooperation between those middlemen and KDE, but I honestly don't see how your emphasizing of defaults and discouragement of defaults diverting from those default defaults is going to make the situation any better.



why is it that whenever i speak of usability or branding, some people get concerned that i’m envisioning GNOME’s approaches to things?



I mentioned it since the way you are emphasizing defaults reminds me of said other project, it is said other project which wants to define "defaults" as an ideal and bury everything else under a thick layer of expert knowledge. And I'm honestly concerned how blown up those needless "usability" discussions about defaults are becoming also within KDE instead that we finally define different targets audiences for which respective setting schemes can be offered in KPersonalizer. Issues like the ridiculous debate about the Konsole icon on Kicker could be completely avoided by KPersonalizer simply asking whether the user has prior experience with command line/console interface/dos prompt and respectively showing or hiding all possible instances of this tool. If we want to offer an ideal environment then its one which adapts to the users need. Static defaults, however well they are chosen, don't adapt to the users needs and consequently will always lead to needlessly controversial discussions stealing everyone's time better spent elsewhere.



as a side note, GNOME hasn’t achieved a level of effective branding either: Bluecurve, Ximian Desktop and Java Desktop all look, behave and feel different and none even mentions GNOME in their name. each of the three have a unique corporate-driven “brand", and in a way that’s what i’m advocating against. in fact, i think that form of branding (which is more “traditional") hurts the source project more than it helps it.



If you seriously think the rebranding of desktops by distributors is not helping us at all I guess you better ignore them and start from zero building your own distribution. I personally would rather prefer encouraging distributors to make use of KDE features in their defaults we can point to them as usage examples, and people who don't like to setup their KDE themselves that way can be referred to that one service provider. The rest willing to setup KDE themselves and for others should be educated how to achieve similar results. KDE, if the numerous comments about the Ars Technica article are honest, already is an unique system which doesn't easily get mixed up with other systems anymore. I believe branding under this circumstance is possible.



while they may not mix and match packages (today) between OSes, people definitely do get exposed to more than one set of KDE packages. [...] and in the future it would be nice if people could mix and match packages from different OSes. in fact, i’ve done it myself between Red Hat and SUSE and it’s rather nice. the future is compatibility, and it would be nice if KDE was a leader there.



As long as KDE doesn't have a strict guideline about how a system should be ideally build (similar to LSB or whatever) KDE should not care about what distributions do. It's a problem distributions are facing and have to handle at the moment. If your stance is that a guideline about the base system is needed to ensure "KDE is compatible with KDE" I agree, but imo this is still not about user visible defaults.



i’m not sure what makes you think that the great Ars Technica article and branding are at odds with each other. could you expand on that for me?



They are only at odds witch each other if you discriminate between features and defaults, as in KDE sets the defaults and educates about the features but forces service provides to ship KDE only with the "original defaults" instead allowing and encouraging them to promote specialized solutions directly.



the point is to get people to be aware of and recognize the name and software that is KDE, to associate it with reliability and quality.



And defaults matter that much here? ;)

I personally think KDE should start to heavily promotion its applications, we already have much to offer but by not promoting them separately we are missing quite some potential. KDE applications still, after all those years, have the widespread dubious outdated fame of being well integrated but no joy to use. MacOS X, being able to run KDE applications without an obvious KDE part around it, is a great testbed for seeing how each of the applications fare on it own, and imo the focus for usability improvements should be applications and not some dubious KDE system defaults which serve well for controversial discussions without any useful result.



Cheers.


By Na at Thu, 02/26/2004 - 21:14

I wrote this and didn't post, then read more, and am more confused :)

Isn't branding essentially an attempt to control the customers view and experience of a product? Branding implies controlling the trademarks, the way it is presented when sold, etc.

If I'm on the right track here, KDE would present good defaults that everyone (or almost everyone) would hopefully use. I can't see how this could be enforced, other than by technical excellence. My impression is that KDE has always done things that way, without much noise produced very good quality stuff that attracts attention on it's merits. For example, Konqueror will replace Mozilla as default when Konqueror is an able replacement.

Derek


By dkite at Fri, 02/27/2004 - 00:47

Branding - why not, defaults - necessary, and when I think about it, they go hand by hand. KDE needs defaults for the following reasons:

-consistency=ease of use: I tried Mandrake, RedHat, Debian, Slackware and LindowsOS, and found that KDE is more-less differently set up in all of these distros, the minute I got used to mdk's KDE I switch to RH and find apps in different sections of the menu- waste of time; various distros organise the desktop in their own ways - home dir is on desktop or not, start here is used or not, or these are oddly mixed... Konsistency, make a theme out of it;).

-recognisable desktop- in RH (they're notorious:)), I couldn't make a difference between KDE and GNOME, and although consistent app-look perhaps needs smth. like that, I still want to see what's what

-integration- all apps should be integrated as tightly as possible, with predefined set of apps, that would, through the process, brand KDE. Gotta choose the best: K3b for burning, Digikam for cameras, Kopete for messaging, GwenView for pics, etc., bundle them together, and make them interact as close as possible, like Apple does with Mac.

Question of freedom of choice - of course, but let the user choose at the first log will he/she use distro- or KDE defaults; for apps - install every good app out there, but don't expect to have flawless interoperability as with those shipped with KDE, although it can be managed through KParts I think.

So, we would have both flexible and recognisable desktop, free for everyone to play with.


By brainkilla at Sat, 02/28/2004 - 15:47