What’s up with KDE.org & Hello GetKDE.org

It’s been a while since I last posted about KDE.org, aka the KDE-www war series. It talked about the current KDE.org design, and how to improve it. The series started with target audiences and conversion goals, picked apart and restructured the sitemap, revealed an initial design proposal with clear-cut priorities, and finally analysed the effectiveness of design.

Since then, the KDE-www team has gotten serious about rebuilding KDE.org from the ground up and has started up project neverland. However, I shall now be continuing the work on KDE.org under a new name, GetKDE.org.

You can visit http://GetKDE.org/ right now – feedback is much appreciated.

One of the most important aspects of the redesign is community involvement. GetKDE.org is built publicly online irrespective of the KDE release schedule. This is so that the community is free to visit it any time and provide feedback and leave comments.

There are a few differences between how GetKDE.org is tackling the KDE.org redo and how neverland is tackling it:

  • The Oxygen team is unpredictable. Neverland’s answer to this is to design without employing the blue-coloured KDE, Oxygen, Air, or Plasma-themed elements as part of the basic design- that way, it will still be relevant no matter what KDE looks like. GetKDE.org instead regards Oxygen’s unpredictability as a fault of Oxygen, and does use the three biggest things which make KDE’s brand recognisable as it currently stands: Blue KDE, Oxygen, and Plasma w/ Air.
  • GetKDE.org is documenting its design process outside IRC. GetKDE.org wants to be 100% transparent with the development process, making sure that the community knows what’s being done, why, and can voice their opinions. This means taking things outside the IRC channel, as well as into real life. This is because any change to a significant visual thing representing KDE may mean changing KDE’s brand. This is not something to be taken lightly. This also means that GetKDE.org doesn’t follow the KDE release schedule.
  • GetKDE.org has a much smaller scope. Only pages within KDE.org will be considered rather that neverland’s objective of a one-size-fits-all solution unifying all sites, including wikis, forums, translated versions, etc. This means that a lot of content will be filtered out, but quality should outweigh the quantity.
  • GetKDE.org is following the previously outlined target audiences and goals. Neverland is following a more rapidly developed, iterative design approach, whereas marketing objectives have been laid out from the start in GetKDE.org, and it will follow that.
  • Neverland is the currently heir to KDE.org. Although GetKDE.org will perform exactly the same functions as KDE.org, most of the team are currently working on neverland. As such, GetKDE.org is being branded as an experimental alternative to KDE.org – and will stay that way until either community or statistics prove otherwise. GetKDE.org will not become KDE.org.

Well, I hope you enjoyed it! More will come soon. Just another week until semester is over and then I’m ready for a sprint :) Updates are being dutifully tracked on its WIPUP project.

Dion Moult

I've been developing software for well over 10 years, work as an architect (not the computer kind, the regular sort), and am classically trained as a pianist. I try to do the right thing when I get the chance in my field, such as through contributing to open-source communities and promoting sustainable living.

More Posts

Spread the love
Tagged: , , , , , , , , ,


  1. Dion, I want to make some suggestions.

    1. Typeface. Is it Arial? Is it Liberation Sans? Is it DejaVu Sans? I suggested Lato here: https://bugs.kde.org/show_bug.cgi?id=283928 , but Lato is lacking internationalization and adjustments for the screen. As the brand designer, can you suggest another typeface? I loved when KDE e.v. used Anivers, but Anivers isn’t even free as in price.

    BTW, you cannot be “seamlessly integrated” when you use Liberation Sans and DejaVu Sans in the same screenshot. A funny mistake.

    2. What does mean “Get KDE 4.7.3?” Getting distro packages? Getting a Windows binary with KDE for Windows? Getting a, say, OpenSuSE image with KDE and endorsed by KDE?

  2. BTW, I talked to the Lato author, and he is absolutely willing to integrate the enhancements we make to the official Lato release. Also, Lato is already released under SIL OFL, so, redistribution is not an issue.

  3. 1. It is Liberation Sans used in the blurb, but Arial throughout the rest of the site. Good point being made about the font – I shall specify Liberation Sans to be the default and Arial as a fallback in the next update.

    There has been discussions about an official KDE font for quite some times now apart from the legacy Kabel font. In my opinion the criteria is basically 1) already popular/distributed, 2) sans-serif, 3) as websafe as possible, 4) not too dissimilar to a fallback font for design reasons.

    2. The Get KDE 4.7.3 page will essentially being a guided page towards directing the user exactly how to get KDE. Eg: Are you Windows/Mac/Linux (OS detection in the page will be used), then if Linux, what distro, etc. We all understand that “Getting KDE” is a very complex thing to tackle but plans have already been made on how to simplify it.

  4. Also, Alejandro, I shall be watching that bug to see if Lato is picked up anywhere else in KDE. If yes, then it makes sense to start using Lato on the site as well.

  5. As one of the main contributors to Neverland I can’t understand 80% of this post.

    I think Neverland is open for the stage we are in. I came in 2 month ago as a ‘new contributor’ and I was very welcome. They helped me ‘forking’ Neverland (never used git before) so I could experiment and merged elements back into master.

    “Unlike Neverland and how Oxygen operates, which is through sneak previews and sudden radical changes”

    I have been in IRC all the time, discussing every detail with plenty of people and explained every choice (…and never heard anything from you). No sneak previews but 20 previes per day, if not more, so I can get feedback from users and designers. Sometimes that means a radical change, but we are still in the first stage so that’s pretty normal. If people don’t like a page I designed and do like, I trow it away and make another one until others are satisfied.

    “Every single page will be considered rather than creating an aesthetically beautiful frame then just copy-pasting the majority of the current content into the new design”

    Neverland is developed on a page-by-page base too. Even while just ‘skinning’ the site and working on the skeleton I rewrite text to show the effect of “better text” than the current ones.

    etc etc.


    irc: kde-www at freenode

    Bart Otten, developer Neverland

  6. Bart, I understand it may be slightly confusing for somebody who has joined a couple months back. In fact, neverland began after this project, through which the main propelling idea behind it was to strip away everything that was Oxygen. This proposal was then scrapped to continue with the new design.

    Unfortunately, the differences between the two projects are far too great for most parts of this proposal to be integrated into Neverland. As a result, instead of competing with Neverland, this proposal is no longer being considered as part of the KDE.org rewrite, and which is why it is now GetKDE.org instead.

    Here are the differences re-explained, hopefully that’ll clear up some misunderstandings:

    1) The first, concerning Oxygen, is a very clear difference and has been since the beginning. To give an example, if we replaced the KDE logo on each design along with all KDE screenshots, and replaced the text with Lorum ipsum, GetKDE still reinforces the KDE brand image, whereas neverland doesn’t. This is _NOT_ necessarily a bad thing. It is simply a difference.

    2) All of GetKDE.org’s development has been documented and communicated to the public – at the time with a secondary objective to help raise design awareness. This also allows people outside the #kde-www channel to post their ideas and comments rather than just the core developers who lurk in the channel. Neverland doesn’t do this – and again, this isn’t necessarily a bad thing – most development is done in this way. GetKDE’s design is also based on research involving a lot of real-life testing against people who know nothing at all about computers. As far as I am aware, neverland doesn’t do this.

    3) Page-by-page basis doesn’t mean “we work on one page, then move on the next”. If it did, that would be true for all web development. Instead, it is referring to how neverland aims to provide a frame through which every single site underneath KDE’s network can insert itself into. GetKDE, on the other hand, tackles a much smaller scope – KDE.org’s content only, and nothing underneath it. This excludes, for example, the wikis, the planets, forum, app sites, etc.

    4) Not much more to be said about this one.

    5) GetKDE.org will _NOT_ replace KDE.org. Neverland will.

    I hope that helped.

  7. Aside the really not understandable decision (read: I reallly can’t understand this post), the whole “getKDE” idea contrasts with the new branding strategy, which focuses on the community.

    I would add that rather than “lack of communication” from #kde-www, it’s lack of manpower. How many people do you think are working on Neverland? And how many people do you think they have time to communicate their results “outside”?

    Honestly, there are few rational reasons for this post.

  8. Luca, the site is being designed to be aligned with KDE being the community – there isn’t a contrast. That will become more apparent as the site unfolds. A common misunderstanding is with how you “pitch” KDE to a newcomer. People don’t join a community then enjoy its product, people come for the product, and stay for the community.

    There are about 5 main people working on neverland. I used to be part of it in the beginning, then stayed to monitor it as it progressed, then went back to this project. As far as communication and research in this project is concerned, it has been a one-person effort. Manpower is irrelevant. Clear visions are.

    This post has not been a random appearance and it based on everything but irrationality.

  9. Dion, thanks for your words.

    IMHO, Open Sans (as used in the Neverland project) is not an ideal font. It looks gorgeous on screen, but it’s the official branding font for Google. We need some branding identity.

    I proposed Lato because I like it, but also:

    – Because it’s free as in freedom (OFL license)
    – Because we can fix its shortcomings.
    – Because it’s available in a respectable set of weights.
    – Because it’s available ;).
    – Because it’s not too dissimilar to Liberation Sans or Arial.

    However, my ideal solution would be a custom-designed font for KDE. We can use the Google Kickstarter to our advantage; http://www.kickstarter.com/ . It’s all about branding, and that’s what GNOME and Ubuntu already did. Let’s do better.

  10. A strong typeface as a branding identity is very important, yes. Lato does seem fine and I will implement it in the next update. (instead of waiting for somewhere else in KDE to pick it up, why not be the first?)

    You are also right that ideally a custom-designed font is best. Do we know an experienced font designer?

    You can’t “Get KDE”. There is no and never will a “KDE 4.7.3”!
    It’s been two (!!) years since the KDE rebranding and you still don’t get it???

  12. I see that in the neverland project the #1 usability mistake is being made again: it assumes a minimal width of the browser. I do have a 1920×1080 screen but my browser windows are never maximized, they are 800 pixels wide and will never be wider than that, because I want to see other things next to my browser. In akregator the left column eats some of those 800 pixels, so if I view a website in akregator, the visible area is even smaller. Nowadays tablets and phones are popular and KDE puts a lot of effort in making their software available for these devices too, so also the KDE website should be usable on these devices (which have a small screen). Also on GetKDE.org this mistake is made, but here (although I’m not a big fan of this situation) it is understandable, because otherwise the layout is totally screwed up and because it is only like this on the front page (I hope) which doesn’t contain text but only links to the good stuff. If the columns at the bottom of the GetKDE.org main page are repeated in the subpages, then I will still have the impression that my browser is too small, even if the main text has the correct width, because it gives me the impression that the site is designed for larger windows. This gives an unwelcoming impression to me, such a website is telling me: “you should visit me using a decent, large webbrowser, please come back when you do”. When I visit such a page, I have a tendency to go away and never come back.

    Also beware of Facebook and other social website buttons: see http://ssrn.com/abstract=1717563 and http://www.guardian.co.uk/technology/2008/jan/14/facebook. Please take into account that a website which links directly to the social websites is much slower to download (because all those social websites must be contacted during load).

    Other than that, nice work!!! But don’t use the legacy K icon (in the first page; in the second page the much nicer new icon is used).

  13. “a”

    Neverland will probably have a different stylesheet for really small screens. As you may have noticed the website does resize to the space available. It’s not yet perfect but it’s WIP. The only thing so far done is 20 mockups and no final yet. Please tell me the use of bugfixing mockups……and how we make the #1 mistake…

    I agree with Dion on the KDE thing. Get KDE SC or Get KDE Software Compilation is just wrong even when promo says different. As far as I know it was not meant to be the way it is used now.

  14. a, thanks for voicing your concerns. Ideally each form factor should have a different design. GetKDE is essentially a fixed-width layout, and so will not scale for smaller screens. It has no plans to either. If a mobile version is needed, a mobile version shall be created separately.

    The neverland project, however, was initially designed to be completely usable on all forms, so any problems you do see about that is something that _does_ need to be addressed.

    As for your comments on social buttons and Oxygen usage, please relay those to the neverland team on #kde-www instead of here. Keep in mind that the neverland project is against using elements of Oxygen, which is why they are using the legacy icon.

  15. Moult may be more right than I thought. I probably judged to fast based on my own perception, openness and not entirely correct information.

  16. FAIL: position of the search box is

    When I want to search for something I don’t want to search for search box!

  17. r0b0t, it is drilled into the heads of every single webdesigner to place the searchbox at the top right of the layout so that it is “easily” seen.

    This is wrong.

    The real reason is that the search bar should be placed next to the navigational items, ie. where somebody would go if they were actually _searching_ for content in the site.

    Thus it is completely appropriate for the search bar to be where it is.

    Given the size of the page, it is also very easy to stumble across. This is a good thing too, as no page should be so monolithic as to have to wade through until you reach the bottom.

  18. Dion, as you can see in the bug report, the “Lato UI” font project, as the KDE default UI font, is gaining traction. The team behind Lato is considering our proposal and is going to launch on its own a Google Kickstarter project.

    I think we must support that once it’s ready, although it’s not official yet. Check https://bugs.kde.org/show_bug.cgi?id=283928 again, there have been some interesting developments.

    A “Lato UI” font for our KDE interface (and web page too) would be a great design decision, in my opinion.

Leave a Reply

Your email address will not be published. Required fields are marked *