This is the work blog for the KDE Visual Design Group a group of dedicated designers striving to improve community participation in design for KDE as well as design work in general for KDE applications and Plasma.

måndag 3 februari 2014

Monday Report #1

In which an institution is born and the stories of the past week is mulled over. But firstly a short thing about my problems to comment on my own blog and a short "sorry".


Right, the issue I thought I should mention is that I have problems commenting on my own blog using chromium so if I don't reply fast enough - please don't think I don't read what you've written, its just that I need to move between browsers. So I might comment a tad slower.


The idea here is to give a weekly update of whats going on in the Visual Design Group until stuff becomes more and more public. Hopefully every monday for a while.

First off right now we have the addition of four new designers:
Acidrums the designer who made the Betelguese icon theme.
Cuan Knaggs a South African designer working on business focused design at Revolver.
Luisa Pires a designer that focus on webdesign and brand design.
Muhammed Nafees a programmer and KDE dev from Jamshedpur.

More on all ten of us later since I plan on doing something like an interview when the time comes.


We're trying to build up the level of trust needed to run a group like this and at the same time compile a good list of jobs that has to be done as well as find a focus for our design ideas.
How do you do that on a very short time when these things tend to take time? You force it. So we've started setting up a short training exercise to get the work going - essentially just spitting out ideas and seeing what sticks. Go without limits, just have fun with ideas and concepts. From there the plan is to go to the realistic.

Simply match up the best ideas from the wild attempts with whats really possible and get a solid design idea out of it. There are some tentative tries with mood boards and suggestions at the moment and more is getting added daily.
By next monday - there will be solid work coming out next monday at this rate.


We're also looking at a good sollution to the icon issue. Lets be frank here - our greatest focus is on Plasma Next - and what we need is a more modern set of icons to be used and its to be released fairly soon. But the Oxygen Icon theme is enormous, its one of the most complete icon themes out there and it took about a year of constant work. So how do we manage to get it done in time?
As it is, one solution would be to replace the central icons, the system icons and then move slowly through the whole icon theme. Maybe replacing them with another icon theme to start with and then move through them all one by one?

Aside from that plans for launch pages for some of the underlying technology coming soon are being done and also promo and presskit ideas are being talked about.


Finally there is the issue of how to organize a design effort beyond the ten of us there are now? Bugtrackers and mailing lists have met with skepticism within the group to say the least - but how do you keep track of the work that needs to be done, what is being addressed by someone else and how to participate? How can we move from what is now a closed group into a larger community effort?

I'm not trying to be cute and start with rhetorical questions - these are all valid and something that I think about daily. The point was "create a design group and a good workflow and environment for designers within KDE" not "isolate all design choices". This is open source and transparency and openness are extremely relevant, but how do you combine them with design work without getting back to the issues that was before? (again, not rhetorical)

14 kommentarer:

  1. Thanks for the update(s), they're much appreciated!

  2. I would suggest you to focus on the application look and feel because they look like win95 apps with some gradients and round borders(ugly)...

    1. Thanks for this really constructive and useful comment!

  3. Use KanbanFlow or something similar like Teambox to manage code/issues. One card has a specific issue. Then different users can comment on the card and discuss. People can indicate what they are working on. It will give you a good picture of what is going on at any moment and how to redistribute resources.

    1. Or KanbanTool has a good editable API too..

  4. The icon issue - new icons would be really cool, but shipping only half of the set and having two, potentially very different icon sets shipped as default and showing up next to each other in one application might be baaaaad.

    As icons need time, I'd say let's start the work, have it distributed through some 3rd party (ie. not shipped by default with Plasma) and replace Oxygen only when ~90% complete. I think that being visually consistent is more important than having 30 new shiny icons (out of 500....or howmanyever Oxygen has).

    1. Well the idea here first off is working on icon designers that are not "in house" to get some of them ported to KDE as a way of "lessening the blow".
      Secondly comes to switcheroo. The thing is that there are some icons that CAN be replaced (like system tray icons) that aren't that tricky to match up to Oxygen as well as another icon theme. Or logout buttons. Then you can move on to the icon theme proper by either using a second set as base, which works better with the new icon theme (with the icon creators blessing of course) and can function as a blue print for the new icons.

      But just so you know - it will either be changed well or not done at all. Which is sorta my philosophy through-out. There are things that need to be changed - but at the same time if the alternative (even if its transatory) isn't good enough: its better not to.

  5. Agree with Jens on this one. A transitonary icon set would be rather a waste of (limited) manpower and time.

    We should as well think about creating new, easy to understand icon guidelines, like for example JOLLA did for Sailfish here:

    This would save quite some headache in the end, plus, make for a more homogenous look of the whole icon ecosystem.

    1. Guessing you mean Martin? :)
      I'd love to create a whole icon set from scratch - the issue is the time needed to do so. It would have to be as extensive as Oxygen (which was about a years worth of full time work). It would also have to go with the general mood and theme of Plasma Next visuals.

      Lets be very clear here - I'd love it! It would be awesome to introduce a new icon theme to go with the next version of Plasma. (actually it would be relevant and important to do so) But again, barring some sort of dark magic I don't know if we can pull it off in time.

      Icons are a tricky bit - because they can't be a communal effort in the same way many other aspects can. They can be commented on but it demands a LOT from the commenters.
      "I don't like the way it looks" = the worst kind of comment, "I don't like the way X looks because of Y but if you do Z then it would be great" = good comment, "I've done Z for you so its not Y so now X looks better" = gold-star-and-free-cake-comment.
      The issue with design is that its often the first kind of comment ("I don't like it") because everyone has an opinion of looks and aesthetics but few have the insight of WHY they like what they like and fewer still have the skill/time to apply that and contribute.
      But since icon themes are so vast in comparison with other details of design where that comment can actually be used - it just clutters the work and make it more difficult.

      So the question is "How to solve that?". If we can create a work flow where you have both the focus and unity you gain from a single designer combined with the collaborative power of communal design - well then we might be able to pull it off.

    2. Well actually i meant both of you ;-)

      Like you said above ""If we can create a work flow where you have both the focus and unity you gain from a single designer combined with the collaborative power of communal design.""

      Even if we'd have the luxorious situation to have a designer working fulltiem on an iconset for a year, it still wouldnt solve the problem of inconsistent thirdparty application icons.

      Again, i think this could be overcome by having very good (and rather strict) guidelines, ideally supplemented with a few final examples and .SVG modules/examples.

  6. Nice to here this.

    Will the new Icon theme be (by design) more scalable than the current Oxygen theme? (Means: being more flexible to scale up or down for high- or low-resolution devices or or different form factors?)

    1. Using SVG as the icon format certainly would be one solution of the problem, but as far as i know special care has the be taken be make lines appear smooth when scaling down a lot if you use svg as format.

      On the other hand you can use a simple script in inkscape to put out png's in a variety of different sizes. Hmm.... :)

    2. Its a tad trickier than that, beyond the fact that you need to do some pixel alignment to avoid blurry icons its also a question of keeping all elements in an icon visible. That means that a small icon tend not to look at all like its larger counterpart.
      The designer removes and enlarges certain sections of the design that need to be visible even at smaller sizes. Check out the icon theme spreads you can find and you'll see that the smaller icons are not just tinier versions of the big one but unique images.

  7. Sebastian Kügler has described at some solutions for this for Plasma. Will there be also a solution for QWidget based applications for KDE 5 framework?