JUL
23
2003

"Raising a Curtain"

Hi, here's Datschge.

I guess I should introduce myself first, but there's really not much to say. I'm a big fan of Motoi Sakuraba and am crying me myself a river for not being able to visit his live concert last Saturday. -_- I'm putting my oar in into KDE for about one year now. Being an ultimately useless non-programmer I've been digging for parts where I can easily teak ("improve") stuff while hoping to reach my goal, making it easier for "outsiders" to effectively contribute to KDE. I started being vocal during the discussions which led to the kde.org redesign, added noise to the usability list, then decided to pick bugs.kde.org, something noone else seems to like to tackle, tweaked stuff here and there, added an hopefully easy to grasp supporting KDE page which I hope will be translated and localizated into other languages/localizations soon. Right now I'm "working" on (I'd rather call it 'keeping myself distracting from') writing a decent help page for bugs.kde.org and putting together a pseudo "KDE HIG" informing people (myself!) about all the build in consistency features KDE offer for ages already. Right now I'm totally excited since I got a hack of weekly-bug-summary for showing the amount of old reports and the ratio of old and all reports per product to work in my local bugzilla installation. Thanks to Steve for the inspiration. ^_^

Hm, I think I wrote a little too much here...

Comments

Cool, it would be great to see a HIG actually based on science for once. Have you started to look at the current HCI research out there yet? Zack and I had talked a while ago about setting up a research project at PSU here in the US where we can test out some ideas and get some more realistic numbers of how things are working and how people are using our apps. One things OSS really gets short sticked on is we dont get the fuzzy warn usability research companies get. We just have to live with "experts" who make things they think are cool....

oh well good luck


By Ian Reinhart Geiser at Wed, 07/23/2003 - 18:02

There's not much really usability related about the document except that I intend it to showcase the existing KDE framework, especially features like Kiosk or stuff like KDialog::spacingHint and KDialog::marginHint you yourself added, which should be considered and extended on when looking for possible usability improvements. So far it's imo a pity that there's much HCI discussion (science based or not) on the usability list without knowledge of the framework in use. Thanks to Henrique Pinto who's managing a school's setup using Kiosk and has rather instant usability feedback this is changing.

As for doing HCI researches I still have good contacts to several teachers responsible for computer stuff at my former high school which I intend to make use of after the current summer break, but that's a completely different topic.


By Na at Thu, 07/24/2003 - 11:16

Hi,

It's brilliant that you're working to try and open KDE to outsiders. I'm currently researching participation in the Free Software community, and KDE is a very interesting example because on the one hand it's one of the most open projects already, and yet on the other hand there are armies of people on sites like kde-forum and kde-look who talk about changes they want to see in KDE, and who seem to get (or feel that they get) no audience with KDE developers.

I think some concrete changes that could be made could include:

- An intuitive bug reporter, so that when a KDE application crashes, there is a big obvious "inform KDE developers of this problem" button (not something obscure like "file a bug"), that gives a form to describe the problem, and submits the data to bugs.kde.org automatically. Not unlike the GNOME bug tool.

- More obvious contact with the developers, or with community representatives, so that if I, as a fairly new KDE user, think that Kate ought to have a particular plugin, I can go to www.kde.org and easily find from there a form, forum or mailing list on which to post my suggestion where I will be heard and will receive a response. Mailing lists are less accessible, so maybe a form or forum. This blog site is a wonderful addition, by the way; it really gives us users contact with the other side of KDE

- Maybe really take on kde-forum, make if more official, and get some developers, or representatives in there. Encourage users from www.kde.org to visit the forum and post their ideas, discuss development issues, etc. Being able to just air your ideas is a far easier way to get into actually contributing to KDE, as opposed to subscribing to a mailing list and jumping in at the deep end.

As I said, KDE is very good compared to othe rprojects of its size, but it could definitely improve its relationship between developers and users, and between the project and those who might contribute but don't want to commit in the way that they currently have to.


By KDE User at Thu, 07/24/2003 - 11:07

...there's no way we can force people to spend more time than they are already spending. Many developers are already unable to follow all the discussions on KDE lists, needless to say that outside sites like kde-look, kde-forum and all the local KDE sites out there can only get even less of a share of important attention. In the end much of the impression that some might feel excluded boils down to a lack of time of those who could help out. The only way I see changing that is encouraging people to inform themselves what possible effective ways there are to contribute, and make use of them without needing to rely on others. That's all we can do (it's much, just do it =).

Now about the changes you suggested:

- I don't know the GNOME bug tool (and am not at all impressed by their Bugzilla web frontend). Reporting bugs is already the most direct and sane way to inform KDE developers of certain problems and suggestions, it's easy to start a report through "Help" > "Report bug..." in any KDE app. Also the KDE crash manager in CVS Head used to have a link to bugs.kde.org afaik. If the wording of all those is your only problem let me know.

- As I wrote above the most direct and sane way to inform KDE developers of certain problems and suggestions is using the existing bug tracking infrastructure. Everything else will just either take up more time for those involved or get more easily lost in obscurity than it's really worth the hassle imo.

- Kde-forum is only one of many many places and ways to talk about anything related (or not) to KDE, hardly anyone is able to keep track of all the multinational forums, mailing lists, IRC channels etc. Also it's important to note that there's no general representative having the power to force KDE his way, so such persons won't show up in selective parts of the community either.

The only way to improve communication different parts of KDE's community afaics is sticking to the few real central meeting points as a common platform for suggestions and reports about problems for everyone. bugs.kde.org, being KDE's sole bug tracker, is one such place, so make use of it as individuals or as community. Another way to increase the attention to eg. kde-forum's output on suggestions and problem reports is letting someone regularly summarize them and send it to KDE Dot News or KDE Traffic. Or look at supporting KDE once again.


By Na at Thu, 07/24/2003 - 12:28

Ok, I take your point about the time developers have. I appreciate that, I'm a develop myself, as well as several other things, and I'd resent being told to spend more time on things I didn't have the time for!

KDE's bugzilla is potentially a *very* powerful and attractive tool, and thanks to your work the web frontend is far better than other bugzillas (yes, GNOME's is horrid). But there are a few objections I have to it:

- The whole thing talks about "bugs". Now if I didn't know any better, I'd think that meant that if something crashed, I'd report a bug. I might even think I'd have to know something about what went wrong to report it. It doesn't suggest that feature requests are a part of the system. Here are some ways you might be able to solve this:

- Where, in application menus, web sites, etc., you currently have "report a bug", why not "request a feature" and "report a problem"? They're far less technical terms, and even if both go to the same web page or application, they're more understandable from the user's point of view. Ideally, you then wouldn't have anyone submitting feature requests to kde-look, kde-forum, etc. because they would all be going onto bugzilla.

- Going back to my idea similar to GNOME's bug reporting tool, let me describe it to you better. In KCrash there is a text link that will open http://bugs.kde.org in Konqueror when you click on it. That's very user unfriendly, since I then have to create an account (something I might not want to do, or be too lazy to do), then I have to search for similar bugs, then I have to file the bug. It's all very intimidating. Instead, there could be a push button that says "Report this problem", that opens a dialogue box. The user can then enter a few keywords, and hit the "Look for similar problems" button, and the application will then search bugzilla for possible duplicates, and present them. If the user feels it isn't a duplicate, he/she can then click on the next push button and fill in a simple form giving the basic details, an e-mail address, and then submit it to bugzilla. The application can submit the error backtrace and all the information correctly formatted, maybe create a user account for the person, and have it all reported transparently. This is far easier for the individual reporting the bug. Of course that's a sketchy idea, but you understand what I mean?

Oh, on an aside, is there a procedure for content from kde-look getting into KDE CVS? A lot of content goes through kde-look, and I think it'd be very valuable if there was such a procedure put down in writing and made known. At the moment, it is difficult to tell how it happens. There are a lot of very good icon themes, window decorations and backgrounds that could made KDE's artwork look far more complete and modern that simply aren't getting through, some having been there since the days of 3.0 and before.


By KDE User at Thu, 07/24/2003 - 16:48

(Look, this blog entry already got moved to page 2 of the blogs page. Are people actually reading that far? ;) )



This is my suggested ToDo regarding your suggestions:

  • On bugs.kde.org there's already a "Enter new wish, bug or crash", I've also replaced several instances of "bug" with "report" where it makes sense and doesn't confuse.
  • Include an "Report new wish, bug or crash" wizard in KBugBuster with the same feature set like the wizard at bugs.kde.org.
  • Rename "Help" > "Report Bug.." in KDE apps to "Report Wish or Bug...", offer "Report Crash..." button in KCrash, both linking to a fitting KBugBuster wizard.

I'll look at improving the online wizard.



Kde-look is an outside (as in 'not residing in KDE's CVS, which is the only official place for official KDE stuff') site so it's up to those commiting the artwork etc. to kde-look to also commit it to KDE's CVS if they wish to do so. There are several artwork related mailing lists where I'd assume one can ask respective questions to members who have access to CVS.


By Na at Sat, 07/26/2003 - 20:08

I don't know who owns/runs kde-forum.org, but it does not render in KHTML, asking the w3c validator for a check yelds numerous errors. Shame on them.

Personally, I dislike sites that does not clearly state who owns/runs them (actually, kdedevelopers.org is guilty of that as well, except for the wellcome message...)


By KDE User at Fri, 07/25/2003 - 19:25

Just wanted to encourage you Datsche, you do great work, keep it up!


By KDE User at Thu, 07/24/2003 - 22:20