Re: Hello and polling users

On Monday 27 September 2010 00:04:12 Ron Katzir wrote:
> Second try, some more trimming, no attachment ...

And then the first one made it through at almost the same time :-) Still, this 
one is shorter so I'll reply here. I've lots of questions, please don't take 
these as discouragement - this is just not something I know much about but 
there are some issues that would need resolving.

> >> 
> >> 1) Actually, posting the idea on the mailing list was the first step of
> >> involvement. I've thought about it after looking at the resources
> >> mentioned of how to get involved. I find it difficult to contribute
> >> (more than submitting ideas) without actually using KDE. Additionally,
> >> writing articles might be too big a commitment right now. With all that
> >> said - I will gladly write questions and procedures for surveys,
> >> starting with this discussion. Firstly, the goals of the surveys, and
> >> other attributes, should be specified (more on that later as well). I'm
> >> somewhat interested in helping the digest, as long as I have time for
> >> it.

Cool. Start with what you want - if later on you want to do some of the other 
stuff that's great too.

> >> 2) This is actually how I came to think of the surveys idea, and also to
> >> bring up the KDE Brainstorm idea long ago. I have a list of set-backs
> >> (first and foremost the amount of time I can invest in converting)
> >> which I will specify at the end, since they are not part of the
> >> discussion about the surveys idea. I'd appreciate any help, especially
> >> from hard core KDE folks

Who are always very busy :-) But will do what we can.

> >> 3) Now to the real stuff :)
> > 
> > I think there could be different goals to surveys, and I want to start
> > with a use case:
> > 
> > Let's say I have an old laptop, and I want to move to Linux. I'm
> > concerned about several issues, especially the battery life and
> > compatibility with this model. I do some research and stumble upon KDE's
> > website. There I can search for a suitable distribution for my laptop,
> > based on information of from surveys of users on identity.kde.org (with
> > permission), and I find openSUSE is best. I install openSUSE with KDE
> > and boot. On startup, I get a notification about support survey
> > (Something like "Would you like to fill in a survey to help us support
> > you better and make your experience more enjoyable?"). I'm a noob
> > without experience, and this is tempting, so I press Yes. A dialog
> > appears and asks me two questions - whether I would like to install a
> > plasmoid on the desktop or open a browser, and whether I would like to
> > register as a KDE user. Let's say I answer Yes for both questions, since
> > the alternative is similar and simpler. A plasmoid similar to Kubuntu
> > Feedback automatically installs and allows me to register to
> > identity.kde.org (or a similar website if we should find a different
> > solution) or just log-in. Then I'm asked to answer questions regarding my
> > system and installation and skills to create a user profile of me and
> > identify the distribution-computer combination I use. I'm a noob, but I'm
> > not afraid to go technical and have very little time - I want it up and
> > running asap. Based on the information I input and information stored in
> > identity.kde.org, the plasmoid presents the problems I might face with
> > this system and distribution, along with solutions to these problems. In
> > addition, the plasmoid presents a list of possible help resources (blogs,
> > websites, manuals) which would help me set up things and a list of people
> > willing to help me set things up, with experience with this kind of
> > system. It might also explain to me how to use IRC (interactively?) or
> > otherwise how to contact these people using KMail, Kopete or similar. It
> > might also be possible to use Telepathy in the plasmoid to facilitate
> > the connection. (This relates to this idea in Brainstorm:
> > http://forum.kde.org/brainstorm.php#idea84164_page1)

This would be cool. It's also masses of work - setting up the infrastructure 
on the KDE side, getting the distributions to include the Plasmoid or 
whatever. I don't want to sound discouraging :-)

> > Surveys (as Google has conducted regularly) are probably the best way to
> > find out why people are dissatisfied KDE. I've indeed started by asking
> > myself why I haven't moved to use Linux and KDE and identifying the
> > causes.

Sure, this is always interesting. However, sometimes the issues are things 
that are not within KDE's control (when it relates to using KDE software on 
Linux and there are hardware problems for example). Still it would be good to 
have the information.

> > Another use case I see is related to promotion - if someone who stopped
> > using KDE fills a survey stating his reasons of dissatisfaction and
> > allows us to register him, we can contact that person when the issues
> > are repaired asking to re-evaluate.

It would need a degree of automation, but allowing a user to cite a load of 
bugs and get notified when they are all fixed - maybe possible but again, not my 
area of expertise. A verbose text report of problems from the user would need 
a team of people to read it, monitor the problems and remember to contact them 
when fixed, which does not scale.

> > The way I see it, we should refer to 3 different user groups - core users
> > (including developers and contributors), floating users (those who can't
> > decide yet) and
> > non-users. The last group can be divided into 2 subgroups - those who had
> > never experienced KDE and deserters :)
> > 
> > I think identity.kde.org can be a most valuable tool to learn about the
> > core-users. Not only to identify their causes of dissatisfaction, but
> > also gain insight into what hardware and distributions they use, which
> > problems they overcame and how they did it and similarly. Registration
> > allows tracking (with the approval of the member, of course), which
> > enables identifying if someone has stopped using KDE, for example.

So the system pings identity.kde.org when the user logs in? Interesting 
privacy issues there, but possible with consent.

> > That's why I think identity.kde.org surveys have a big potential to
> > provide significant information, albeit the registration required.
> > Registration has another benefit of dividing core-users into groups -
> > since developers and users have some different requirements, thus would
> > probably be dissatisfied with different issues. This might help with
> > assigning tasks: for example, if Nokia is more interested in expanding
> > the developer-base, it might be more appropriate to assign its funded
> > developers tasks regarding developers. Integration of the desktop with
> > the website is another idea which I find beneficial.

Who would determine which group a user is in? The user?

> > Surveys without registration are more relevant to floating-users and
> > non-users, I believe. It does not allow tracking, which means their
> > results is a snapshot of time, and it's harder to measure progress. I
> > think there is no substantial difference between the surveys themselves.
> > Nevertheless, having non-registration surveys allows to reach non-users
> > more easily. Having them conducted or referred to from other websites
> > such as the distributions' websites (Kubuntu, openSUSE, etc.) or
> > significant F/OSS websites (Phoronix, blogs, SourceForge etc.) will
> > increase the possibility of reaching this audience.

Sticking something on the Dot gets us to several places, however possibly 
biased towards KDE people. Some other sites would probably run a short story 
if we asked them

> > Some other goals for surveys, affect different kinds of questions. I
> > mentioned the first-time survey. I imagine a time/usage-based survey
> > measuring satisfaction after 2 weeks of use or 10 log-ins or so will give
> > a good indication of why people desert KDE or continue to use it. I
> > think a regular 2-or-3-months survey for core-users should be conducted
> > as well, to measure progress and influence development during cycles
> > (might affect importance of bugs). This survey might even include a "5
> > most annoying bugs" question :) (This is redundant to the bug-system,
> > but is a more active approach to rating bugs), among questions about the
> > hardware used and issues with it. For the general audience (without
> > registration), the survey should start with identifying to which of the
> > 3 groups the person taking the survey belongs, then identify the
> > subgroup. For deserters, the questions should be meant to identify
> > causes for desertion and what would bring them back to KDE. For these
> > who haven't used KDE, questions should identify strengths of desktop
> > environment they currently use, whether they would switch, would the
> > strengths of KDE appeal to them.

Ok. We have to be careful not to annoy our users by asking too many questions 
too often: "Why did you stop using KDE software? Because of all the annoying 
surveys." ;-)

> > Regarding results, I believe these of the mentioned surveys would help
> > discern problems at upstream and downstream more easily, and address them
> > better. Hardware surveys, I believe, would help make problems clearer,
> > and bugs could be submitted and explored more easily. Thus specific
> > driver problems could be submitted to ATI or NVidia or Intel, and maybe
> > this would help to better communicate with them.

Could be very useful. What are people using to run KDE software? We really 
don't have a clue at present.

> > (*) I wish to say something about what are the strengths of KDE for me
> > and promotion, since that's part of the purpose of surveys.
> > 
> >> I've seen a lot of promotion for applications. The reason I want to use
> >> Linux and KDE is that I'm under the impression that if I'll make them
> >> work properly and to my satisfaction, I'll be able to utilize my
> >> computers more and in a much better way and would benefit more as well
> >> (then when using Windows).

For sure :-)

> >> The technologies manifested in Linux and KDE
> >> attract me, even if they are far from being usable. Nepomuk,
> >> PulseAudio, network transparency, Akonadi, Telepathy - I see great
> >> (unfulfilled) potential in them, and this keeps me a floating user.

It can take time to appreciate them. When I first switched to KDE software it 
bothered me that I couldn't find a decent ftp client for web development. It 
was a bit later that I realised I could edit the remote files and images 
directly in KDE applications because they didn't care the stuff was on a remote 
server - or that I could do any file transfers that were necessary in 
Konqueror. Network transparency in KDE software is beyond awesome. The other 
things you mention are not awesome yet ;-)

> >> I
> >> want to use that potential in my computer, and these things would help
> >> me. In addition, the ability to use more fully the potential of older
> >> machines is a great benefit for me, as well as part of the essence of
> >> being free. For example, Windows 7 does not allow me 3D acceleration on
> >> my laptop, while Linux does (old intel 915). Application wise, I like
> >> Okular a lot, and wish I could easily use it on Windows.

> >> Nepomuk and Akonadi can't be promoted yet since they don't hold up to
> >> their promise or potential.

Yet we need to explain why they are going to be important so people understand 
where the pain is leading.

> >> I guess they require more development
> >> resources than available. I believe the survey system would help this
> >> as well, since it might indicate what direction for development is more
> >> important to certain groups. Thus, for example, the project might
> >> decide than 4.7 should have significant performance improvements to
> >> attract users, or to the contrary, implement many new features for
> >> media centers. Results can also be used to indicate what promises are
> >> most highly anticipated. Imagine having a sort of advertisement on
> >> KDE's and Nepomuk (and amaroK) websites stating that Nepomuk
> >> performance is considered one of the most annoying issues, and calling
> >> for help for this specifically. It's a tailored "Getting Involved" ad.
> >> Wildly imagining, we might even approach people showing a tendency to
> >> contribute asking them to help on issues they're most concerned about
> >> and giving them an easy way to do it.

This is a weakness (also a strength) of KDE. People work on what they want to 
work on and what they are good at - so we get very skilled people working on 
stuff they're passionate about, rather than under-motivated people who were 
just the best available for the job. We know that printing was an issue at the 
start of the 4.x series, but there weren't people around with the skills and 
inclination to fix it.

Tailoring a get involved message has some potential. This has been tried 
before to some extent with the concept of 'jobs', but not really followed 
through or high profile.

> >> A comment about Brainstorm - I think brainstorm would benefit from some
> >> tool that would help identify similar ideas, or ideas belonging to
> >> similar areas, or applications, or KDE components. It's not easy - I
> >> wonder if a tool similar to this for identifying similar bugs is
> >> possible.

People :-)

> >> This relates to surveys in that ideas could be linked to
> >> surveys' results, maybe even linking the results to the people who
> >> submitted the ideas related to them (for feedback or other involvement)
> >> and developers related. Use case is contacting a user who has submitted
> >> an idea to implement something more efficiently to test the
> >> implementation, since it came up as important in survey results and was
> >> implemented as a consequence.

Yep. Of course it's possible to some extent through the bug tracker - you get 
notified of follow ups to your report so you know where things are going. 
However, testing the fix might require building from source, a big turn off for 
a lot of people.

> >> Since sharing is very important in F/OSS, I believe the system I
> >> described could benefit other projects as well, including GNOME. If so,
> >> it might be better to get help building such a system, by building it
> >> jointly.

A generic system would also help us get a more unbiased view - hosting jointly 
with GNOME would give us feedback from GNOME users and vice versa. More people 
to get together on this though...

> >> Maybe some of the ideas described above would help KDE receive
> >> funding as well.
> >> 
> >> Tasks and components I see related:
> >> 
> >> 
> >> 1) Surveys written for specific audiences, goals and times
Todo
> >> 2) Survey servers
Available (from what Kenny said)
> >> 3) Database (on identity.kde.org?) of users and their
> >> distributions-hardware, issues and solutions, etc. (Similar a bit to
> >> ThinkWiki)
Yep ThinkWiki is amazing
> >> 4) Survey (and support) Plasmoid
> >> 5) Survey and support data-engines(?)
> >> 6) Survey results analyzer
> >> *) Some mechanism to integrate data and search the database mentioned
> >> above, on servers or on desktops or both.

So, how long would it take to get all the above in place, bearing in mind that 
unless someone else appears with the time and interest to push this, you'll 
have to do most of the harassing of people?

2 is more or less there. 1 can be done fairly quickly. So if it was me I would 
probably got for those, get a first survey out to prove the point that it is 
useful and maybe get some more people interested (and also so I felt like I'd 
actually got somewhere). Then try and hit the others, which are all decent 
goals. The downside otherwise is that you can spend a year on this and find 
that 1-6 are all in various states of being done and you didn't see any result 
yet for your hard work :-( However, it's your call - which is the beauty but 
also the curse of getting involved in this kind of project.

> >> If you find the written above interesting, I believe that Aaron Seigo
> >> and Tom Albers should become correspondents as well.

Aaron is on this list at least

> >> I hope this is helpful
Lots to think about :-)

> >> About my systems and why I fail to use Linux. This is actually more
> >> related to Linux then to KDE. I don't have much time to invest in
> >> getting things to work, so I stick to Windows.
> >> I have 3 computers at my disposal, neither of which is extremely strong
> >> nor new. I use 2 regularly with Windows 7, and 1 is intended as a
> >> server. I'll list their hardware and usage:

Thanks, this is interesting

> >> 1) IBM X41 tablet - 1.5GHz Pentium M, 1.5GB RAM, i915, 30GB HDD + 16GB
> >> SD. Used for note-taking, mails, reading and writing. Sometimes for
> >> Matlab calculations. Problems - power consumption and tablet support.
> >> Tablet requires a lot a hassle. Power consumption is crucial - Win7 has
> >> around 3.5 hours of work off the battery. Linux consumes much more
> >> leaving about 2.5 hours of juice (I think. I need to try in fact ...)

This doesn't surprise me. I think people are starting to wake up to this - 
Intel has powertop for tracking power consumption and distros etc are starting 
to work on this I think. But we're behind Windows? Very likely. Having said 
that, I get four hours out of my five year old laptop which is good enough for 
me :-)

> >> 2) AMD 3.2GHz (Dual-core) Athlon X2 6400+, 2GB RAM, integrated ATI
> >> HD3200, plenty of HDD room. Used for calculations and simulations, HD
> >> movies and music, with 2 screens. I'm mostly worried about Full HD
> >> movies playback and sound, and the dual-screen configuration.

I think the free drivers are in a reasonable state for R600 at this point, at 
least on some of the cutting edge distros (Fedora etc). I run an R500 with no 
problems. Dual screen is unknown territory for me though...

> >> 3) 2.8GHz Intel 4, 768MB RAM, NVidia GeForce 440MX, 80GB HDD. Currently
> >> unused, I meant to use it as a server (possibly for surveys as well) and
> >> media center for regular TV. Been trying to make LIRC work with and old
> >> Creative iNFRA CDROM without success. Have not idea how to set up a
> >> server.

I've never had much joy with LIRC :-( However, xbmc seems to just work...

> >> By the way, I thought to pick up programming by starting writing a
> >> plasmoid. I have a nice idea for it. Sounds like a good plan?

I think so :-)

Cheers,
Stu
 
_______________________________________________
This message is from the kde-promo mailing list.

Visit https://mail.kde.org/mailman/listinfo/kde-promo to unsubscribe, set digest on or temporarily stop your subscription.