[kde-edu]: [RFC] KPhysics - Interactive physics experiments for KDE

[Jure Repinc]

On Fri, April 28, 2006 16:43, Jason Harris said:

> On Thursday 27 April 2006 12:59, Jure Repinc wrote:
>> Yes the framework will be the first priority. And there will be just
>> a
>> couple of simple experiments of different types to test the
>> framework. I
>> think that unit testing could also play big part here. Right? I just
>> saw
>> the first part of some tutorial on unit testing with Qt/KDE that I
>> could
>> take a look at.
>>
> Ok, great.  I don't know how plugins work either, but let's worry
> about that
> later.  For SoC, you can just code one or two demo experiment modules
> directly in the project.

That would be an option for the SoC. I guess I'll just do it this way
for now and then after that covert it to full plug-in based system.

>> Nice to hear this. I think it will not be hard to throw away many
>> habits. Simply because I don't have so many as I'm still a beginner
>> in
>> C++/Qt/KDE programming. All that I was concerned with is this moving
>> kdelibs target. Do you know what is the best way to keep in touch
>> with
>> the changes?
>>
> Unfortunately, I don't know of a way to monitor the changes in
> kdelibs4_snapshot.  I monitor the kde-core-devel list, and still I am
> usually
> unaware of what is coming in the next update.  What I usually do is
> update
> kdelibs4_snapshot, then try to recompile kstars and look at what has
> broken.
> It's usually not too much.

Yeah others have suggested the same way. just compile and see if
anything brokes. I was also told that there is a HTML file
(KDE4PORTING.html) in kdelibs folder that lists the changes.

> Carsten brings up a good point: you could make KPhysics a pure Qt4 app
> to
> start, if you don't want to keep up with kdelibs changes.  Actually,
> it would
> probably be easiest to have a little KDE code at the beginning (i.e.,
> KAboutData, KCmdLineArgs), but when making class choices, you could
> choose
> the Q* variants for now, and later we can KDE-ify everything.

I agree. This was basicly my idea at the very start. To use only Qt
classes at first (if both KDE and Qt one exist) and only then try to
convert Qt ones to KDE ones slowly one at a time.

>> Thanks for your offer. I hope that someone with more time will be
>> able
>> to help and that I don't eat to much of your valuable time. I like
>> KStars a lot and I don't want it to suffer just because of me :)
>>
> Sure.  I'd like to be involved at some level, even if I'm not the
> official
> mentor.  What is the next step?  We need to add this project to the
> list on
> the KDE wiki, I guess, and you need to write up your proposal.

Yeah I guess if someone will be willing to be my mentor then it can go
to the KDE SoC2006 Wiki page. And I can prepare the application for
Google so that it is ready for 1st of May when they start accepting
them.

>> And I agree, the size of this project and the fact that I'm a
>> beginner
>> and it would take a lot of time to do it are the reasons I ask if
>> this
>> is even sutable for SoC application. The money would sure help as I
>> wouldn't have to work during summer and which would allow
>> concentrate
>> only on the project. But if this isn't possible I sure hope that I
>> could
>> have a mentor who would help me as if I was in SoC.
>>
> It's perfect for SoC, as long as we keep the scope down.  You're
> proposing to
> do a proof-of-concept of the physics backend for KPhysics, not
> KPhysics
> itself (but you should mention that you will continue to develop and
> maintain
> the project as part of KDE beyond SoC).

OK. So I'll try to keep it as simple as possible and use that for the
application for SoC.

>> P.S.
>> Would you know anything about the problems with KDE mailing lists
>> server. I had to send the first message a couple of times before it
>> showed up. And I have already sent this reply to the kde-edu list a
>> couple of times but it still didn't show up. Who should I contact
>> about
>> the mailinglist server?
>
> Hmm.  I'm the list manager for kde-edu, and I don't remember seeing
> any
> message errors recently.  Could be a server problem.  If it isn't
> cleared up
> in a couple of days let us know.

I'll wait for a couple of days and see if it gets better. I've also
noticed that if I send the mail from my webinterface SquirrelMail it
always shows up, but if I use Thunderbird 1.5 it does't. So maybe it
could be some problem with parsing the e-mail that Thunderbird
creates. I'll try later with KMail also and see how it goes.


-- 
JLP's Blog - http://jlp.holodeck1.com/blog/