Stuck on FreeplaneFX

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Stuck on FreeplaneFX

theworldbright
Hello everyone,

        It has been around three months since I started to attempt a JavaFX port for Freeplane. At this point, I would appreciate a quick two minute evaluation on my implementation of the JavaFX version of the OptionPanel. How many problems/bugs can you immediately see?
        I am not exactly a fan of giving up but I feel like pushing forward would be fairly inefficient. Even for a simple task of porting the OptionPanel, I have already noticed two bugs that I have no idea how to fix:

1) Occasionally the property panel just shows up blank
2) A huge vertical space is created when text is typed into the search box

        Both problems are the result of me directly modifying the library, ControlsFX, which I have been using to construct the accordion pane. Moreover, they do not happen consistently, and I am sure there are more bugs lurking around. I feel like I would never be reasonably assured that the implementation is stable.
        Furthermore, most of the classes under freeplane/core/resources/components are needed by the StyleEditor. Hence, when I write JavaFX specific code, I cannot delete the Swing specific code, causing a huge amount of duplication, and thus dirty code. In short, I am unconfident that I would be able to “support and fix every single issue which could come as a consequence of this development.”
        In case I was not clear, I am not asking for a fix to the bugs I mentioned above. I know most of you are more experienced programmers than I am. Hence, I would like advice whether you would continue in my situation or work on something easier and come back to this effort later?

Regards,

Kent
Reply | Threaded
Open this post in threaded view
|

Re: Stuck on FreeplaneFX

Felix Natter
Administrator
"theworldbright [via Freeplane Developer]"
<[hidden email]> writes:

> Hello everyone,

hello Kent,

cheers for your work on JavaFX!

>         It has been around three months since I started to attempt a JavaFX port for Freeplane. At
> this point, I would appreciate a quick two minute evaluation on my implementation of the JavaFX
> version of the OptionPanel. How many problems/bugs can you immediately see?
>         I am not exactly a fan of giving up but I feel like pushing forward would be fairly
> inefficient. Even for a simple task of porting the OptionPanel, I have already noticed two bugs that
> I have no idea how to fix:
>
> 1) Occasionally the property panel just shows up blank
> 2) A huge vertical space is created when text is typed into the search box
>
>         Both problems are the result of me directly modifying the library, ControlsFX, which I have
> been using to construct the accordion pane. Moreover, they do not happen consistently, and I am sure
> there are more bugs lurking around. I feel like I would never be reasonably assured that the
> implementation is stable.
>         Furthermore, most of the classes under freeplane/core/resources/components are needed by the
> StyleEditor. Hence, when I write JavaFX specific code, I cannot delete the Swing specific code,
> causing a huge amount of duplication, and thus dirty code.

I really appreciate your work, but I think we should keep the Swing
code, because I believe (without knowing it) that JavaFX would be too
slow on systems without OpenGL support (e.g. older systems, Linux
systems).
--> Is it feasible to support two (UI) renderers?

> In short, I am unconfident that I would
> be able to “support and fix every single issue which could come as a consequence of this
> development.”
>         In case I was not clear, I am not asking for a fix to the bugs I mentioned above. I know
> most of you are more experienced programmers than I am.

Not me ;-)

> Hence, I would like advice whether you would
> continue in my situation or work on something easier and come back to
> this effort later?

Cheers and Best Regards,
--
Felix Natter
Reply | Threaded
Open this post in threaded view
|

Re: Stuck on FreeplaneFX

Dimitry Polivaev
Administrator
In reply to this post by theworldbright
Hello team,

thank you Kent for sharing your state. I was always appreciating your work. You have never told me
anything about your programming background and I never asked you about it but I always felt
confident about your skills and judgment.

Also I was always thinking that JavaFX is a pretty big story which costs a lot of efforts, and that
there is a high risk of getting stuck at some point here. I think the question is not about "giving
up" but about how to deal with our limited resources namely our time and our motivation.

I think that motivation that is the believe to do sensible work is the most important resource for
getting any kind of work or try done. It is like momentum which keeps us going, and it is getting
lost by frustration if there is no visible success for a long time. In my view every motivating
action is helpful. If you think there are tasks which you could accomplish in a shorter time and it
could make you happy, there is neither need nor reasons to postpone them until JavaFX is finished or
you loose your confidence and your interest.

(However I do not see any unsolvable problems with the technical part of your questions: as long as
we have both JavaFX and Swing related parts some duplication is unavoidable, and it can be removed
automatically if Swing is completely replaced by JavaFX. And I do not think we need JavaFX for
option pane only, rather for the map itself. I think that at the time it happen computers are even
more powerful than now and I do not think we have to keep the both. Certainly using unstable
libraries is a risk, and if they continue to make troubles I think we should not keep using them.)

Now about me.

Actually the main job of a project lead is to help other project members to develop inspiration and
motivation, and I feel I am not successful at doing it. The opposite is the case, I have lost my
momentum myself. I do not feel great any more about what I am doing myself. As I started Freeplane I
wanted to do better work than I could ever do in FreeMind and have a good running project. There
have been some ideas about design and mind mapping I wanted to try. And it was running pretty well.

Now I feel like being out of ideas.

Current freeplane code base is still a kind of mess with possibility to make it better at any place
we need it. What for? What can be significantly changed by adding a dotted edge line or another
layout? What is the vision behind it? The less I understand it the less I can help to motivate you
for doing any work.

So I think that both Kent and me but also the rest of the Team need some arguments and motivating
new goals to continue spending our time and efforts. I would appreciate your help here.

Thank you!

Dimitry

> Hello everyone,
>
>         It has been around three months since I started to attempt a JavaFX port for Freeplane. At
> this point, I would appreciate a quick two minute evaluation on my implementation of the JavaFX
> version of the OptionPanel. How many problems/bugs can you immediately see?
>         I am not exactly a fan of giving up but I feel like pushing forward would be fairly
> inefficient. Even for a simple task of porting the OptionPanel, I have already noticed two bugs that
> I have no idea how to fix:
>
> 1) Occasionally the property panel just shows up blank
> 2) A huge vertical space is created when text is typed into the search box
>
>         Both problems are the result of me directly modifying the library, ControlsFX, which I have
> been using to construct the accordion pane. Moreover, they do not happen consistently, and I am sure
> there are more bugs lurking around. I feel like I would never be reasonably assured that the
> implementation is stable.
>         Furthermore, most of the classes under freeplane/core/resources/components are needed by the
> StyleEditor. Hence, when I write JavaFX specific code, I cannot delete the Swing specific code,
> causing a huge amount of duplication, and thus dirty code. In short, I am unconfident that I would
> be able to “support and fix every single issue which could come as a consequence of this development.”
>         In case I was not clear, I am not asking for a fix to the bugs I mentioned above. I know
> most of you are more experienced programmers than I am. Hence, I would like advice whether you would
> continue in my situation or work on something easier and come back to this effort later?
>
> Regards,
>
> Kent