Latest experimentation and what direction for VCC

[Thomas Sant'ana]

Last couple weeks I've been working somewhat disperse on several ideas. Among them:

- Easier update. Goal: get the latest version when you start the program so you don't have to download, restart, download patch, restart.

- Chrome Extension: 

- Web UI for the combat

VCC 1.7.2 metrics information show that very few people are using VCC. This was really disappointing. This project is something I do on my spare time to sharpen my programming skills and learn new technologies. Given that, I'm still willing to work on this project and try out new ideas and technologies.

So my real target now is moving the VCC UI away from Swing toward HTML5 + Javascript. This seems to be a good move because I want to learn this, it works on mobile devices, and is the Common Tongue of the internet these days. 

The following link points to an experiment with AngularJS and Twitter's Bootstrap them. This is just a rough cut with some of the navigation ideas. Works on the iPad too. 

http://www.exnebula.org/files/ui-mock/src/index.html

However before anyone gets too excited there we must do a really big change on the Tracker core logic. I will still keep the tracker running in Scala and on the "server side". But the current logic has the assumption that the tracker can reach out to the UI synchronously in the same process. To use JS we need to deal with the fact that the UI is remote and not synchronous. This is a somewhat big refactor.

What I really want to do is develop the concept of the UI in parallel to changing the backend.

The other thing I want to do is to have hang-out (web conferences) with you guys to discuss the ideas directly. 

Once this migration is done we can tackle other things like resource management (power, surges), and other improvement.

Thomas

[Vladimir B. Grebenschikov]

On Aug 16, 2012, at 20:51 , Thomas Sant'ana <mail...@gmail.com> wrote:

- Web UI for the combat

VCC 1.7.2 metrics information show that very few people are using VCC. This was really disappointing. This project is something I do on my spare time to sharpen my programming skills and learn new technologies. Given that, I'm still willing to work on this project and try out new ideas and technologies.

So my real target now is moving the VCC UI away from Swing toward HTML5 + Javascript. This seems to be a good move because I want to learn this, it works on mobile devices, and is the Common Tongue of the internet these days. 

The following link points to an experiment with AngularJS and Twitter's Bootstrap them. This is just a rough cut with some of the navigation ideas. Works on the iPad too. 

http://www.exnebula.org/files/ui-mock/src/index.html

However before anyone gets too excited there we must do a really big change on the Tracker core logic. I will still keep the tracker running in Scala and on the "server side". But the current logic has the assumption that the tracker can reach out to the UI synchronously in the same process. To use JS we need to deal with the fact that the UI is remote and not synchronous. This is a somewhat big refactor.

Well HTML5+JS looks like a trend now, although, for rich applications (like vcc) it requires much work.

What I really want to do is develop the concept of the UI in parallel to changing the backend.

Di you think it is really possible with acceptable quality?

The other thing I want to do is to have hang-out (web conferences) with you guys to discuss the ideas directly.

Once this migration is done we can tackle other things like resource management (power, surges), and other improvement.

Thomas

--
You received this message because you are subscribed to the Google Groups "Virtual Combat Cards Development" group.
To post to this group, send email to vcc...@googlegroups.com.
To unsubscribe from this group, send email to vcc-dev+u...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/vcc-dev?hl=en.

--

Vladimir B. Grebenschikov

vgreben...@gmail.com

[Thomas Sant'ana]

On Mon, Aug 20, 2012 at 12:48 PM, Vladimir B. Grebenschikov <vgreben...@gmail.com> wrote:

On Aug 16, 2012, at 20:51 , Thomas Sant'ana <mail...@gmail.com> wrote:

- Web UI for the combat

VCC 1.7.2 metrics information show that very few people are using VCC. This was really disappointing. This project is something I do on my spare time to sharpen my programming skills and learn new technologies. Given that, I'm still willing to work on this project and try out new ideas and technologies.

So my real target now is moving the VCC UI away from Swing toward HTML5 + Javascript. This seems to be a good move because I want to learn this, it works on mobile devices, and is the Common Tongue of the internet these days. 

The following link points to an experiment with AngularJS and Twitter's Bootstrap them. This is just a rough cut with some of the navigation ideas. Works on the iPad too. 

http://www.exnebula.org/files/ui-mock/src/index.html

However before anyone gets too excited there we must do a really big change on the Tracker core logic. I will still keep the tracker running in Scala and on the "server side". But the current logic has the assumption that the tracker can reach out to the UI synchronously in the same process. To use JS we need to deal with the fact that the UI is remote and not synchronous. This is a somewhat big refactor.

Well HTML5+JS looks like a trend now, although, for rich applications (like vcc) it requires much work.

Thanks for the comment. There are 3 major tasks to handle:

1) Make tracker REST + JSON friendly 

2) Improve compendium so we don't lose data (like powers), this is critical to getting new things done (like tracking powers).

3) Improve update, this is not mandatory but would make changing the application much faster.

Although I listed them like this I think the order is reverse. Not having all the information from the compendium is bad. Having complicated update means we can't release constantly. I think 2 and 3 could be the core of a 2.0 release, since this would break a lot of things for the users. Once we've done that we can start getting other benefits.

 

What I really want to do is develop the concept of the UI in parallel to changing the backend.

Di you think it is really possible with acceptable quality?

Yes. It will take some time but I think we could make it much better.

Quality breaks down to two parts (the tracker logic) which is good and would stay highly tested in scala; and the HTML5 UI, if the asynchronous nature of the web does not make it bad it should work.

My current concept is to have a Activities/Wizard based approach. Example of an attack:

1. Select some targets  (just click on them)
2. Select power to use (click on power name or a small use button)
3. This makes UI switch to a wizard
4. Select hit/miss/critical for each target (here we can say what is the target dice roll for each target just to help out)
5. Compose effects for hit/miss/critical, or get one from past uses of the power (this again can be more complete)
6. UI now becomes modal as DM provides additional input (ex, resistance, immunity,  dwarfs resistance to tripping etc)
7. Back to combat screen with updated status

Of course we can start with a  more VCC 1.x approach then evolve to this. Going with the Wizard/Activities means I can use the same screen space for different tasks. That's a bit of what I tried to capture in the mock.

To me the pros are:

 - Tablet friendly

 - Current technology should be easier to get support for design and improvement (let's face it Scala+Swing is a bit obscure these days)

 -  Possibly better UX. 

Cons: 

 - A lot of work, specially if we have to keep backwards compatible.

I'm really split on doing a big bang approach or evolving slowly. I hate the current compendium format and update mechanism. They are old, ill designed and complicated.

Thomas