Whoa nelly! I still think you should pick one or two simple vector operations. AFTER that we can think about the user interface.
(By the way, I looked today for a Swing Grid component, but I didn't find anything I thought would work. And I don't think that we want to use a JTable. This means we'll have to cook something up, which will be a fair amount of work.) Why don't I see if I can come up with an interface or two for you in the morning? I think that might help get you started. The Sunburned Surveyor On Wed, Apr 30, 2008 at 1:46 PM, Leandro Leal Parente <[EMAIL PROTECTED]> wrote: > All right Sunburned Surveyor, > > That implementation is the most better possible for this project. We will > focus on user interface and later add a lot of procedures around that. I > will start tomorrow. > > Thanks, > Leandro Leal Parente > > 2008/4/30 Sunburned Surveyor <[EMAIL PROTECTED]>: > > > > Leandro, > > > > Stefan's advice on using an iterative development process was > > excellent. If I remember correctly the original goal for this project > > was to allow the user to develop a graphical model for data > > processing. > > > > So I will take this a step further than Stefan. Pick just ONE or TWO > > vector layer processing operations. These may be operations that are > > already implemented in OpenJUMP that we can extract, or they may be > > new operations. This will be your first module. If you decide to go > > with new operations I can help you access the OpenJUMP and JTS API. > > > > Then we need to construct the user interface that allows the user to > > construct the graphical model for processing. This could get really > > ugly, really fast, because of the potential complexity. So I think we > > start by making the user interface very constrained. Let's used a > > fixed grid. We can have a button or list that allows the user to > > assign a single input to the grid (this will be an OpenJUMP Layer), > > and then a single vector operation (one of the ones you create in your > > first modue). > > > > I am starting to get a little nervous, and I don't want to bite of > > more than you can chew. > > > > Your next assignment: Choose one or two vector operations that we can > > allow the user to add to the graphical model. The simpler these > > operations the better. > > > > I will poke around on the net to see if there is an existing Swing > > grid component that we can use. > > > > The Sunburned Surveyor > > > > OpenJUMP Developers: I imagine we will want to encapsulate the "vector > > operation" as an interface. Do we have an existing interface kicking > > around in JTS/deegree/GeoTools that can be used for this purpose? > > > > > > > > > > On Wed, Apr 30, 2008 at 4:02 AM, Leandro Leal Parente > > <[EMAIL PROTECTED]> wrote: > > > Hei Stefan > > > > > > Now I understand what is a case studies. You be propose to me a > implement a > > > little application with only 3 or 4 functions, implements a step 2) and > 3). > > > All right, I think that is a excellent idea. > > > > > > My first module is a little implementation of GMGDP with vectors > procedures > > > manipulating only layers, all right. > > > > > > Thanks, > > > Leandro Leal Parente > > > > > > 2008/4/30 Stefan Steiniger <[EMAIL PROTECTED]>: > > > > > > > > > > Hei Leandro, > > > > > > > > the outline is ok for me with steps 1-3. > > > > But > > > > I would again suggest, that > > > > a) you start with one or two case studies first on paper! > > > > > > > > b) That we use an iterative development process: > > > > that means: you start only with 3-4 functions in the first step (e.g. > > > union, intersection, ...). Then you do the first version for 2) and 3). > > > > Doing it in an iterative way will make it easier if we discover > problems > > > in a later stage. > > > > > > > > c) If all basic tasks of steps 1-3 are implemented, then we start can > add > > > more vector procedures (So this would be something like a new Step 4). > > > > > > > > For the procedures (vectors operations) we should make the > Case-Studies. > > > Look for instance in the OpenJUMP Tutorial by Uwe. Having the case > studie we > > > can define our first 4 vector operations. > > > > > > > > A couple of function we can start with are those, which are offered by > > > Tools/analysis/geometry functions... But I am not sure what you mean by > the > > > following questions: > > > > > > > > > > > > > > > > > What procedures we will implements ? Does has some procedures we can > use > > > ? > > > > > The procedures will be a .JAR ? > > > > > What Workbench datasources we will use ? > > > > > > > > > > > > > your code should be in a jar file at the very end (such as a plugin). > But > > > we need to decide on that later, when we know what are the requirements > of > > > your work > > > > > > > > Workbench datasources will be a vecor "layer" only. I would suggest. > So > > > the model should only manipulate layers (i.e. take it from one layer and > > > write it to another one - or write it to a file) > > > > > > > > > > > > > > > > > PS: I get all jump-pilot code, how I build the OpenJUMP from source > code > > > in Netbeans 6.0.1 > > > > > > > > > > > > > Unfortunately I have never used NetBeans.. so you need to find it out > by > > > yourself. But on the wiki we have some stuff for Eclipse IDE: > > > > http://openjump.org/wiki/show/How+to+use+and+make+own+Plugins > > > > > > > > stefan > > > > > > > > maybe it would be good if you communicate on some things with Nacho in > > > portugese, if it is to difficult to translate your questions. > > > > > > > > > > > > > > ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Jump-pilot-devel mailing list Jump-pilot-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/jump-pilot-devel