Centum

12:42am: Tuples
did a tuples module
too lazy to post it right now, tomorrow

3:51pm: Stuff
Fixed a problem in the array operations modules. Obviously a big issue since it is in the main release. Have been in contact with matai, who has lots of questions about centum.

5:01pm: Finsihed with list box for now
finished up the list control and sent it off to adrien for review. Also posed it onto sourceforge.

3:28pm: feelin listy
Finished the rest of the list box dll today. Will write header #dll and test tomorrow.

11:40pm: more listy
added more list box functions today. hirt my hand so i may be a bit slow on the coding. ouchie.

2:06pm: More on windows
I have no excuses for my long haitus.
Today I upgraded the windows module slightly so that other dlls could insert a custom control and wrote the list box creation methods in a new module.

4:50pm: threadfin
Implemented _THREADFIN
version 0.1.4 should come out today or in a few days.

6:01pm: #include madness
Today I added a macro to the standard macro file that will append .txt to the end of an include statment if it does not contain a '.'. Thus #include console and #include console.txt are equivalent. Yes! 4 saved charachters of typing!!.

9:16pm: Done with docs
I finished the documentation today. Tomorrow i'll add a cover page and send it out.

11:37pm: whoopsies
I forgot to post yesterday, sorry. Partly that is becuase I have been falling a little behind on my work. I have finished all the documentation except bytecodes.cpp. however I did not go through and check ThingManipulation or Function calls as well as I would have liked.

7:08pm: too little
I am ashamed to say that today i was distracted and accompilished too little work. I documented a few functions of ThingManipulation.cpp and did a little code revising but that was it.

11:26pm: More and more
Documented the FunctionCalls.cpp and began some of ThingManipulation.cpp. Had a really really stupid error. And after many hours of frustrating work tracked it down solely by examination, yay me.

8:06pm: mucho doco
did a whole lot of documentation today, only ThingManipulation.cpp, FunctionCalls.cpp, and bytecodes.cpp remaining. However these are large files, so I am not sure that I will be able to finish tomorrow, altough that would be nice.

7:59pm: further documentation
Today I documented the CDllManager.cpp file, the CStringManager.cpp file, and the CProcessData file. I also fixed a small bug in the default loader that would cause an error in debug mode. I also re-wrote the memory manager to only keep an numerical count instead of a list of the pointers, which imrproved preformance drastically. The next performance enhancements will probably come from reducing calls to the string manager.

5:23pm: More docs
Since i was busy I only documented the 4 array classes, CThingArray, LongArray, Int8Array, and PointerArray because it was mostly copy-paste.

6:33pm: Doxygen
Today I began documenting the Centum API / shared code base with Doxygen, a very awesome tool. So fara I have documented the CThing.cpp, StackOperations.cpp, and ThingReferences.cpp files. I will do more tomorrow, altough tuesdays and thursdays are always so busy I don't know how much will get finished.

5:08pm: Docs
Today I wrote appendix A on OOP programming in Centum version 0.1.4. Gotta run.

12:11pm: Internet Down
There was no post yesterday due to network outages. Finished the registry module and the thread control module, as well as modifications to the windows module and the quick type module. That is all for the moment.

10:42pm: Registry
I made a small change the the sleepy IDE at Adrien's request which allows you to double click an entry in the dictionary list box and have it inserted in the edit space. I also finished the registry module (well minus the documentation) which worked on the first try. I am very happy with the registry module since it makes the resistry much much simpler to use, usually requiring only 2 or 3 parameters for a function instead of five or more. Anyways ill try to document and post it to sourceforge before the night is out.

10:02pm: Double Click
Due to my move back and forth today and seeing my friend mike (and sin city for only $1.50) I didnt get much time to program today (i know excuses excuses). I did however add double click selection to the IDE because its lack was driving me nuts ... !!!

11:44pm: File Dialogs
Today I whipped out a file dialgos module, which is one of the requirements that Adrien had. It wasn't that hard, and I think it is a very good module to have. However it also means that I will be pushing back the threading module and the registry module for a bit while I work on the IDE reqs. I am anxious for diana to come back, not only because I miss her :-), but because I want to post some new news and documentation to the website. I also made a few miscelanious fixes to the Sleepy IDE.

9:55pm: Inheritance now possible
I have modified isA? so that you can declare a class in such a way so that functions chedcking for an instance of the inherited class will accept a rebi that inherits from/extends that parent rebi. Here are is an example class list (p.s. you create a type function that returns the classes of the parent(s) to make inheritance work)

#include tests.txt

myclassA
    {
    }

myclassB
    {
    }

myclassC
    {
    static type
        {
        return myclassB myclassA
        }
    }

myclassD
    {
    }

myclassE
    {
    static type
        {
        return myclassD
        }
    }

myclassF
    {
    static type
        {
        return myclassE
        }
    }

In other news I have had some inexplicable problems with the Sleepy IDE today. It crashed twice when trying to put two newlines at the beggining of a file and then never again. It also refused to display the console window for this project, despite the fact that it works fine for every other project. I fear that I may not be able to release it in the manual if it fails to create the console sometimes, else some users may think that their installation is not functioning properly ARG!!

10:45pm: Finished with tie and untie?
I think I finished with tie and untie today. I am of couse eternally grateful (to me) for may garbage collection system that can collect circular references. Wihtout it I think any simple soultion to this problem would have been impossible. Before I continue I must also note that centum launcher, default loader, macro compiler, acontrol folw, and string additions A had to be either modified or re-compiled. Now to get to what tie and untie do. Basically local variables (and local variables only, not mys or statics) can be "tied" so that they are treated as a single variable, meaning that if on is modified so is the other, even by aliasing, which = will not do. Any number of variables can be tied together, and removing one variable from a tie of three or more will not dissolve the tie. The motivation behind this is that in centum inheritance is accomplished by having a copy of the parent class. However if you wish to be able to pass the class/rebi to functions expecting its parent you must declare in it duplicate local varaibles so that functions using it can do name:var ... ect. You could allways pass the included object directly of course, but then any redefined member function would not be called in the function it was pased to, only the originals. With tie in the constructor you can connect these new local varaibles with the ones in the "parent" rebi so that they stay in sync and allow you to pass the "inherited" rebi in its place. With a few modifications to isA? that will allows derived rebi to idenftify themselves as their parent to tests, I will create a section on OOP in Centum, which will be greatly enhanced by these. The syntax for tie is tie var1 var2 var3 ... and untie is untie varname variables do not need to be untied if they are tied, however being tied to another variable counts as a reference and will prevent the tied object from being released until they all can be collected at once.

11:27pm: Broken & Fixed
Finished wrtitng the (still untested) code for the tie and untie commands, as well as required modifications to the findinthingsub function and the garbage collection functions. The code I wrote made a mysretious break, and after about an hour and a half i realized that is was a simple scoping mistake. Even I can be stupid i guess. I spent so much time looking for the big errors that I didn't notice such a small detail. With luck (lots of luck) tie and untie will be finished tomorrow.

11:04pm: Tie and Untie
Today I began work on the tie and untie instructions that will aid the use od pseudo-inheritance. They might be good for other thigs as well, who knows. In any case I have thd bad feeling that work will be slow until I get back to school. When I have too much free time I never seem to have the energy to work on my own projects. At least I brought some books to read.