As Thomas and myslef are now back in university, its looking like the time allowed to continue our efforts on the Object Creator/Editor tool will be slowed somewhat.

However, we do have the Object model finished and optomized for fast and ez object classes in MASM32, where much of the syntax is simular to HL syntax so you dont need to mess with crazy switches etc.

My poll question is this: Do you want me to pre-release the Objects.inc version, and all supporting example code, with a revised Creator tool able to support Inheritance?

If i do, i would probably take my conversations from here to the COM section (the forum seems to not be moving at all these day :O ), and begin helping people create classes etc, walk thu examples, and provide basic tutorial over the next couple of months (Its really ez to do, but all the same, its new too!).

I check this page about once a day, so if this poll turns out promising, i will give it my attention in the months to come. As well this should help provide some Tutorial base with actual user quesitons, rather than me guess what people might ask :)

For those worried about overhead, i have fully commented every macro with the amount of actually generated code that is inserted. As well, there is no secrets in our model, so i would be happy to explain how it works in detail if this is a concern.

So what do you think.. ready to give MASM32 Objects a shot?
Posted on 2001-09-17 21:01:20 by NaN
i will take all that is given

but

i will take no prisoners
Posted on 2001-09-17 22:52:51 by titan
I've always been curious about what you two have been up to, mostly to see if I can steal anything for CoLib. ;-)

But personally, I've just moved and have a house to repair, rebuild, redecorate.... plus my job.

Arggggg... no time to code!
Posted on 2001-09-17 23:19:07 by Ernie
Nan,

Apart from my personal prejudice against OOP programming, I think many people will be interested in the work that you and Thomas have done so if you are both heading back to university studies and will not have the same time available, it would be a good idea to post the work you have already done so that other can benefit from it.

Regards,

hutch@pbq.com.au
Posted on 2001-09-18 02:00:26 by hutch--
Yes do the pre-release.

I am also interested in what you both are doing. It seems to be interesting ... I think everyone who deals sometimes with a HLL has made his own thoughts of realising a oop-model in assembler but - hugh - it's a hard work of designing macros and data-models (RESPECT !) So many wanted to know the code inside your implementation as early as possible.

(Yes, pretend to be our SantaClaus (well the christmas-shopping-season has started already ;) )

Greetings, CALEB
Posted on 2001-09-18 02:27:40 by Caleb
One sidenote I want to make here: With this prerelease, you can use all the OOP things we've done so far and it won't differ much from the final version I think (besides some optimizations and maybe new functions), so you can do all the object stuff, but the next version of the tool will be just as important. As right now, with this prerelease, you can do everything, but you'll have to do it manually. That means you have to design prototypes, initial data sections, modify the class definition etc. All of this can be done by hand and is not so hard, but takes a lot of time if you need to do many modifications.
The tool we are now working on is totally different from the tool we will be supplying with this prerelease (it isn't even based on it, although the new tool is created with it :) ).
The current tool will only 'help you on your way'. It writes the right code to get you started but will not do any modifications.
The new tool will be able to load every existing .asm file with a class in it, and parse & process all the class stuff inside. Then you will have a nice window that looks a bit like the class view tab in Visual C++. You can add methods, variables, make modifications, change the inital values of members, see what other classes it inherits etc... All with treeviews/tabs/dialogs/etc for easy editing. The tool will not harm your existing code, it only modifies the class definition and everything related. It will neither do anything you can't do yourself or something that is impossible to understand, it's just that it saves you time and boring modifications.
So I recommend you all to look at the prerelease, work with it, ask about it, etc.. If you find it hard to work with, remember that the tool will solve that.
Like NaN said we're both back to university, but still the progress is going well. The technical implementation (the actual parsing of existing code and processing it) is about 80% finished, we'll have to build the other 20% :), the user interface and the communication between both for the tool to finish it, so stay tuned!

Thomas
Posted on 2001-09-18 12:33:34 by Thomas
I posted yes, just so I can see how you guys did it. But, I have no interest in using it because I cannot understand it, and my brain doesn't think that way. However, I might use it as a reference when I start my template project. :)
Posted on 2001-09-18 12:42:13 by Kenny
I dont want to seem contradictory to Thomas, but in my opinion its actually not all that bad to make revisions or add another method, now.

In the past it was... expecially with the old tool. But it sounds like he's overlooked some of the streamlining ideas he had :) . It now makes adding a new method the same as a high level language, except you also define the PROTO separately (just like PROC's).

But none the less, i agree with the overall message, the final tool will definitely be advantagous to use over manual editing.

I will work on packaging a 'box set' and perhaps have something posted by this weekend :)

Thanx for the good responce, (few) :) . I know in the ASM world its hard to go against the flow, and offer something that is "High level"... I hope we can keep you impressed.

PS: I will post an early discription of how our model is 'framed' now in the COM section to start people's minds on the right track..


NaN
Posted on 2001-09-18 17:05:33 by NaN
Well, NAN, i don't think that offering something high-level-like is something against the flow, especially your thing ... the fact, i believe is, as follows:
Nearly every programmer who used modern technologies uses polymorphic and oop-code. further more com is an object without many don't want to code evermore ... most of them believe when it came up it was revolutionary but it isn't this at all - it's only a way to handle and order data, functions and their relationships and it is useful to do oop in a shorter time.
If you make it good, implementations of oop in assembly need not to be slow or to be something only snops will handle with. It will help most of the masm-freaks to do a faster realising of their projects, because oop makes many things much easier if polymorphic-code is possible. Everyone who says he don't need this, does only short projects or is a real hardliner ;)

Greetings, CALEB
Posted on 2001-09-19 02:09:02 by Caleb
NaN: You are right, it's now a lot easier than the first version, although setting up the initial data table can cause some troubles, so check this part first when you get any crashes.. Objects.inc should warn you if you forgot one item, but of course it doesn't protect against the wrong order.

Thomas
Posted on 2001-09-20 09:01:18 by Thomas
Good Point!
Posted on 2001-09-20 10:31:16 by NaN
Ok...


    [*]Objects.inc v2.27
    [*]OOP Creator v1.5
    [*]OOP2HTML


    Is now all posted... you can find them in the COM section threads.

    NaN
Posted on 2001-09-20 19:34:19 by NaN