flat assembler
Message board for the users of flat assembler.
Index
> IDE Development > Fresh 1.0.09 pre-alpha |
Author |
|
scientica 15 Sep 2003, 14:28
Here is a feature fix (not bug fix, because there was no bug, rather a feature not implented ):
Now the ex styles are checked too. [EDIT]Outdated attachement removed. There should be newer version. possibly in another thread.[/EDIT] _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
15 Sep 2003, 14:28 |
|
scientica 15 Sep 2003, 17:44
Here is the current htmldocs (extract to any folder, if you want the links to the sources to work specify the locaiton of the fresh folder in the text field that shows when settings are shown, for instance if you have the fresh folder in:
D:\Dev\FASM\FreshLast1_0_09\ then the text field should contain this line: file:///D:\Dev\FASM\FreshLast1_0_09\ (note the bold typed parts) so far these files are "done": include\libs\form.asm include\libs\gui.asm include\libs\strlib.asm
_________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||||||||||
15 Sep 2003, 17:44 |
|
JohnFound 15 Sep 2003, 18:38
Hi, scientica.
Hm, it works good for me. But, do you know, I still think that all this is too hard to maintain. At the end we will have hundreds or thousends of procedures, structures and constants for documenting. In FPR, the JavaScript code is too mixed with description codes. IMO we need some more "separated" reference system: "Data base" and "interface engine". For example: plain text files for information, maybe with some easy tags, and some (I don't know is JavaScript code can read from external file) engine that will format the output. Even I think, why not, (Hm, this is not so bad idea... ) to make the header comments in the source files in some standard format and to try to extract the help information directly from source files, without using separate data files... Regards. |
|||
15 Sep 2003, 18:38 |
|
scientica 15 Sep 2003, 20:29
JohnFound wrote: In FPR, the JavaScript code is too mixed with description codes. IMO we need some more "separated" reference system: "Data base" and "interface engine". For example: plain text files for information, maybe with some easy tags, and some (I don't know is JavaScript code can read from external file) engine that will format the output. Java script can't read files, there are workarounds (most common side effect = headace). Perhaps a Java based solution is more efficient, as I'm having java as one of my subects in school and where a part of that couse is a project work. If you think that a java based solution would be good (java can read (and write) files), plus that I think that there is a "real" data base support for java. Your idea about in source comment/reference, good idea, can be done, but then there must be a strict stanard for the comments which would end up in the output from the reference interface. Here is one suggersion: (btw I saw Joshua's "comment fix", maybe it can be used, see 2nd suggersion for example) Code: 1) ;*********...**** ;Description: <General Description> ;[Arguments: ; <arguments> - <description>] ;[Returns: ; <return vaule + possible notes on return value>] ;[[Remarks:] <Remarks, important highlights from the general ;description>] ;[History: ;<[ ] || [+] || [-]> <when (YYYY-MM-DD HH:mm)> <who> <what>] ;*********...**** 2) ;maybe "docomment" could be a better name? comment fix macro comment ;*********...**** comment{ Description: <General Description> [Arguments: ... <[ ] || [+] || [-]> <when (YYYY-MM-DD HH:mm)> <who> <what>] } *********...**** It's late and it's school tomorrow, so good nigth. _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
15 Sep 2003, 20:29 |
|
scientica 19 Sep 2003, 14:31
Here is how far I've gotten on the styled editor:
I've begun work with making the un/checking of styles do it, I use the lParam of the listview item to store the value of the equate (best solution I could came up with) - lot's of debug code (and anoying message boxes, but debugging code isn't the Mona-Lisa painting ), the .buff is temorary, so the use of NumToStrU (will (try) to implent it instead of wsprinf when there is a "wrapper" for it). Should I use the PRM_UPDATEPROP message to update the prop editor after modifying the style(s)? btw, any comments on above post [EDIT]Uppdated attachment!!![/EDIT] [EDIT]Outdated attachement removed. There should be newer version. possibly in another thread.[/EDIT] _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
19 Sep 2003, 14:31 |
|
JohnFound 19 Sep 2003, 16:50
Hi scientica.
Post the values for LV_CHECKED and LV_UNCHECKED, please. They are not in my set of Win includes. |
|||
19 Sep 2003, 16:50 |
|
scientica 19 Sep 2003, 16:52
Ok, I'm doing something wrong. Somewhere after/in UppdateFlags completes the system tries to do something at eip=0, crashing Fresh, is it my SendMessage that has the wrong params that causes it or is there some other mistake from my side?
Howto reproduce: Create a form, open styles, un/check any style, klick ok, -sigh- oops, here is the .inc file (i modifyed that one too, forgot it). the equates aren't official afaik had to look in C++ macros for the values. [EDIT]Outdated attachement removed. There should be newer version. possibly in another thread.[/EDIT] _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
19 Sep 2003, 16:52 |
|
JohnFound 19 Sep 2003, 18:27
Hi, scientica.
Look at: SetFlag and ClearFlag - there was an error that cause the GPF. I fixed it but there are some more bugs. For example, the computing of FlagtoSet/Clear is wrong. regards. BTW: User "int3" for breakpoints. I dunno why, but my OllyDebug don't stop on "int 3". There was an article about this on the board... [EDIT]Outdated attachement removed. There should be newer version. possibly in another thread.[/EDIT] Last edited by JohnFound on 22 Sep 2003, 09:39; edited 1 time in total |
|||
19 Sep 2003, 18:27 |
|
scientica 19 Sep 2003, 20:46
JohnFound wrote: BTW: User "int3" for breakpoints. I dunno why, but my OllyDebug don't stop on "int 3". There was an article about this on the board... Didn't know of "int3", does it generate db CCh? Olly disbehaves on that as well as the two byte version (don't know why, but sometimes olly even (tries to) mess with registers when bumping into int 3 during debugging). I'm going to get some sleep (I know a bad habbit ) now and will look more on the sources to morrow. (Ran a quick compile, ok, I see one thing that needs to be fixed) _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
19 Sep 2003, 20:46 |
|
roticv 20 Sep 2003, 04:07
haha..
"int3" generates 0CCh in fasm while "int 3" generates 0CDh, 03h. Just something I noticed. Ollydbg seem to not work well with 0CDh, 03h. |
|||
20 Sep 2003, 04:07 |
|
scientica 20 Sep 2003, 12:12
The flag bug was the fact that I neglected to preserve ecx when calling apis (they may trash it since it's not esi, edi, ebx or ebp).
It's fixed now (at least when doing quick tests ) [EDIT]Outdated attachement removed. There should be newer version. possibly in another thread.[/EDIT] _________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||
20 Sep 2003, 12:12 |
|
scientica 20 Sep 2003, 14:49
Here is my work so far, I've ran into some big problem -- I just can't get the styles to apply propperly.
First of all I'd like to explain how it works: When a style is set or unset, the style is updated, for both set and unset of flags UppdateFlags is called. UpdateFlags calls ApplyStyleChange, which is the funciton that determinates if a style will be visible updated for the control, in the future I plan to let UpdateFlags deal with "inter-flag depencies" (see end of styleeditor.asm, some flags affect each other). The ApplyStyleChange, which resides in ApplyStyles.asm, looks for the style which has been changed in a table (currently all styles are applyable - will fix it later) for the controll, not all styles can be visibly applied - or would cause errors/problems if set. This function in calls either SetControllStyle or ClearControllStyle, these functions actually apply the style to the controll, these functions should (IMO) reside in the form lib, I've added dummy procs for them in ApplyStyles.asm, so that it will compile. I've failed to apply the styles to the controlls (got wierd side effects, for instance some buttons became windows with scrollbars when selected). I'm not done with parts of the ApplyStyles.asm, but if some one could get the SetControllStyle and ClearControllStyle to work when I complete the ApplyStyles then it would be great. I've used the Button class for testing, setting/unsetting the BS_FLAT style, sonce I feel that that's the most harm less, and it would be quick to see if it works. btw, "int3" will cause olly to set eip to the instruction above it -- don't forget to change eip in olly when "launching" it with int3.
_________________ ... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself. - Bradley Kuhn |
|||||||||||
20 Sep 2003, 14:49 |
|
JohnFound 20 Sep 2003, 15:03
At first glance, it trashes the TBaseWin structures of all windows in the tree. I will look in the sources.
BTW: Today I will post some work I made this days. It's about controls. The main idea is that controls must be loaded from some kind of libraries on run of the Fresh, not to be hard coded in the fresh source. So, please, read it, because it will concerns, your work too. regards. |
|||
20 Sep 2003, 15:03 |
|
< Last Thread | Next Thread > |
Forum Rules:
|
Copyright © 1999-2024, Tomasz Grysztar. Also on GitHub, YouTube.
Website powered by rwasa.