cs161akx@sdcc10.ucsd.EDU (cs161akx) (01/25/89)
Can anyone tell me what differences are between Zortech C++ v1.06 and Zortech C++ v1.07 ? The reason I ask for these differences is because I bought my Zortech C++ v1.06 and a week after the Zortech people shipped out a new version v1.07. Thanks in advance. PS: I prefer if you could post your opinion in the news group rather than e-mail to me. Thanks again.
bright@Data-IO.COM (Walter Bright) (01/26/89)
In article <69@sdcc10.ucsd.EDU> cs161akx@sdcc10.ucsd.edu () writes: >Can anyone tell me what differences are between Zortech C++ v1.06 >and Zortech C++ v1.07 ? >PS: I prefer if you could post your opinion >in the news group rather than e-mail to me. The following is extracted from the READ.ME that comes with 1.07: ------------------ 1.07 Notes ------------------- CODEVIEW SUPPORT At last! Symbolic information is output and passed to Microsoft CodeView! Here's how to do it: 1. Compile with: -gl line number info only -gs symbolic info only -g line number and symbolic info -S always generate stack frame (otherwise CodeView exhibits some strange behavior) 2. Link with the Microsoft Linker (not Zortech's!) with: /CO specify CodeView Limitations and problems: o Zortech C sometimes assigns variables to the registers AX,BX,CX,DX, and sometimes the byte registers. There is no way to tell CodeView about this, the variable locations come up as ??. o In order to conserve the size of the symbol table, typedefs and enums are not emitted. Only the minimal type information for variables actually used is emitted. o Because the optimizer moves variables and sections of code around, there becomes only one scope per function. Thus, avoid using the same variable names in nested scopes if you want to use the debugger on them. o Use the -S switch to cause the code generator to always generate a stack frame for each function if you want symbolic debugging. CodeView seems to expect this. ZTC.COM will automatically invoke -S if the -g or -gs switch was specified. o C++ stuff is 'converted' to C-ish types, because CodeView doesn't know anything about C++. This makes symbolic debugging of C++ rather ugly. o This was tested with the CodeView that came with MASM 5.1. ZTC.COM Now recognizes files with a .CXX extension as being C++ files. .CXX is used by Glockenspiel C++. Added -S switch to cause the code generator to always generate a standard stack frame. Added -C switch to prevent inline function expansion in C++ files. Added -L switch to run LINK.EXE instead of BLINK.EXE when linking. C COMPILER Fixed a problem with expanding circular nested macros. Fixed cast to void bugs. Casting an expression to void can be used to suppress the warning about the value of an expression not being used. C++ COMPILER Added -C switch to prevent inline function expansion. This is useful if you're debugging the logic of your program, since inline expansion can be confusing when you look at it with the debugger. It's also interesting to use this switch to look and see if you're getting any real speed/size improvements by using inlines! Fixed problems with initializing references. Fixed problems where destructors sometimes were called in the wrong place. BLINK Now is called BLINK.EXE instead of LINK.EXE, which caused annoyance to many people. Use MS-LINK if linking for CodeView. If BLINK encounters CodeView records, it will ignore them. C LIBRARY Added sleep(), msleep() and usleep(). Added _rotl(), _rotr(), _lrotl(), _lrotr(). Fixed problem with 8087 support. FLASH GRAPHICS Added support for CGA in medium resolution mode, and the IBM 8514A graphics adapter. The new initialization routines are: fg_init_cgamedres() fg_init_8514a() The "FG_" is no longer necessary as the value of the environment string for FG_DISPLAY used by fg_init_all(). Thus, fg_init_all() supports the following: a) b) c) * HERC FG_HERCHALF fg_init_herc() HERCFULL FG_HERCFULL fg_init_herc() CGAHIRES FG_CGAHIRES fg_init_cga() CGAMEDRES FG_CGAMEDRES fg_init_cgamedres() EGAECD FG_EGAECD fg_init_egaecd() VGA11 FG_VGA11 fg_init_vga11() VGA12 FG_VGA12 fg_init_vga12() EGAMONO FG_EGAMONO fg_init_egamono() EGACOLOR FG_EGACOLOR fg_init_egacolor() TOSHIBA FG_TOSHIBA fg_init_toshiba() 8514A FG_8514A fg_init_8514a() a) Value for environment variable FG_DISPLAY. b) Return value from fg_init_all() if successful initialization. Values are defined in fg.h. c) Initialization routine actually called by fg_init_all(). * fg_init_herc() determines if FG_HERCHALF or FG_HERCFULL is returned based on testing if two pages is possible.
raw@ushiva.UUCP (Roland Wilcher) (01/26/89)
In article <1843@dataio.Data-IO.COM> bright@dataio.Data-IO.COM (Walter Bright) writes: >In article <69@sdcc10.ucsd.EDU> cs161akx@sdcc10.ucsd.edu () writes: >------------------ 1.07 Notes ------------------- >CODEVIEW SUPPORT > At last! Symbolic information is output and passed to Microsoft > CodeView! Here's how to do it: Please excuse a possibly (likely) stupid question but... When using codeview I found that the only variables I could examine were in main. All vars in function calls were unknown to CV although I could single step at the source level. Even parameters to the function could not be examined. Did I blow it somehow ? I have 1.07 . --- "If aint broke, don't fix it" raw@neologic%hal.crwu.edu -Bert lance {hal,decvax}!ncoast!ushiva!raw Roland A. Wilcher ------------------------------------------------------------------------------