knagano.blogspot.com
hi this is keiichiro nagano living in tokyo, japan.2005-10-27
portable dumper, revisited.
now i don't feel like describing what is Meadow, what is GNU Emacs, and in the first place, what is portable dumper. just google them. one day i should do it especially on portable dumper, but it's not the time.
our software Meadow is an text editor for win32 based on the GNU Emacs with many extensions. portable dumper is one of the extensions. recently FSF people has changed internal data structure of the editor (or to say, the lisp engine) for the next 22 release. the problem is that portable dumper is very vulnerable for that kind of changes. my goal is to sync up the dumper, call it pdump, with the changes donw on the original.
three years ago i implemented pdump first for Emacs 21.1 on Linux machine and then ported it for Meadow 1.99. i just suceeded only the Meadow team to take my code. now the problem is the code and the environment. i should use pdump code in the Meadow codebase which is timely maintained, but i think i should firstly work on the GNU one cause
from the viewpoint of debugging i'm not good at win32. pdump could be called as a kind of memory-to-disk gc subsystem and this kind of code might not be completed without a neat debugger. three years ago i used gdb. and got the goal. now i think i can beat the mysterious bugs with it again.
i would do this as follows. (1) pick the latest pdump code from the Meadow codebase. (2) apply it for GNU Emacs 22-pre codebase. (3) sync up, compile and debug the pdump code. (4) backport to the Meadow codebase.
here i would log states of my work.
our software Meadow is an text editor for win32 based on the GNU Emacs with many extensions. portable dumper is one of the extensions. recently FSF people has changed internal data structure of the editor (or to say, the lisp engine) for the next 22 release. the problem is that portable dumper is very vulnerable for that kind of changes. my goal is to sync up the dumper, call it pdump, with the changes donw on the original.
three years ago i implemented pdump first for Emacs 21.1 on Linux machine and then ported it for Meadow 1.99. i just suceeded only the Meadow team to take my code. now the problem is the code and the environment. i should use pdump code in the Meadow codebase which is timely maintained, but i think i should firstly work on the GNU one cause
from the viewpoint of debugging i'm not good at win32. pdump could be called as a kind of memory-to-disk gc subsystem and this kind of code might not be completed without a neat debugger. three years ago i used gdb. and got the goal. now i think i can beat the mysterious bugs with it again.
i would do this as follows. (1) pick the latest pdump code from the Meadow codebase. (2) apply it for GNU Emacs 22-pre codebase. (3) sync up, compile and debug the pdump code. (4) backport to the Meadow codebase.
here i would log states of my work.
posted by knagano # 20:36
Archives
