Re: Alpha Test - A standalone windows executable
- Hi Jon,
Nice ! :P
One thing I forgot to refer yesterday was that it would be nice that the container is protected against corruption by appending to the beginning headers of the container some md5sum of the whole "cake", that would then be checked at startup just to be sure that we dont have "bad" contents inside.
Just some more cents ... :-)
--- In email@example.com, Jon Green <jon@...> wrote:
> azynheira wrote:
> > Hi Jon, Just some thoughts on the Win32 version: 1) I maybe trying to be
> > perfectionist, but do you think that it could also be a good approach to put
> > the bfs into a resource reference. This way one could replace it as wish.
> > 2) Another approach would be to append the bfs to the end of the executable
> > and at run time the executable would search for the bfs at as trailing
> > information append to the end of the exe file and load it. This way it would
> > also allow some flexible plug&play bfs. If nothing was found, then it would
> > be regular stripped me32.exe.
> Well done Pedro - top marks; this is exactly what we should do.
> This means that we do not need to do a build to change the macros and any Joe
> can easily customize. Also the executable is standard, if there is anything on
> the end then you can run it, if it is missing then run as normal. Also the
> stuff on the end can be huge i.e. spelling dictionaries, the works! This also
> satisfies Bryan's requirements of being able to load up with his own private files.
> From what I have at the moment to this is just a little step - a small tweak
> to the container format will mean I can access it from the bottom of the file
> to find the root node entry point and then I have the root directory.
> Well done - exactly why I alpha'ed it and what I should have thought of - guess
> I was not thinking that far ahead just to proof of concept.
> Thanks, I will pursue this path as the next revision.
> > What do you think about this?
> > Regards, Pedro