Re: eecs bugs or euphoria bugs
On Tue, 25 Jul 2000 15:32:02 +0200, =?iso-8859-2?B?qWtvZGE=?=
<tone.skoda at SIOL.NET> wrote:
>first two bugs i think are bugs in euphoria interpreter:
>
>BUG # 1:
>in EE:CS editor, when you try to run program that uses more complex
graphics, like bump.ex (which i think is the best graphics demo in
euphoria) picture is not like it should be. i think there is bug in
system_exec command, where you define something about graphics with second
parameter. i tried all three numbers, none seem to work.
I don't _think_ the problem is with EE:CS, but I never rule out the
possibility these days. How exactly are you using system_exec()? The help
file states clearly that that the command must be a valid executable (not
just any DOS command), and that system_exec() is a function, not a
procedure like system(), so it needs to be used like this:
integer errorlevel
errorlevel = system_exec("somecommand.exe /parameters", 2)
>BUG # 2: current_dir bug: inside editor you run program. if that program
then tries to open file just by name, like fn=open("file.txt") which exists
in it's current directory, it cant open it, but it should. this is because
current_dir is the one where eecs runs and not where current program runs.
This is a genuine bug, but thankfully it's not a serious one.
Use full pathnames if you find it really annoying.
The bugfix will probably consist of me adding "lfnchdir(currentDir)" before
the ex/exw system-call if-statement, so if you want to run off of the
source code rather than the compiled version, you could probably add the
fix yourself...
Sorry about the bug tho...
>another little bug: if you open file which has last line blank and with
null length (just '\n' character), it is not read.
Which mode does it happen in? I can't seem to recreate this bug even when I
create a 1 byte file containing a '\n'. It loads fine.
Carl
|
Not Categorized, Please Help
|
|
