1. Greetings again all
Granted, there are none of you who remember me..
About 2 years ago I picked up euphoria for linux and began coding a mud.
With the help of some of you, I have a wonderful, smooth running game on
the internet! I have told everyone I could about euphoria, simply
because I found it to be such a good language for me to code in.
Now, I have been completely out of touch with any new developments in
Euphoria for about the last year or so (The Euphoria to C translator had
JUST come out).
Now the question..
The only problem I still have with euphoria is the lack of error
trapping. One little coding error and down goes my game. Has there, by
any chance, been any developments in this area? Is it possible now in
euphoria to trap the errors somehow and avoid the crashes?
Thanks in advance for any assistance.
Paul
2. Re: Greetings again all
Hello, Paul
----------
> Îò: draegur at earthlink.net
> Êîìó: EUforum <EUforum at topica.com>
> Òåìà: Greetings again all
> Äàòà: 30 èþëÿ 2002 ã. 0:52
>
> Granted, there are none of you who remember me..
>
> About 2 years ago I picked up euphoria for linux and began coding a mud.
>
> With the help of some of you, I have a wonderful, smooth running game on
>
> the internet! I have told everyone I could about euphoria, simply
> because I found it to be such a good language for me to code in.
>
> Now, I have been completely out of touch with any new developments in
> Euphoria for about the last year or so (The Euphoria to C translator had
>
> JUST come out).
>
> Now the question..
>
> The only problem I still have with euphoria is the lack of error
> trapping. One little coding error and down goes my game. Has there, by
> any chance, been any developments in this area? Is it possible now in
> euphoria to trap the errors somehow and avoid the crashes?
>
> Thanks in advance for any assistance.
>
> Paul
Yes, Euphoria v2.3 Public Domain has no any
restrictions on trapping of the run time errors now.
Do you remember that magic 300 number ?
Just the trace feature has that restriction now.
Regards,
Igor Kachan
kinz at peterlink.ru
3. Re: Greetings again all
On 0, Matthew Lewis <matthewwalkerlewis at YAHOO.COM> wrote:
> > -----Original Message-----
> > From: draegur at earthlink.net [mailto:draegur at earthlink.net]
>
> > The only problem I still have with euphoria is the lack of error
> > trapping. One little coding error and down goes my game. Has there, by
> > any chance, been any developments in this area? Is it possible now in
> > euphoria to trap the errors somehow and avoid the crashes?
>
> There are some libs that work with exceptions of some sort.
Yes, but those are mostly for user-defined exceptions. There is no way
to
catch a builtin run-time exception except for checking the
conditions/parameters
etc. in advance.
Another hack I attempted was to modify David Cuny and Delroy Gayle's
euphoria
interpreter written in euphoria code to add this support. However,
ignoring
the bugs already contained in the simulated interpreter, this is very
slow
and kludgy. (One plus of this would be hacking the eu coded interpreter
to be an include file rather than a stand-alone program. Then
scripting,
such as exec("a = b + c"), would become more feasible.
> Also, my
> modified interpreter could help you. You can specify a crash routine. You
> can also get a dump of the symbol table, allowing you to possibly recover
> the program state after an error. You might be able to dump vars, start a
> new instance of the mud, and allow it to recover from there. Of course, I
> suppose you'd have to reestablish connections and so forth (but I'm sure you
> know a lot more about that end of it than I do)...
>
This would be a much better idea than using a simulated interpeter.
> Oh, and it would also need to be recompiled for linux (I could only get
> Borland to work for me), which I don't have. If anyone is willing, I'd be
> happy to share my modifications for you to recompile under linux/dos
> (assuming you've registered the source code already, of course :).
>
I would love to try out your modified interpreter, but I can not
because of
this restriction. (I use linux and do not that the source code.)
> Matt Lewis
>
Linux User:190064
Linux Machine:84163
--
http://fastmail.fm -- Does exactly what it says on the tin
4. Re: Greetings again all
jbrown105 at speedymail.org wrote:
> > > The only problem I still have with euphoria is the lack of error
> > > trapping. One little coding error and down goes my game. Has there, by
> > > any chance, been any developments in this area? Is it possible now in
> > > euphoria to trap the errors somehow and avoid the crashes?
I may be looking at this all cockeyed, but I'd rather let the program crash than
have all that extra processing overhead going into error-checking. I mean, when
it
comes right down to it, a programmer's job is to hang on for dear life while the
code runs. A programmer's machine's job is to crash. Sure, it would be easier if
I
didn't have to deal with crashes, but it would also be easier to overlook bad
code
or sluff it off and not bother fixing things.
Just MHO.
-Ron Tarrant
