1. IDE error
Hi all,
I have just tried the new compiled IDE 0.15.
Each time I try to change the name of a control I get the following
error :-
MSG 193
Can not find OtherC.ini
Any ideas ?
--
Peter Blue
IT Director - Legend Interactive Ltd
Technical Director - Platinum Communications Ltd
Please visit our web sites
http://www.yes-property.co.uk - International Estate Agents
http://www.legend-i.demon.co.uk - Legend Interactive Ltd
2. Re: IDE error
Oops! Sorry, i guess in my haste i forgot to include all the files! I just
made sure the program RAN before sending it in! I'll update it as soon as
possible, check back soon!
~Greg
g.haberek at comcast.net
----- Original Message -----
From: <Peter.Blue at legend-i.demon.co.uk>
Subject: IDE error
>
> Hi all,
>
> I have just tried the new compiled IDE 0.15.
>
> Each time I try to change the name of a control I get the following
> error :-
>
> MSG 193
> Can not find OtherC.ini
>
> Any ideas ?
>
> --
> Peter Blue
> IT Director - Legend Interactive Ltd
> Technical Director - Platinum Communications Ltd
>
> Please visit our web sites
> http://www.yes-property.co.uk - International Estate Agents
> http://www.legend-i.demon.co.uk - Legend Interactive Ltd
>
>
>
>
> TOPICA - Start your own email discussion group. FREE!
>
3. IDE error
There is an error in IDE_edit.ew in the routine editorSetText. Look for
the statement:
if vscroll > length( edited ) then
setScrollPos (vscroll, 1 )
end if
and change it to:
if showScroll > length( edited ) then
setScrollPos (vscroll, 1)
end if
Editor will now go to the correct line after ex.err.
4. IDE error
Why is it that when i try to run ide.exw in the win32lib package, I get the
following message:
C:\EUPHORIA\win32lib\ide_edit.ew:248
attempt to redefine flush - defined already in C:\EUPHORIA\include\file.e
procedure flush(integer newColor)
^
Anyone know?
Nick
5. Re: IDE error
Nick Jaohnson wrote:
> attempt to redefine flush - defined already
> in C:\EUPHORIA\include\file.e
Replace all the occurances of 'flush' with 'flushLine'. Robert made flush a
new keyword in Euphoria, so it breaks some code.
-- David Cuny
6. Re: IDE error
Nick writes:
> Why is it that when i try to run ide.exw in the win32lib package,
> I get the following message:
> C:\EUPHORIA\win32lib\ide_edit.ew:248
> attempt to redefine flush - defined already in
> C:\EUPHORIA\include\file.e
> procedure flush(integer newColor)
> ^
In Euphoria version 2.2, a new global routine, flush(),
was added to file.e. This conflicts with flush() in ide_edit.ew.
A few days ago I renamed flush() in the version of Win32Lib.zip
on the Recent User Contributions page. You can download
this new .zip, or get David's bleeding edge version from his site,
or you can rename flush() in ide_edit.ew as something else.
This is an example of a "namespace" problem that
I'll try to improve for the next release. Probably
the local symbol "flush" should take precedence over
the global symbol "flush" in that one file,
instead of triggering a fatal error. A simple warning would
probably be better.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
7. Re: IDE error
- Posted by Everett Williams <rett at GVTC.COM>
Jan 21, 2000
-
Last edited Jan 22, 2000
Robert Craig wrote:
>In Euphoria version 2.2, a new global routine, flush(),
>was added to file.e. This conflicts with flush() in ide_edit.ew.
>A few days ago I renamed flush() in the version of Win32Lib.zip
>on the Recent User Contributions page. You can download
>this new .zip, or get David's bleeding edge version from his site,
>or you can rename flush() in ide_edit.ew as something else.
>This is an example of a "namespace" problem that
>I'll try to improve for the next release. Probably
>the local symbol "flush" should take precedence over
>the global symbol "flush" in that one file,
>instead of triggering a fatal error. A simple warning would
>probably be better.
>
>Regards,
> Rob Craig
> Rapid Deployment Software
> http://www.RapidEuphoria.com
If namespace is handled with prefixing allowed but not required, this
whole situation could be improved without breaking existing code.
The precedence with warning would allow code that now fails to execute, to
possibly execute with unexpected side effects. This seems to be less
desireable than having the code fail as it does now. With the failure, the
programmer can then add prefixing to his use of the include or rename his
function to avoid the problem. With multiple includes possibly containing the
same function or procedure name(actually possibly a desireable state),
precedence would appear to be a chancy and difficult to control method that
would not allow the use of multiple functions or procedures with identical
names when that might be a desireable or necessary thing. All those includes
could be shrouded and not subject to change or examination by the
programmer. As a practical matter, programmers without great expertise
would be greatly challenged to unscramble such a situation in any case,
even with the source available.
Since you are not a supporter of open source for your own code, it would
seem to be inconsistent to create a situation where intimate knowledge of
someone else's source is necessary. If you're going to fix this thing, please
do it in a way that will solve this problem permanently, not create further
quandaries.
Everett L.(Rett) Williams
rett at gvtc.com
8. Re: IDE error
- Posted by Robert Craig <rds at ATTCANADA.NET>
Jan 22, 2000
-
Last edited Jan 23, 2000
Everett Williams writes:
> The precedence with warning would allow code
> that now fails to execute, to possibly execute with
> unexpected side effects. This seems to be less
> desireable than having the code fail as it does now. With
> the failure, the programmer can then add prefixing to his use
> of the include or rename his function to avoid the problem.
I used to think the way you do. That's why it currently
fails when a global symbol conflicts with a local-to-one-file
symbol.
I originally wanted to avoid any possible confusion in
this situation, so I made it a fatal compile-time error. I now
think that confusion in this case is not too likely, since the
confusion will be limited to a single file. Most single files are
developed by a single programmer who is familiar with the
symbols defined in that file.
The current situation causes problems whenever I add
new global routines to the standard include files.
With each release, some existing program, such as
Win32Lib.ew, breaks due to naming conflicts, often with
a local symbol conflicting with one of my new global symbols.
In this case it is very clear which symbol should have precedence.
Unfortunately, people who are new to Euphoria will download
the latest release of Euphoria and then find that some
programs in the Archive fail to run. Most of these people
will not try to understand or fix the problem. They will just
give up in frustration.
The same problem can happen when any author of a
popular include file decides to add some new functionality,
along with some new global symbols. It's obvious that
existing code will not yet contain actual references to the
new global symbol, so any conflicts should be quietly
(with a warning) decided in favor of the local symbol.
So far I'm just talking about global vs. local conflicts.
I haven't decided what to do about global vs. global conflicts,
but insisting on prefixing with the name of the file seems likely.
I wouldn't want the symbol that you are referencing to depend
on the order that you included various files.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
9. Re: IDE error
- Posted by Everett Williams <rett at GVTC.COM>
Jan 22, 2000
-
Last edited Jan 23, 2000
Thankyou for the correction. I am in 100% agreement with what you
have proposed here. I'll even pay for an upgrade if that is in it.
Everett L.(Rett) Williams
rett at gvtc.com
Robert Craig wrote:
>Everett Williams writes:
>> The precedence with warning would allow code
>> that now fails to execute, to possibly execute with
>> unexpected side effects. This seems to be less
>> desireable than having the code fail as it does now. With
>> the failure, the programmer can then add prefixing to his use
>> of the include or rename his function to avoid the problem.
>
>I used to think the way you do. That's why it currently
>fails when a global symbol conflicts with a local-to-one-file
>symbol.
>
>I originally wanted to avoid any possible confusion in
>this situation, so I made it a fatal compile-time error. I now
>think that confusion in this case is not too likely, since the
>confusion will be limited to a single file. Most single files are
>developed by a single programmer who is familiar with the
>symbols defined in that file.
>
>The current situation causes problems whenever I add
>new global routines to the standard include files.
>With each release, some existing program, such as
>Win32Lib.ew, breaks due to naming conflicts, often with
>a local symbol conflicting with one of my new global symbols.
>In this case it is very clear which symbol should have precedence.
>Unfortunately, people who are new to Euphoria will download
>the latest release of Euphoria and then find that some
>programs in the Archive fail to run. Most of these people
>will not try to understand or fix the problem. They will just
>give up in frustration.
>
>The same problem can happen when any author of a
>popular include file decides to add some new functionality,
>along with some new global symbols. It's obvious that
>existing code will not yet contain actual references to the
>new global symbol, so any conflicts should be quietly
>(with a warning) decided in favor of the local symbol.
>
>So far I'm just talking about global vs. local conflicts.
>I haven't decided what to do about global vs. global conflicts,
>but insisting on prefixing with the name of the file seems likely.
>I wouldn't want the symbol that you are referencing to depend
>on the order that you included various files.
>
>Regards,
> Rob Craig
> Rapid Deployment Software
> http://www.RapidEuphoria.com
