1. Testing 2.3
Well, I've not done a lot with 2.3 yet, but I did play my JokoSoko
sokoban game using the new interpreter. It is about 3000 lines of code
plus win32lib and possibly a few other small includes and it seems to
work fine without any problems so far. Compatibility looks good!
Looking forward to further discussion and testing.
-jkr
2. Re: Testing 2.3
Dear Eu users,
I have done the simplest test.
I just have *copied* graphics.e into graf.e
and then ran the program:
-- test.ex
include graphics.e
include graf.e
-- end of program
Nothing happened...
Then:
-- test1.ex
include graphics.e
include graf.e
integer BRIGHT_WHITE
BRIGHT_WHITE=1
-- end of program
I got:
Warning: there is more than one global
symbol with name "BRIGHT_WHITE"
But integer BRIGHT_WHITE
is not *global* in my main file.
Then:
-- test2.ex
integer BRIGHT_WHITE
BRIGHT_WHITE=1
include graphics.e
include graf.e
-- end of program
I got too many Warnings to fit in one screen,
about *all* globals in graphics.e
Then I have copied graphics.e into graff.e:
-- test3.ex
include graphics.e
include graf.e
integer BRIGHT_WHITE
BRIGHT_WHITE=1
include graff.e
-- end of program
I got too many Warnings to fit in one screen,
about *all* globals, but *after* include
graff.e statement, for which my integer
BRIGHT_WHITE is not *global*.
Then:
-- test4.ex
include graphics.e
include graf.e
include graff.e
-- end of program
I got too many Warnings to fit in one screen,
about *all* globals, but *after* include
graff.e statement.
Such a behaviour of namespaceing is strange
enough, I think.
Regards,
Igor Kachan
kinz at peterlink.ru
3. Re: Testing 2.3
Igor Kachan writes:
> I got too many Warnings to fit in one screen,
Well if it's just a matter of too many warnings,
rather than incorrect warnings or errors, then
perhaps I can cut down on them.
(You could also say "without warning")
I assumed that the case of duplicate globals
would be fairly rare, and people would want
to be warned about it, but maybe I should take out that
warning, since you'll get an error if you try
to actually use one of the globals ambiguously
without a namespace qualifier.
Regards,
Rob Craig
Rapid Deployment Software
http://www.RapidEuphoria.com
4. Re: Testing 2.3
Robert Craig writes:
----------
> ïÔ: Robert Craig <rds at RapidEuphoria.com>
> ëÏÍÕ: EUforum <EUforum at topica.com>
> ôÅÍÁ: Re: Testing 2.3
> äÁÔÁ: Friday, December 07, 2001 00:21
>
> Igor Kachan writes:
> > I got too many Warnings to fit in one screen,
>
> Well if it's just a matter of too many warnings,
> rather than incorrect warnings or errors, then
> perhaps I can cut down on them.
> (You could also say "without warning")
>
> I assumed that the case of duplicate globals
> would be fairly rare, and people would want
> to be warned about it, but maybe I should take out that
> warning, since you'll get an error if you try
> to actually use one of the globals ambiguously
> without a namespace qualifier.
Sorry please, my first perplexity was :
> > I just have *copied* graphics.e into graf.e
> > and then ran the program:
> > -- test.ex
> > include graphics.e
> > include graf.e
> > -- end of program
> > Nothing happened ...
Why "Nothing happened ..." if two includes
are fully non-compatible ? - this is my question,
sorry please my English again.
The usual practice, I think, is testing
of includes as the first action in programming
of any new program.
And if interpreter will abort program on any
*first names conflict* then anyone can understand
immediately what actually happens.
Program can not run fine with names conflicts,
so abort it on the first one, then on the second
one and so on. There are tools for resolving any
conflicts in the bud - renameing and qualifiers.
Just some thoughts.
Regards,
Igor Kachan
kinz at peterlink.ru
