Very insightful, Joel!
I found it interesting, however, that you omitted a quote
in the EWD498 document that reads:
[Projects promoting programming in "natural language" are
intrinsically doomed to fail.]
Since Rebol has been designed to have a "natural language"
slant (esp. dialects) - is it d
A hash in an object doesn't seem to act like a block.
Here is a console session that shows the problem.
>> o!: make object! [h: make hash! 10 b: make block! 10]
>> probe o!
make object! [
h: make hash! []
b: []
]
>> obj: make o! []
>> probe obj
make object! [
h: make hash! []
b: []
]
>> append
Great stuff from both of you - I am thinking of changing my
INI file handling code to one of these methods (where I have
a hash of hashes so that a section value is really itself an
associative array).
But Elan, what do you mean when you say that there is a
problem with function code duplicatio
Under NT4, I have a mapped drive u: that contains
some directories that I have no permission to view
the contents of - but can see that they are there and
view properties about the directory from a file manager.
Problem:
>> dir? %/u/employees/ksmith/
** Access Error: Cannot open /u/employees/ksm
A while back Ryan added this Rebol list to the eScribe site.
This is my opinion on the better archive (so don't get mad
anyone - it's just my opinion...)
My suggestion to those wanting answers (and knowing you have
seen it posted but can't remember when) is to first start with
the eScribe archive
Thanks Holger. I'll try the xper but I really need to stick
with a stable release at this point. It looks like I can use
the /binary refinement and copy/part to get the data and that
should work for now.
One possible problem I saw though. I want to check whether there
is data waiting on the po
I need to implement a telnet client in Rebol for a fixed task (that is, I
don't
want a terminal type thing - all client commands are fixed and will be run
with no user interaction).
Anyway, so I open up a telnet connection to the box of interest:
>> c: open tcp://192.168.15.233:23
>>
Cool. Now
Yes, possibly a bug. The wierd thing is that it works
as I expect when the variable assignment is added (sprops: )
to the save.
But I see what you mean about load/save and that load always
puts a block around the data so I should always be ok.
Thanks a lot for your help!
Rodney
-Origina
Thanks for the 'compose lesson. I've never had to use it.
I can get something close to what I want, but I still am having
trouble getting the following to work.
Here is one file %in.dat:
REBOL []
[
ss
[
item1 "hello"
item2 "goodbye"
]
other
[
n "bad"
]
]
Fine. Now a
I'm trying to do something that Rebol should handle so
easily but it's taking me way too much time so I could
use a couple of tips to get going.
I have a small DB that I usually (in C++) construct
by using an INI (Win32) file - where each section is
a new record and it has some number of keys/va
Hi all - slow weekend on the Rebol lists - must be
getting into vacation time :)
I have a /Core script that needs to perform something
different when running on Win32 vs UNIX (the only two
OS's I'm currently concerned about). My script shows
directory paths and the "/" path is valid under UNIX
b
Thanks for the suggestions. I like Tim's idea better
since it's easier on the caller (caller can just 'do
the script like normal) and probably a little
less overhead (try/catch exception handling).
Still seems like maybe a script should be able to just
return a value at any point.
Thanks again
I have a script that 'do(es) another script.
The called script assembles a string and returns
it to the caller by placing the string word as
the last line in the script. Fine.
But the called script has numerous places where
an error can occur and I would like to return
a partially assembled stri
Ah, I finally get it. I was letting whitespace confuse me.
Basically, I think I'll stay away from default arguments
from now on.
Thanks for all the help!
Rodney
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Thursday, June 22, 2000 12:31 PM
To: [EMAIL PRO
Thanks for your patience. I do understand what
Andrew is saying. But maybe you guys can then
explain the following behavior. Type the following
into a script and 'do it:
REBOL[]
print-val: func [val [any-type!]] [
print either value? 'val [val]["No value given"]
]
print-val 1234
print-v
Yes, it doesn't appear that Rebol has an ICMP
protocol handler. This is unfortunate because with
all of it's shortcomings, a ping is one of the most
useful network debugging tools (in my experience -
especially on a LAN).
This also points to a potentially larger problem in
that Rebol doesn't ap
You must perform your test in a script file and
'do it to see the bug. The bug does not occur
in a console session.
Rodney
-Original Message-
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]
Sent: Wednesday, June 21, 2000 9:17 PM
To: [EMAIL PROTECTED]
Subject: [REBOL] Optional Argumen
Hi Larry,
My user.r is minimal - just a set-net[] and a
system/view/ check for /View beta 4.1 (v 0.9.9.3.1).
No other scripts get invoked before this (it happens
on a fresh instance of /View after a reboot).
I am using WinNT4 SP3.
Here is one difference I just noticed, though. If
I type the f
Maybe this has been discussed on the list before
but if I copy/paste the example in the users-guide
about how to use optional arguments in a function:
print-vals: func [arg1 arg2 [any-type!]] [
print either value? 'arg2 [[arg2 arg1]][arg1]
]
print-vals "abc"
abc
print-vals "abc" 123
123 ab
Although I currently don't have the need to do this, I
am wondering how best to perform a /Command function
from /View. For instance, let's say I have a /View app
that is working great but I really need the Win32 version
to access the Registry when an edit box is filled.
No problem - just use /
Excellent Larry - thanks for the explanations (although
the GetSystemTime() vs. time() one still baffles me - I'll
dig into the assembly code when I get some time).
It's good to know that Rebol is probably a (nearly) pure
interpreter especially when dealing with loop calculations.
The fact that
Regarding timestamp - of course it would be possible to create
your own timestamp string from the pieces you want but maybe
a new refinement on either the time or date value or 'now function
could be created. Call it /nonano to retrieve the classic time/date
stamp. Just a thought.
Regarding the
Unfortunately, neither the daytime (RFC 867) or time (RFC 868)
protocols appear to handle subsecond times. The Network Time
Protocol (RFC 958) does, but someone would have to write that
protocol for Rebol. In any case, you really want to be able
to access subsecond times locally.
Rodney
I would think that /Core really needs to return the
known subsecond value when now/time is called.
RT obviously recognized the need because they allow you
to represent times down to the nanosecond (look at the
time! datatype description in users guide) and it seems
restrictive to me to not allow n
This may help in understanding the allocation/insertion times:
A hash preallocates a chunk of memory (into 'bins'). When
you insert an item, a special value (hash value) is calculated
for the item to determine which bin to place it in. This is
all done for fast access when dealing with large n
My 2 cents...
I have personally been a little disappointed that /view is getting all
the RT attention. This is simply because my problems are better
solved with /command and who knows when that will be ready.
I also wish that /core was updated as quickly as /view with new
features and bug fixes.
One problem I can see with the solution of:
to-integer head reverse #{A102}
is that this is assuming reverse-byte storage which is
dependent on the platform (processor) where the script is
being executed. The same script would not run correctly
on all platforms (ie. it would break on 68000
27 matches
Mail list logo
