We've talked about renaming the destructive versions of words to be
more consistent with the non-destructive versions for quite some time,
but until now, little has been done in this direction.
First off, I renamed change-each to map! and made map! return the
sequence so the stack effect is
On Thu, Oct 29, 2009 at 9:35 AM, Doug Coleman doug.cole...@gmail.com wrote:
I removed a couple of unused words: remove-all and substitute-here.
How do you know they're unused?
Chris.
--
http://www.bluishcoder.co.nz
I used grep. I can add substitute-here as substitute! if you'd like.
Remove-all, if it's useful at all, should be renamed to something else
since sequences:remove is something quite different, and the name was
confusing.
In general, if there are no usages of a word in the Factor git
Hi Slava, Hi everyone,
I've got the native thread message passing working really well on Linux,
you can just do a '[ blah ] spawn-vm'.
Unfortunately I'm having problems doing the same thing with io.pipes on
windows. Here's a boiled down test case 'pipe-race.factor', which blows
up with the error
As I threatened to a few days ago, I've gone ahead and streamlined the
locals vocabulary. Instead of having three different binding forms
([let | x [ y ] | ], [let* | x [ y ] | ], and y : x), we now only
have : . [let ] remains as a way to establish a lexical scope in a :
definition or
On Thu, Oct 29, 2009 at 10:32 AM, Doug Coleman doug.cole...@gmail.com wrote:
I used grep. I can add substitute-here as substitute! if you'd like.
Remove-all, if it's useful at all, should be renamed to something else
since sequences:remove is something quite different, and the name was
I went ahead and added a new in-place accumulate word.
accumulate! ( seq identity quot -- final seq )
Some more words have been renamed:
(normalize-path) - absolute-path
canonicalize-path - resolve-symlinks
The difference between normalize-path and absolute-path is that
normalize-path
2009/10/28 Paul Moore p.f.mo...@gmail.com:
2009/10/28 Phil Dawes p...@phildawes.net:
Hi Slava, Hi everyone,
I've got the native thread message passing working really well on Linux,
you can just do a '[ blah ] spawn-vm'.
Unfortunately I'm having problems doing the same thing with io.pipes on
2009/10/28 Paul Moore p.f.mo...@gmail.com:
Yes, the default pipe creation in io.pipes.windows.nt makes the pipe
handles non-inheritable. Rather than hacking the security attributes,
you could use DuplicateHandle (from windows.kernel32) to create an
inheritable copy of the handle. That should
Reading the older docs for [let and [let* I see that the [let* form
evaluates the bindings sequentially rather than in parallel.
What order are the : bindings evaluated?
I'm just curious as I haven't used the : form yet.
On Wed, Oct 28, 2009 at 2:34 PM, Joe Groff arc...@gmail.com wrote:
As I
There is no evaluation at all with :. All values have to be on the
stack and : does the binding only. In some sense, you decide the
evaluation order in the word generate-x-y-z.
On Thu, Oct 29, 2009 at 11:20 AM, Adam hiat...@gmail.com wrote:
Reading the older docs for [let and [let* I see that
On Oct 28, 2009, at 9:20 PM, Adam wrote:
Reading the older docs for [let and [let* I see that the [let* form
evaluates the bindings sequentially rather than in parallel.
What order are the : bindings evaluated?
I'm just curious as I haven't used the : form yet.
As Jon said, : binds the
This makes perfect sense.
What do you think of : or : rather than : to align with the way
accessors/converters have the convention of? The current : just
seems jarring but that's just my opinion.
Thanks again for the clarification.
-Adam
On Wed, Oct 28, 2009 at 8:09 PM, Joe Groff
13 matches
Mail list logo