World: r3wp

It was built rapidly off of another function that I built in TRETBASE.

If you see a series it doesn't work on let me know.

I see some errors in it already in the code that I should patch real 
quick

Here you go:

replace-all: func [series oldval newval /local sub-ic][
        sub-ic: func [sd][
            forall sd [
                either equal? first sd oldval [
                    poke sd 1 newval
                ][
                    if series? first sd [sub-ic first sd]
                ]
            ]
        ]
        sub-ic series
    ]

Not bad. You use an inner function for the recursion, which should 
allow you to go to greater recursion depth before running out of 
stack space. I'd change the series? to any-block?, and the EITHER 
IF to a CASE, and the FORALL to a WHILE, and add type specs to the 
outer function, and change /local sd to /local sub-ic?, but otherwise 
good stuff.

Sorry, I wrote most of that before you posted your fixed version.

There you BrianH..  You just thought of ways to improve it.  I just 
whipped it up when we were discussing it from another function built 
for a different purpose.

I like your recommendations also.  Maybe I will change the ip-check? 
function I have in TRETBASE which this is based on.

Be sure to change your oldval and newval references to get-words 
for safety.

Not sure I follow you.

Oh so it doesn't evaluate them you mean.

either equal? first sd :oldval [
                    poke sd 1 :newval

Yeah good point

Did i spark your interest for changing the replace function?

If so I suggest you keep the replace function and make this an entirely 
different replace function.  Call it replace-deep if you like.

Then just cut out out the /all from your current replace function 
and leave as is.

The reason is that I believe this will be more tasking on the stats/evals 
then replace is by far.

By the way that recursion method with the function inside the function 
is the only way I found to get round most of those stack errors you 
mentioned.

Have you seen the replace function? We are getting to the point of 
diminishing returns on new features - too much complexity overhead. 
As it is, we are going to have to use APPLY in the R3 version just 
to add another option.

No I haven't really paid much attention to it.  i know there was 
some discussion on the 2.7.6 discussions but i haven't followed R3 
development much at all.

Do source replace - it's a good way to learn some interesting optimization 
techniques.

is it the same in 2.7.6 as it is in 2.7.5?  I'm currently using 2.7.5.

No, it really isn't. The 2.7.6 REPLACE was one of the backports, 
as was EXTRACT.

Extract is in 2.7.5 also which is the one I was using.  Maybe that 
is why we are on different pages on that issue.

I'll check them out later.

I didn't realize that you weren't using 2.7.6. I wrote REPLACE and 
EXTRACT in that release.

oh - lol.

replace-all: func [series oldval newval /local sub-ic][
        sub-ic: func [sd][
            while [not tail? sd ][  
                either equal? first sd :oldval [
                    poke sd 1 :newval
                ][
                    if series? first sd [sub-ic first sd]
                ]
                sd: next sd
            ]
        ]
        sub-ic series
        series
    ]

>> replace-all b 2 1
== [[1] [[[1]]] [1]]
>> stats/evals
== [219 105 47]

Do you really want to do this:
>> replace-all ["1"] 1 2
== ["2"]
Otherwise, change series? to any-block?

Actually, that is exactly what I want

replace-all: func [series oldval newval /local sub-ic][
        sub-ic: func [sd][
            while [not tail? sd ][  
                case [
                    equal? first sd :oldval [poke sd 1 :newval]
                    series? first sd [sub-ic first sd]
                ]
                sd: next sd
            ]
        ]
        sub-ic series
        series
    ]

CASE is faster than 2 comparisons with EITHER or IF.

I recall you saying that before I never really checked it out.

I profile code patterns to see which is faster.

I think your case is messed up

I think I had that if for a reason but not sure why

in the case of mine both conditions will not be true whereas in your 
case arrangment both can can't they?

Not unless you use case/all. Did you have /case as an option to your 
function at some point?

all yeah I always forget that.

Keep thinking of like switch when I see case and I use case often.

It's more like cond from Lisp.

yeah I never used another language other than REBOL so I wouldn't 
know.

Wow. I've never heard that before.

lol - I'm just a hobbiest with REBOL.

I have no programming education.  Had to self teach me this stuff.

That is why when you use a lot of programming terms I'm lost.

I'm self-taught too, but after more than 20 years you pick up some 
stuff :)

I think that replace function is good Brian.  I would push for that 
to be a mezzanine at this point.