World: r3wp

Instrumenting R2 would be a good idea, but is not likely happening 
in the 2.7.7 release. I'll bring it up with Carl.

Later, I mean. He can read it now here, but we need to track these 
things for the long run :)

Gab: your LOAD bug has been fixed for this release.

Robert: Sounds like a job for REBOL Consulting?

And that means?

www.rebol.com/consulting.html

There is 1 day left for V2.7 fixes/changes.  Please contact BrianH 
if you have any critical items for this build.  You must have test 
code available to validate any changes.

I will be integrating changes starting this evening (Chicago time, 
starting 4 hours from now). If you want your fixes in, get them to 
me before I go to sleep (late) and I'll see what I can do. Code that 
patches memory should include a version of the function that is supposed 
to be the result of the patch; we're patching the source, not hotpatching 
the runtime.

Gab. has mentioned an updated imap protocol that might be donated 
... dunno if that is still on or not.

I guess he'll be up in 4 hours or more ...

If it can get to me this evening (which will be for quite a while), 
it can go in. Otherwise, 2.7.8 next month :)

what time is it for you now?

7:41pm. I'll be up til early morning tomorrow.

Hope you are leaving the installer for last ...

Yes. And it looks like it might not make the cut for this version, 
instead doing just a simple fix to make the existing installer work 
again. To do a full installer revamp would require testing on every 
Windows platform, and I don't have time to set up all of the VMs 
required.

It is likely to make in into the next version a month from now though.

R2 2.7.6 SDK sources posted to R3 chat in the appropriate headers. 
Time to integrate the backwards-compatible R2/Forward changes.

If you need temporary access to windows 2008, windows 2003 servers, 
you can start them up in Amazon EC2

Thanks, that will help for next month :)

Brian is it possible to backport that "func" with all words as local 
? 
I don't recall exactly ... maybe "funct" in R3

It's in R2/Forward already. With any luck I'll have enough time to 
backport the /with option too :)

R2/forward ? Is it a patch ?

It's a full rewrite of a lot of the R2 mezzanines, making them less 
buggy and more R3 compatible.

but it's a separate .r file or a new rebol exe ?

It's a separate .r file, but its functions can be swapped into the 
SDK source and a new .exe built. And that is my task for the evening: 
to swap in the more R2-compatible functions in R2/Forward into the 
mezzanine source.

Great ! Brian, you are our avatar

in 3d !

10 feet tall and blue?

Aren't you ?

Sure, why not? :)

Thanks for all your work.

Yes, great work!

Proposed 2.7.7 mezz changes have been submitted to DevBase - look 
in R3 chat for details. Haven't gotten to the protocols or VID yet.

Brian, for R2 can we make sure that a base component isn't doing 
any unnecessary reads or writes - such as registry values etc...

I get killed trying to use stuff like REBOL in the enterprise because 
it does stuff that I didn't advertise.

Core doesn't do anything like that, just View. I'm going to do a 
full review of the installer for the next version though.

I hope core or base doesn't do it.  We need those as lean as possible.

Graham, I'm looking over your http scheme changes now. They won't 
maker the cut this release but should make the next release once 
it's been through the REBOL optimizer. The installer revamp didn't 
make this release either, though in theory the current installer 
might be working again. The 2.7.7 release has been mostly the licensing 
changes, a few native fixes, and the integration of a lot of R2/Forward 
backports and fixes.

What is the licensing changes looking like?

Brian, I thought you were the Rebol optimizer!

You get SSL, ODBC and encryption for free in 2.7.7.

Got ya.  Cool.

As for the REBOL optimizer, for R2 networking code there are better 
optimizers than me: Doc, Gabriele, Maarten, ... :)

By the way, the backports include FUNCT, closures (faked), typesets 
(faked), the REFLECT and *-OF functions, APPLY, MAP-EACH, ...

Carl might even add the missing operators != and !== :)

If we're really lucky we'll get the STRICT-NOT-EQUAL? integer! integer! 
fix (crossing fingers...).

Shame .. the changes don't break anything .. just add new functionality 
needed for REST etc

It's a time thing. The next release is next month so it's not much 
of a shame :)

'func  Defines a function with all set-words as locals.  but if you 
use a parse rule which sets a word, it appears in global name space.

'funct