Slides/notes, RubyConf2003

[Hal Fulton]

Hello, all.

It recently came to my attention that I never did put up my slides
from the last Ruby conference. <sarcasm>This was only five months ago,
you know.</sarcasm>

At any rate, I just polished them a tiny bit, so here they are:

http://rubyhacker.com/ff/slide01.html

The title is "Toward a Refactoring Framework for FreeRIDE."

Of course, some bright people have obsoleted some of this work since
that time. My hope is that parts of my code or my ideas will find their
way into their project.

If you read these slides, don't miss the commentary at the bottom, and
don't miss the (HTML version of the) demo at the end.

As I'm not parser-savvy, the bulk of what I did turned out to be:

1. Develop a reasonable skeleton (including a few GUI and scripting
issues)
2. Work on a Ruby-like editor API to augment the "literally
translated" API

Also note:

1. This is NOT just about refactoring. It's also about generalized
editing scripts, macros, and so on.
2. Even the editor API, the most complete part, is not really released
yet, and is very much a work in progress.
3. The Editor API is not dependent on FreeRIDE, so it could be used by
(for example) Scite fans also.


Cheers,
Hal Fulton

 

[Lothar Scholz]

Hello Hal,

Sunday, April 18, 2004, 4:47:59 AM, you wrote:

HF> Hello, all.

HF> It recently came to my attention that I never did put up my slides
HF> from the last Ruby conference. <sarcasm>This was only five months ago,
HF> you know.</sarcasm>

HF> At any rate, I just polished them a tiny bit, so here they are:

HF> http://rubyhacker.com/ff/slide01.html

HF> The title is "Toward a Refactoring Framework for FreeRIDE."

Hmmm, but i don't find anything about refactoring in this slides.
It's all about some really simple editor things (i don't want say
anything about the so called GUI API).

Refactoring is about code transformation and most importantly about
semantik neutral code transformation. This is the important task and
we should talk about this first. Because if they turn out to be
impossible or very very unsatisfying in ruby (thats my point of view) we
don't need a framework.

Have do done any work on this topic ?

 

[Hal Fulton]

HF> The title is "Toward a Refactoring Framework for FreeRIDE."

Hmmm, but i don't find anything about refactoring in this slides.
It's all about some really simple editor things (i don't want say
anything about the so called GUI API).

Don't you find the manipulation of code as text a necessary
condition for automatic refactoring?

Hal

 

[Mauricio Fernández]

Don't you find the manipulation of code as text a necessary
condition for automatic refactoring?

Isn't refactoring normally performed at the syntax-tree level (hence
the need for a parser)?

--
Running Debian GNU/Linux Sid (unstable)
batsman dot geo at yahoo dot com

On the Internet, no one knows you're using Windows NT
-- Submitted by Ramiro Estrugo, (e-mail address removed)

 

[gabriele renzi]

il Sun, 18 Apr 2004 14:16:43 +0900, Lothar Scholz

Because if they turn out to be
impossible or very very unsatisfying in ruby (thats my point of view) we
don't need a framework.

Have do done any work on this topic ?

well, it seem that the best one is the ruby refcatoring browser for
emacs:
http://www.kmc.gr.jp/proj/rrb/index-en.html

looking at the freeride ml it seem work is going on to port it to
freeride (I believe hal was talking about this in the first msg)
This code allows:
* Rename local/instance/class variable
* Rename method
* Rename constant(including class name and module name)
* Pull up/Push down method
* Extract method
* Extract superclass

I'm not much into automagic refactoring but this seem the core things
that need to be allowed , am I wreong ?

 

[Charles Comstock]

Isn't refactoring normally performed at the syntax-tree level (hence
the need for a parser)?

Yea but syntax trees breaks when you can add stuff to the tree at
runtime. Well it doesn't ACTUALLY break since we can run it, but from a
editor level it breaks because you have to actually run the program in
order to walk the syntax tree and you can't exactly tree search all
possible syntax trees in order to do automatic refactoring.
Charlie