This post originated from an RSS feed registered with Agile Buzz
by James Robertson.
Original Post: A Tribute
Feed Title: Travis Griggs - Blog
Feed URL: http://www.cincomsmalltalk.com/rssBlog/travis-rss.xml
Feed Description: This TAG Line is Extra
As I sit here late at night waiting for the build server to complete a cycle... I thought I'd catch up on a post I've meant to make for awhile.
I was reminiscing what life was like before the RefactoringBrowser. I was privileged to use the RefactoringBrowser from its early days when it was just a filein/parcel in/envy app to the 5.x series of VisualWorks.
It was a better browser. By far. Everything about it was just great. Of course, being compared to the "full browser" for VisualWorks with namespaces, this wasn't saying much. But I'd used the rest, and it was just better than those too. It wasn't a revolution or paradigm shift or anything. I think what made it great was the numerous evolutions it went through from a decently wide group of people, the author's endless tweaking, and the fact that it was designed, improved, and maintained by people who were programming real world solutions with Smalltalk. The tool, as a tool, eventually found its way into a must have for VisualWorks users, and Cincom ended up integrating it. Today, the tool, as a browser/environment continues to slowly evolve under Vassili Bykov's steady hand.
All that aside, that's not what made the RefactoringBrowser great. It was the stuff under the hood. The automatic refactorings. I sometimes have to remind myself that to rename a method, I first browsed "all senders" so that I'd be able to affect the change elsewhere. AND, I bragged to programmers of other languages how vastly more efficient Smalltalk IDEs were. Chuckle. Extract method. Wow. I remember the first time I used Extract to Component. Rename class. Today, I mostly take these for granted. And these are the things that other language IDEs are attempting to reproduce. Not the browser really.
The sad thing to me, is that after the Refactory guys' contract expired with Cincom, I am aware of no significant work that has taken place with the refactoring engine. It's like it died. Cincom does not really have an engineer dedicated to maintaining/improving the refactoring engine. I'm sure it falls to someone, but what with workloads what they are. Maybe we have enough. Maybe there's just no other refactorings worthy of exploring and implementing. I'd like to think otherwise though. Three point releases later, and it's still not possible to move objects from namespace to namespace as a refactoring. The targets for extract to component have never been improved.
Regardless, I remain acutely impressed that the refactoring engine was the single greatest Smalltalk innovation in the last 10 years. It has had a far profounder affect on how I program and solve problems (both in Smalltalk and out) than any other library/VM enhancement. And though I may be frustrated that Cincom doesn't put much into the engine itself, I AM grateful that they did go as far as taking the steps to integrate it as far as they did. Thank you to John and Don for the RB. Thank you to Cincom for adopting it.
Previous Topic
