I think this transformer needs a vote. I for one will cast my vote for this transformer.. Stephen ----- Original Message ----- From: "Jason Birch" <[EMAIL PROTECTED]> To: "Dale Lutz" <[EMAIL PROTECTED]>; <[email protected]> Sent: Wednesday, February 16, 2005 12:40 AM Subject: RE: [fme] Compound Transformers > > > I think that your release notes are somewhat unique in the proprietary industry. I know I find myself checking them at least a couple times a week to see what goodies are getting rolled in. And also for the rare bit of humour. > > "Compound transformer" does make sense, as it's not truly custom (just an aggregation of existing functionality). If you intend to design a mechanism for truly custom transformers (actually, doesn't TCLCaller already do this?) then staying with "compound transformer" makes sense. However, if you do not need to reserve the word "custom" for future use, a rename to be consistent with "custom data source" makes sense. Consistency is a good thing in a product that already requires Jedi training (sorry, can't think of a Battlestar equivalent; how about Gallifreyan wisdom?) to remember all of the functionality. Either stay consistent or my vote is for WhizBangTransformerComplexityAbstractor. > > I know that some people use your betas in production. In my experience they are typically more stable than most released commercial products. Perhaps this encourages people to test them more, letting you find the edge cases before final release. However, I would personally be happier playing with bleeding edge stuff; whether it makes it into the distribution or not. Have you considered making your alpha/developer release builds public, or are they just too chaotic before you move the code back into the main branch? > > Jason > > ________________________________ > > From: Dale Lutz [mailto:[EMAIL PROTECTED] > Sent: Tue 2005-02-15 10:48 PM > To: [email protected] > Subject: RE: [fme] Compound Transformers > > > > Wow, many of you noticed this functionality that was beginning to appear in > the current betas. > > Here's the deal. We had to make very very substantial changes to the core > workbench codebase in preparation for unveiling "compound transformers" > (which may still yet be named Custom Transformers -- I'd be interested in > opinions on what the FME audience thinks a good name would be). > > Anyway, we wanted to get the core Workbench out into people's hands as soon > as possible, even before we were done our initial internal testing of the new > Compound Transformer functionality. We didn't want your first experiences > with this to be bad. So the easiest way was to disable the menu items > relating to compound transformers. We didn't think anyone would notice :-) > > So, to get the full compound transformer stuff you have to wait for a newer > build to come out. This should happen within the next week or so. We'll > post here when its available. > > > BTW Jason's definition of what compound transformers are is correct. We are > letting you basically select a chain of transformers in your existing > workspace, and give them a name. They are then replaced by a single > "compound transformer" which does all the work that the sequence used to do. > You can then duplicate that transformer within the workspace, and perhaps > re-parameterize it, and then you've got all those "elemental" transformers > working for you again. You can also edit the definition of the compound > transformer in another tabbed window within workbench. You CANNOT include > any source or destination feature types within a compound transformer. You > also can't use a FeatureTYpeFilter in one either (but I can't figure out why > you'd want to anyway). There may be some other restrictions but so far we've > not had much time to explore this brave new world of nested transformers. > (Oh ya, that reminds me, you can't recursively define transformers either, > but you can use one compound transformer within another). You can also > "publish" parameters from the elemental transformers to become parameters of > the compound one. You can copy compound transformers across workspaces -- > the definition of it is then embedded in the other workspace. Right now you > can't save the compound transformer into a 'library', that will come sometime > later, probably not until after we've let this stuff settle down. In the > meantime, you'll be able ot make a workspace that acts as your library if you > are so inclined, and then copy and paste out of it. > > > We think this will greatly help simplify many workspaces and make them easier > to deal with as the transformations become decomposed into smaller pieces. > Combined with the new bookmarks,we're hoping this greatly improves the > usability of workbench in the presence of larger transformations. > > We'll let you know when all this appears in a beta. > > > Dale > > ---------------------------------------------------------------------- > Dale Lutz Safe Software Inc. > VP Development Surrey, BC, CANADA phone: (604) 501-9985 > http://www.safe.com fax: (604) 501-9965 > ---------------------------------------------------------------------- > > > > > > > > Get the maximum benefit from your FME, FME Objects, or SpatialDirect via our Professional Services team. Visit www.safe.com/services for details. > Yahoo! Groups Links > > > > > > > > Get the maximum benefit from your FME, FME Objects, or SpatialDirect via our Professional Services team. Visit www.safe.com/services for details. Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/fme/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
