sweet, can't wait > -----Original Message----- > From: Mike Aizatsky [mailto:[EMAIL PROTECTED]] > Sent: 28 February 2002 10:09 > To: [EMAIL PROTECTED] > Subject: RE: [Eap-features] TestFirst XP and create class on unknown > type > > > Tim, > > We are planning to "teach" our intention action to recognize coding > conventions. We also have the action to create class not from "new" > operator but from first mentioning in our request base. Just > stay tuned > and give as a bit more time. > > Best regards, > Mike Aizatsky. > ------------------------------ > IntelliJ Software, "Develop with pleasure!" http://www.intellij.com > > > > -----Original Message----- > > From: [EMAIL PROTECTED] [mailto:eap-features- > > [EMAIL PROTECTED]] On Behalf Of Tim Mackinnon > > Sent: Thursday, February 28, 2002 3:52 AM > > To: [EMAIL PROTECTED] > > Subject: [Eap-features] TestFirst XP and create class on > unknown type > > > > I work on an XP team and Martin Fowler often sings your > praises so we > > thought we would give it a go. In our early tests its very good - we > are > > very impressed. > > > > However I want to give you some XP feedback - in a test first > environment > > you often refer to types that are not there yet (you refer > to them in > your > > test with the notion of creating them later) and I think you can do > better > > (although you are not bad).. > > > > I often create a test method and then refer to a type that > isn't there > > e.g. > > > > ServiceCalculator > > > > at this point you show red and only offer to create a field - this > isn't > > very helpful in an XP world and to be honest not many > people create a > > field > > with a capitalised name - you really should offer a second option - > create > > Class and third Create Interface - this really would support XP and > java. > > > > I say this becuase the next bit you type is a variable name (more > often > > than > > not named a<typename> or f<typename> or my<typname> > according to your > > house > > style (p.s. you handle this nicely with a nice LiveTemplate called > prefix > > with a $variable$ and a definitoin of > captalize(suggestTypeName())) ) > then > > you type "new" and want to complete again (ahhh the tedium of Java). > > > > If you have already defined the type - and type "new" then complete > works > > well - it expands to the type you mentioned at the beginning. If the > type > > is > > undefined (the normal case in test first) then you can't > complete and > have > > to type the name again - at this point you do offer to > create a class > - > > but > > in test first its just too late and tedium has already set in. > Creating > > the > > type at the beginning would be just so handy and then your normal > rules > > appy > > (unless you created an interface - and then after the new you want > either > > a > > list of conforming classes or an offer to create a new type that > conforms > > to > > the interface). > > > > I know teh above is a bit tricky to read - but you do have > the ability > to > > be > > a leader in XP developement tools. > > > > Tim > > --- > > Outgoing mail is certified Virus Free. > > Checked by AVG anti-virus system (http://www.grisoft.com). > > Version: 6.0.325 / Virus Database: 182 - Release Date: 19/02/2002 > > > > > > _______________________________________________ > > Eap-features mailing list > > [EMAIL PROTECTED] > > http://www.intellij.com/mailman/listinfo/eap-features > > > _______________________________________________ > Eap-features mailing list > [EMAIL PROTECTED] > http://www.intellij.com/mailman/listinfo/eap-features >
_______________________________________________ Eap-features mailing list [EMAIL PROTECTED] http://www.intellij.com/mailman/listinfo/eap-features
