Re: Constructor annotation

2013-10-23 Thread Eva Krejcirova
To summarize: the final agreement was on: @Retention(RUNTIME) @Target(PARAMETER) public @interface NamedArg { public String value(); public String defaultValue() default ""; } in javafx.beans package, right? If nobody objects, I will go with this. Regards, Eva On 16.10.2013 21:07, Ev

Re: Constructor annotation

2013-10-16 Thread Eva Krejcirova
We already have RT-32860 (Add constructor annotations) for this. I've copied this discussion to the JIRA issue. Eva On 16.10.2013 19:27, Stephen F Northover wrote: Eva, Perhaps @NamedArg is shorter and makes the code more readable? If you don't have a JIRA already, please create one and pas

Re: Constructor annotation

2013-10-16 Thread Claus Luethje
+1 > Am 16.10.2013 um 19:28 schrieb "Richard Bair" : > > Looks good to me. > >> On Oct 16, 2013, at 10:02 AM, Stephen F Northover >> wrote: >> >> It seems we are settling on @NamedArgument ... anybody disagree strongly? >> >> Steve >> >>> On 2013-10-16 11:45 AM, Richard Bair wrote: >>> Ya

Re: Constructor annotation

2013-10-16 Thread Richard Bair
NamedArg, like a pirate :-) > On Oct 16, 2013, at 10:27 AM, Stephen F Northover > wrote: > > Eva, > > Perhaps @NamedArg is shorter and makes the code more readable? > > If you don't have a JIRA already, please create one and paste in this > discussion. Interested parties can add themselves

Re: Constructor annotation

2013-10-16 Thread Stephen F Northover
Eva, Perhaps @NamedArg is shorter and makes the code more readable? If you don't have a JIRA already, please create one and paste in this discussion. Interested parties can add themselves to the watchers list. Steve On 2013-10-16 1:25 PM, Richard Bair wrote: Looks good to me. On Oct 16,

Re: Constructor annotation

2013-10-16 Thread Richard Bair
Looks good to me. > On Oct 16, 2013, at 10:02 AM, Stephen F Northover > wrote: > > It seems we are settling on @NamedArgument ... anybody disagree strongly? > > Steve > >> On 2013-10-16 11:45 AM, Richard Bair wrote: >> Ya that works too. >> >>> On Oct 16, 2013, at 8:41 AM, Eva Krejcirova >

Re: Constructor annotation

2013-10-16 Thread Stephen F Northover
It seems we are settling on @NamedArgument ... anybody disagree strongly? Steve On 2013-10-16 11:45 AM, Richard Bair wrote: Ya that works too. On Oct 16, 2013, at 8:41 AM, Eva Krejcirova wrote: Good point! In FX sources, we already use the @Default annotation which was used by annotation p

Re: Constructor annotation

2013-10-16 Thread Richard Bair
Ya that works too. > On Oct 16, 2013, at 8:41 AM, Eva Krejcirova wrote: > > Good point! > In FX sources, we already use the @Default annotation which was used by > annotation processor when generating the builders. Because of this, it has > source retention policy, so it cannot be used by FXML

Re: Constructor annotation

2013-10-16 Thread Eva Krejcirova
Good point! In FX sources, we already use the @Default annotation which was used by annotation processor when generating the builders. Because of this, it has source retention policy, so it cannot be used by FXMLLoader. I was thinking about promoting this to runtime annotation but maybe your s

Re: Constructor annotation

2013-10-16 Thread Tom Schindl
One thing that just came to my mind is that maybe also need a way to define the default value to be used, with a builder I could e.g. define that the default for fields are different from their real native default. class MyBuilder { private boolean a = true; private int x = -1; private Inset

Re: Constructor annotation

2013-10-16 Thread Sven Reimers
+1 for base module +1 for option 2 Only thing that troubles me is the readability if the annotation is long.. you will end up with a lot of boilerplate annotation characters vs. content in the constructor argument declaration... -Sven On Wed, Oct 16, 2013 at 5:12 PM, Tom Schindl wrote: > To m

Re: Constructor annotation

2013-10-16 Thread Tom Schindl
To me the JavaBean solution with one annotation looks error prone, does anybody know why they did not use an annotation per field? Tom On 16.10.13 16:58, Stephen F Northover wrote: > +1 for base. Should we not follow closely what Java Beans is doing for > consistency? I realize that we can't ha

Re: Constructor annotation

2013-10-16 Thread Kevin Rushforth
Not to mention Tom's point that it can't be in the fxml module without created unwanted (and circular) module dependencies. Seems like it needs to be in the "base" module then, right? -- Kevin Richard Bair wrote: +1 this is my preference. It is useful for things other than FXML, and should

Re: Constructor annotation

2013-10-16 Thread Stephen F Northover
+1 for base. Should we not follow closely what Java Beans is doing for consistency? I realize that we can't have the reference. Steve On 2013-10-16 10:53 AM, Kevin Rushforth wrote: Not to mention Tom's point that it can't be in the fxml module without created unwanted (and circular) module d

Re: Constructor annotation

2013-10-16 Thread Richard Bair
+1 this is my preference. It is useful for things other than FXML, and should be considered part of our javafx.beans API. > On Oct 16, 2013, at 4:20 AM, Tom Schindl wrote: > >> On 16.10.13 11:22, Eva Krejcirova wrote: >> Hi All, >> >> when we retired builders, we caused a problem for FXML whic

Re: Constructor annotation

2013-10-16 Thread Eva Krejcirova
On 16.10.2013 11:26, Tom Eugelink wrote: Would the duality of using the parameter info when available or else the annotation (2nd option) be an viable approach? This is probably doable but I am not sure whether we want to turn on the parameter info for the whole JavaFX - last time I checked i

Re: Constructor annotation

2013-10-16 Thread Tom Schindl
On 16.10.13 11:22, Eva Krejcirova wrote: > Hi All, > > when we retired builders, we caused a problem for FXML which doesn't > have a way to create classes without default constructors. Back then we > decided to use an annotation for this but never actually got to > implement it and we need to fix

Re: Constructor annotation

2013-10-16 Thread Eva Krejcirova
ely. Eva *From:* Claus Luethje [mailto:claus.luet...@osys.ch] *To:* Eva Krejcirova [mailto:eva.krejcir...@oracle.com], openjfx-dev@openjdk.java.net [mailto:openjfx-dev@openjdk.java.net] *Sent:* Wed, 16 Oct 2013 11:26:12 + *Subject:* RE: Constructor annotation

RE: Constructor annotation

2013-10-16 Thread Jack Moxley
:12 + Subject: RE: Constructor annotation Hi I'd prefer the second option, because the correlation of the order of arguments in the annotation and in the constructors parameters is irritating and error prone. The way it is structured in option two is seen elsewhere also. So, nothing n

RE: Constructor annotation

2013-10-16 Thread Claus Luethje
name to describe what's going on. My 2 cents... Regards Claus -Original Message- From: openjfx-dev-boun...@openjdk.java.net [mailto:openjfx-dev-boun...@openjdk.java.net] On Behalf Of Eva Krejcirova Sent: Mittwoch, 16. Oktober 2013 11:22 To: openjfx-dev@openjdk.java.net Subject: C

Re: Constructor annotation

2013-10-16 Thread Tom Eugelink
Would the duality of using the parameter info when available or else the annotation (2nd option) be an viable approach? Tom On 2013-10-16 11:22, Eva Krejcirova wrote: Hi All, when we retired builders, we caused a problem for FXML which doesn't have a way to create classes without default c

Constructor annotation

2013-10-16 Thread Eva Krejcirova
Hi All, when we retired builders, we caused a problem for FXML which doesn't have a way to create classes without default constructors. Back then we decided to use an annotation for this but never actually got to implement it and we need to fix this for FX8. I am in the process of adding this