On Sat, Jan 18, 2020 at 3:11 PM Michael Osipov <[email protected]> wrote:

> Am 2020-01-18 um 19:05 schrieb Gary Gregory:
> > I think the idea behind the HttpRequests enum is that it defines a
> specific
> > set of supported values.
> >
> > If we want to allow users to extend this set of values, then the enum
> could
> > implement an interface and we would code to the interface as opposed to
> the
> > enum internally.
> >
> > I suppose this all depends on how we want to represent 'we support RFC N
> > and that RFC defines HTTP verbs A, B, and C'...
>
> I would not make it too complex and academic. A public final class makes
> pretty clear that we are in control of this class and provide it as
> courtesy. Anything else more sophisticated can implemented by the user
> or donated.
>
> It really looks like an enum abuse for fluent builders.
>
> Gary, are you OK making them "public final classes"?
>

Sure, go ahead.

Gary


> M
>
>
> > On Sat, Jan 18, 2020 at 11:48 AM Oleg Kalnichevski <[email protected]>
> wrote:
> >
> >> On Sat, 2020-01-18 at 16:17 +0100, Michael Osipov wrote:
> >>> Am 2020-01-18 um 11:02 schrieb Oleg Kalnichevski:
> >>>> On Wed, 2020-01-15 at 22:39 +0100, Michael Osipov wrote:
> >>>>> Am 2020-01-15 um 18:24 schrieb Oleg Kalnichevski:
> >>>>>> On Wed, 2020-01-15 at 17:36 +0100, Michael Osipov wrote:
> >>>>>>>> On Mon, 2020-01-13 at 20:20 +0100, Michael Osipov wrote:
> >>>>>>>>> Am 2020-01-13 um 15:00 schrieb Oleg Kalnichevski:
> >>>>>>>>>> Michael, Gary
> >>>>>>>>>>
> >>>>>>>>>> How long do you intend to block HttpClient 5.0-beta7
> >>>>>>>>>> release?
> >>>>>>>>>>
> >>>>>>>>>> Are you done with all this renaming and enum stuff?
> >>>>>>>>>
> >>>>>>>>> One remaining PR and I am done. Will merge tonight.
> >>>>>>>>>
> >>>>>>>>
> >>>>>>>> Michael
> >>>>>>>>
> >>>>>>>> Is there anything left blocking the release?
> >>>>>>>
> >>>>>>> No! I have merged immediately after my email.
> >>>>>>>
> >>>>>>> Go ahead.
> >>>>>>>
> >>>>>>
> >>>>>> OK. And what about HttpRequests, SimpleHttpRequests and
> >>>>>> ClassicHttpRequests?
> >>>>>
> >>>>> I will have a look at them tomorrow.
> >>>>>
> >>>>
> >>>> Hi Michael
> >>>>
> >>>> Any update?
> >>>
> >>> Wen't through the enums. First of all, I think HttpRequests should
> >>> be
> >>> BasicHttpRequests to be consistent with the rest.
> >>>
> >>> As it looks like to me these are basic builders/factories. Enum has
> >>> been
> >>> used (to forbid inheritance?). In my opinion this is not what enum
> >>> is
> >>> for. I would expect them to be like:
> >>>
> >>> public final class BasicHttpRequests {
> >>>
> >>>     public BasicHttpRequests delete(String) ...
> >>>     public BasicHttpRequests delete(URI) ...
> >>>     public BasicHttpRequests delete(HttpHost, String) ...
> >>>     // and so on
> >>>
> >>> }
> >>>
> >>> A bit similar to the HTTP method operations in
> >>> org.springframework.web.client.RestTemplate, but w/o the execution.
> >>>
> >>> If that sounds reasonable, I will provide a PR today for all three of
> >>> them.
> >>>
> >>
> >> Michael
> >>
> >> I personally do not have an opinion strong enough and will just stay
> >> away.
> >>
> >> If my memory still serves me the first of such classes was contributed
> >> by Gary and I modeled similar classes based on the first one.
> >>
> >> Oleg
> >>
> >>
> >>
> >> ---------------------------------------------------------------------
> >> To unsubscribe, e-mail: [email protected]
> >> For additional commands, e-mail: [email protected]
> >>
> >>
> >
>
>

Reply via email to