Hi,

> Why duplicate the component hierarchy in Java if it's already defined in
the markup?

I'm not completely sure that component queuing could lead to a more
maintainable code. Reflecting component hierarchy in Java code might seem
an additional burden, but I have the feeling that "flatting" this hierarchy
with component queuing makes debugging and trouble shooting harder as it's
more difficult to understand what contains what.
my 2 cents.

> having missed the introduction of component queuing, I was surprise now
how many places were touched by it - take a look at the size of the commit,
that's insane for a mere convenience feature.

That's right, component queuing should have been provided in a more modular
fashion, without touching so many parts of the framework.

On Thu, Jun 3, 2021 at 4:31 PM Sven Meier <[email protected]> wrote:

> Hi,
>
> having missed the introduction of component queuing, I was surprise now
> how many places were touched by it - take a look at the size of the
> commit, that's insane for a mere convenience feature.
>
>  > Why duplicate the component hierarchy in Java if it's already defined
> in the markup?
>
> Maybe if Wicket had offered component queuing from the beginning, we'd
> had a better solution available.
>
> Regards
> Sven
>
>
> On 03.06.21 15:52, Martin Grigorov wrote:
> > On Thu, Jun 3, 2021 at 3:57 PM Andrew Geery <[email protected]>
> wrote:
> >
> >> I personally find queueing really useful.
> >>
> >> I'm curious why people aren't using it.
> >>
> >> Why duplicate the component hierarchy in Java if it's already defined in
> >> the markup?
> >>
> > well, it is all in the code
> > see Sven's commit at
> >
> https://github.com/apache/wicket/commit/adbf35e6176ea69a36e5c7eaf1f514fac99cb91e
> > it adds complexity and fails at random corner cases
> > and no one wants to maintain this extra complexity and add more on top
> of it
> >
> >
> >> Thanks
> >> Andrew
> >>
> >> On Thu, Jun 3, 2021 at 8:09 AM Sven Meier <[email protected]> wrote:
> >>
> >>> Hi Ernesto,
> >>>
> >>> if you want to do some tests:
> >>>
> >>> https://github.com/apache/wicket/tree/remove-queuing
> >>>
> >>> Have fun
> >>> Sven
> >>>
> >>> On 02.06.21 12:04, Ernesto Reinaldo Barreiro wrote:
> >>>> Hi,
> >>>>
> >>>>
> >>>>> I already removed queuing on an experimental branch, and it was quite
> >>>>> simple.
> >>>>>
> >>>>>
> >>>> Nice. We have the master  branch of our application running on 9.3.0.
> >>> and a
> >>>> large set of selenium tests: I will try it against this branch.
> >>>>
> >>>> Regards
> >>>>> Sven
> >>>>>
> >>>>>
> >>>>> On 02.06.21 08:36, Ernesto Reinaldo Barreiro wrote:
> >>>>>> Hi,
> >>>>>>
> >>>>>> On Wed, Jun 2, 2021 at 9:14 AM Martin Grigorov <
> [email protected]
> >>>>> wrote:
> >>>>>>> On Tue, Jun 1, 2021 at 10:51 PM Ernesto Reinaldo Barreiro <
> >>>>>>> [email protected]> wrote:
> >>>>>>>
> >>>>>>>> Hi Martin,
> >>>>>>>>
> >>>>>>>> How much work do you think this will imply?
> >>>>>>>>
> >>>>>>> Removing the component queueing ?
> >>>>>>> I am not sure. Removing should be easier than implementing it.
> >>>>>>> Why do you ask ?
> >>>>>>>
> >>>>>> It is removing +  fixing back all things that were changed because
> of
> >>>>> this
> >>>>>> feature.I have the impression this implied a lot of minor changes.
> >> Now
> >>> it
> >>>>>> is reverting all that and making sure everything works again. Thus I
> >>> was
> >>>>>> just wondering if it will be worth the effort to do that just to
> >> remove
> >>>>>> something that is almost working in all cases.
> >>>>>>
>


-- 
Andrea Del Bene.
Apache Wicket committer.

Reply via email to