Btw, there have been talk of a dependency graph algorithm thing to go with this feature. Apparently Guice of before just instantiated the objects as it went along, proxying here and there if it didn't have the objects, not distinguishing between constructor and setter/field injecting - thus implementing a setter to "resolve" the fact that you were given a proxy didn't necessarily work out. Guice should rather instantiate after some logic; I guess building some kind of tree of the dependencies, instantiating the leaf nodes (those w/o dependencies) first, then going upwards. Any circle in the constructor graph should resolve correctly by having any of the classes in the circle implement a setter or a field injection taking the object(s) it depends on which is a part of the circle.
Is this how it was implemented? Thanks again, Endre. On Mon, Jul 26, 2010 at 00:56, Sam Berlin <[email protected]> wrote: > Yes, it should disable it. I *think* the only time guice created a proxy > was when it detected a circular dependency. If you build the injector with > this option, then where guice would have created a proxy, it instead throws > an exeption saying circular proxies are disabled. > > Sam > > > On Jul 24, 2010, at 5:00 AM, Endre Stølsvik <[email protected]> wrote: > > Oh, I didn't realize that it had been implemented! > > If this method totally disables Guice's use of proxies (so that it never > proxies an object as it did before, you will get the actual objects > injected, always), then that is what I've been wanting since day 0! Thanks! > > Kind regards, > Endre. > > On Fri, Jul 23, 2010 at 14:16, Sam Berlin < <[email protected]> > [email protected]> wrote: > >> I'm not sure what you mean by "please evaluate the proxies". Do you mean >> circular proxies, and you want to disable them? See >> InjectorBuilder.disableCircularProxies. >> >> sam >> >> 2010/7/23 Endre Stølsvik < <[email protected]>[email protected]> >> >> Please evaluate the proxies - an option to turn that idea off is my >>> personal top priority for Guice. The construction of objects should be done >>> such that they aren't needed, or, if cycles exists, any setter/field >>> injection on any of the classes in question should resolve it. >>> >>> Endre. >>> >>> On Sun, Jul 18, 2010 at 01:56, Sam Berlin < <[email protected]> >>> [email protected]> wrote: >>> >>>> Here's another status update (in an effort to get 3.0 out soon) -- >>>> >>>> Open issues: >>>> >>>> * 366 <http://code.google.com/p/google-guice/issues/detail?id=366> - >>>> @Provides doesn't respect @Nullable >>>> Status: Patch attached [from me], needs review (it adds APIs). >>>> >>>> * 436 <http://code.google.com/p/google-guice/issues/detail?id=436> - >>>> Provide access to Guice's own internal TypeConverters >>>> Status: Patch attached [from Stuart], needs review (it adds APIs, >>>> makes some classes public, and changes a bit of the internals). >>>> >>>> * 508 <http://code.google.com/p/google-guice/issues/detail?id=508> - >>>> AOP visibility checks should include the method return type. >>>> Status: Patch attached [from Stuart], just waiting on some tests >>>> before applying. >>>> >>>> * 509 <http://code.google.com/p/google-guice/issues/detail?id=509> - >>>> Cache result of Class.getAnnotations() >>>> Status: Perf improvement request, no patch attached. >>>> >>>> * 510 <http://code.google.com/p/google-guice/issues/detail?id=510> - >>>> Remove InjectorBuilder >>>> Status: Rationale for keeping InjectorBuilder listed in the comments >>>> along with other possible scenarios, awaiting replies. >>>> >>>> * 513 <http://code.google.com/p/google-guice/issues/detail?id=513> - >>>> Why do we need Injector.getAllBindings? >>>> Status: Rationale for why it's useful in the comments, awaiting >>>> replies. >>>> >>>> * 514 <http://code.google.com/p/google-guice/issues/detail?id=514> - >>>> Do we need Injector.getExistingBinding()? >>>> Status: Rationale for why it's useful (but not strictly necessary) in >>>> the comments, awaiting replies. >>>> >>>> * 515 <http://code.google.com/p/google-guice/issues/detail?id=515> - >>>> What are the use cases for @Toolable? >>>> Status: Use cases listed in the comments, awaiting replies. >>>> >>>> * 516 <http://code.google.com/p/google-guice/issues/detail?id=516> - >>>> Investigate duplicate binding filtering >>>> Status: Performance implications & rationale explained in comments, >>>> awaiting replies. >>>> >>>> * 517 <http://code.google.com/p/google-guice/issues/detail?id=517> - >>>> Fix classloader leak >>>> Status: More questions asked in the comments, awaiting answers. >>>> >>>> * 518 <http://code.google.com/p/google-guice/issues/detail?id=518> - >>>> Hide Key.ofType() >>>> Status: Question opened, no replies. Awaiting comments. >>>> >>>> * 519 <http://code.google.com/p/google-guice/issues/detail?id=519> - >>>> toProvider() should use Guice's Provider, not JSR-330 >>>> Status: Lots of discussion in comments, no resolution. >>>> >>>> * Cleanup new guice-persist extension. >>>> Status: Some code review changes applied... is it stable? >>>> >>>> * maven2 POM? >>>> Status: I don't know? >>>> >>>> * Include guice-grapher extension in release >>>> Status: Open question, awaiting comments. >>>> >>>> Closed issues: >>>> * Undeprecate Guice.createInjector. [DONE, >>>> r1180<http://code.google.com/p/google-guice/source/detail?r=1180> >>>> ] >>>> * Fix servlet extension to allow multiple modules/filters/injectors in >>>> one JVM. [DONE, >>>> r1187<http://code.google.com/p/google-guice/source/detail?r=1187> >>>> ] >>>> * Add "@since 3.0" to new APIs. [DONE, >>>> r1183<http://code.google.com/p/google-guice/source/detail?r=1183> >>>> ] >>>> * Remove ScopeChecker [DONE, >>>> r1179<http://code.google.com/p/google-guice/source/detail?r=1179> >>>> ] >>>> * Refactor internal package (move utility class to .internal.util) >>>> [DONE, r1185<http://code.google.com/p/google-guice/source/detail?r=1185> >>>> ] >>>> * Fix overridden @Inject method behavior [DONE, >>>> r1177<http://code.google.com/p/google-guice/source/detail?r=1177> >>>> ] >>>> * JIT providers / autobinders [SHELVED] >>>> >>>> sam >>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "google-guice-dev" group. >>>> To post to this group, send email to >>>> <[email protected]>[email protected]. >>>> To unsubscribe from this group, send email to >>>> <google-guice-dev%[email protected]> >>>> [email protected]. >>>> For more options, visit this group at >>>> <http://groups.google.com/group/google-guice-dev?hl=en> >>>> http://groups.google.com/group/google-guice-dev?hl=en. >>>> >>> >>> -- >>> You received this message because you are subscribed to the Google Groups >>> "google-guice-dev" group. >>> To post to this group, send email to <[email protected]> >>> [email protected]. >>> To unsubscribe from this group, send email to >>> <google-guice-dev%[email protected]> >>> [email protected]. >>> For more options, visit this group at >>> <http://groups.google.com/group/google-guice-dev?hl=en> >>> http://groups.google.com/group/google-guice-dev?hl=en. >>> >> >> -- >> You received this message because you are subscribed to the Google Groups >> "google-guice-dev" group. >> To post to this group, send email to <[email protected]> >> [email protected]. >> To unsubscribe from this group, send email to >> <google-guice-dev%[email protected]> >> [email protected]. >> For more options, visit this group at >> <http://groups.google.com/group/google-guice-dev?hl=en> >> http://groups.google.com/group/google-guice-dev?hl=en. >> > > -- > You received this message because you are subscribed to the Google Groups > "google-guice-dev" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]. > For more options, visit this group at > http://groups.google.com/group/google-guice-dev?hl=en. > > -- > You received this message because you are subscribed to the Google Groups > "google-guice-dev" group. > To post to this group, send email to [email protected]. > To unsubscribe from this group, send email to > [email protected]<google-guice-dev%[email protected]> > . > For more options, visit this group at > http://groups.google.com/group/google-guice-dev?hl=en. > -- You received this message because you are subscribed to the Google Groups "google-guice-dev" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-guice-dev?hl=en.
