May 2011 - Pharo-users - lists.pharo.org

Re: [Pharo-users] Object>>copyTwoLevel

by Mariano Martinez Peck

On Thu, May 5, 2011 at 4:30 AM, Igor Stasenko <siguctua(a)gmail.com> wrote: > On 5 May 2011 01:59, Johan Fabry <jfabry(a)dcc.uchile.cl> wrote: > > > > On 04 May 2011, at 17:30, StÃ©phane Ducasse wrote: > > > >> > >> mariano got a really nice proxy model based on the stratified proxy of > igor (which is in 1.3). > > > > > > Question: could this be used for PHANtom? We are now using Methodwrappers > but there are some issues (fixed in PHANtom). Mariano, Igor, can you explain > the advantages of your model wrt methodwrappers? > > > > Sorry i'm not familiar with methodwrappers. > > The basic model is separation of proxy and handler. Proxy acts only as > a message trap, > while handler is actually responsible for handling messages trapped by > proxy. > > Mariano builds on top of that, because he wants to replace classes > and/or methods with proxies, > and this requires a special handling (class proxy should obey > VM-specific Behavior format, > and method proxy should implement #run:in:as: when installed in method > dictionary ). > > So, i guess that proxies which representing compiled methods could act > similar to method wrappers. > But of course you still have to implement all intall/deinstall > management et al.. > > The only advantage of new proxies comparing to usual proxies based on > #DNU pattern that > they guarantee a clear separation between meta-levels, i.e. even if > you send #doesNotUnderstand: > message to a proxy, this message will be trapped and handled by > handler as any other. > > Yes, and the cool thing is that proxies can intercept ALL messages, not only those that are not understood. Hence, you don't need to subclass from ProtoObject or nil. In addition (I think, I didn't measure) that they are faster because they don't need to the method lookup until getting to the dnu..but I am not sure about this last one. Another problem with the DNU is what Igor say, you are mixing both things: error handling and message trapping. So...when a proxy receives DNU, how do you know if it was a real dnu or because it needs to trap the message? Implementing method wrappers on top of such proxies is a matter of 5 minutes as much as I can see. In fact, in the paper I "implement" method wrappers as an example. You can take a look by Gofer new squeaksource: 'Marea'; package: 'GhostProxies'; package: 'GhostProxiesTests'; load. But remember it is just a prototype ;) The paper is almost finish... cheers -- Mariano http://marianopeck.wordpress.com

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Panu Suominen

2011/5/5 StÃ©phane Ducasse <stephane.ducasse(a)inria.fr>: > why copy is not enough for you? > and you redefine postCopy to copy the instance variables? Because it would involve manual labour for any object that is going to be used in business logic. This would create artificial requirement that is easily forgotten. However first try-out of this approach used postCopy in domain objects. I thought that copyTwoLevel is going to be deprecated because it is too specific. However I am not sure whether this is actually true. It depends how often copying objects is necessary. The deprecation is not a problem. I can easily just create utility class that copies objects like copyTwoLevel. -- Panu

May 5, 2011

Re: [Pharo-users] [Pharocasts] New comment on SandstoneDb, simple ActiveRecord style persistence....

by Bernat Romagosa

Thanks a lot everyone, I'll start migrating it right way :) Bernat Romagosa.

May 5, 2011

Re: [Pharo-users] WebClient for 1.2.2?

by Sven Van Caekenberghe

Andy, On 04 May 2011, at 18:01, Andy Burnett wrote: > Out of curiosity, what will it take to get SSL working? Is it a difficult > technical problem, or is it just a question of time/money? You are not the first to ask for this. This is really a recurring theme: why isn't HTTPS working ? I think it is doable, but it is certainly not trivial. It requires both a good understanding of the Smalltalk level (Socket, SocketStream, SSL concepts, streams/networking) as well as an understanding of the VM Plugin level (C code). Furthermore, there are 3 versions of the plugin (Windows, Mac OS X and Linux) each talking to a different OS API, to develop this further would potentiallyt/probably require knowledge of 3 platforms. Basic SSL is not that difficult, the finer points (certificates for example) are more challenging. Reaching adequate performance will not be easy (SSL is always slower, the question is how much slower). Finally, to test HTTPS, you also have to understand HTTP, of course. On the other hand, there is something that works more or less and that can be used as a basis to start from. Money would help, since you keep on asking, how much would it be worth to you ? Regards, Sven

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Norbert Hartl

Am 05.05.2011 um 08:24 schrieb Marcus Denker: > > On May 5, 2011, at 8:19 AM, Panu Suominen wrote: > >> 2011/5/4 andres <andres(a)lifia.info.unlp.edu.ar>: >>> Hi Panu. I don't know if immutability is available in Pharo, but if it is it >>> may work for you. You basically mark an object as immutable and register an >>> exception handler. When an immutable object's i.v. is assigned a new value, >>> an exception is raised and the assignment is left pending. There you can log >>> the change and effectively apply the assignment to the object. >> >> This sounds good but it seems that Squeak VM does not yet contain >> immutable objects. >> I am not sure. > > > No. They were implemented in the Squeak VM that NewSpeak uses. > So the code is out there. > Isn't there a WriteBarrier implementation somewhere. I don't know really but I think at least magma is using it. Norbert

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Marcus Denker

On May 5, 2011, at 8:19 AM, Panu Suominen wrote: > 2011/5/4 andres <andres(a)lifia.info.unlp.edu.ar>: >> Hi Panu. I don't know if immutability is available in Pharo, but if it is it >> may work for you. You basically mark an object as immutable and register an >> exception handler. When an immutable object's i.v. is assigned a new value, >> an exception is raised and the assignment is left pending. There you can log >> the change and effectively apply the assignment to the object. > > This sounds good but it seems that Squeak VM does not yet contain > immutable objects. > I am not sure. No. They were implemented in the Squeak VM that NewSpeak uses. So the code is out there. Marcus -- Marcus Denker -- http://www.marcusdenker.de INRIA Lille -- Nord Europe. Team RMoD.

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Panu Suominen

2011/5/4 andres <andres(a)lifia.info.unlp.edu.ar>: > Hi Panu. I don't know if immutability is available in Pharo, but if it is it > may work for you. You basically mark an object as immutable and register an > exception handler. When an immutable object's i.v. is assigned a new value, > an exception is raised and the assignment is left pending. There you can log > the change and effectively apply the assignment to the object. This sounds good but it seems that Squeak VM does not yet contain immutable objects. I am not sure. -- Panu

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Igor Stasenko

On 5 May 2011 01:59, Johan Fabry <jfabry(a)dcc.uchile.cl> wrote: > > On 04 May 2011, at 17:30, StÃ©phane Ducasse wrote: > >> >> mariano got a really nice proxy model based on the stratified proxy of igor (which is in 1.3). > > > Question: could this be used for PHANtom? We are now using Methodwrappers but there are some issues (fixed in PHANtom). Mariano, Igor, can you explain the advantages of your model wrt methodwrappers? > Sorry i'm not familiar with methodwrappers. The basic model is separation of proxy and handler. Proxy acts only as a message trap, while handler is actually responsible for handling messages trapped by proxy. Mariano builds on top of that, because he wants to replace classes and/or methods with proxies, and this requires a special handling (class proxy should obey VM-specific Behavior format, and method proxy should implement #run:in:as: when installed in method dictionary ). So, i guess that proxies which representing compiled methods could act similar to method wrappers. But of course you still have to implement all intall/deinstall management et al.. The only advantage of new proxies comparing to usual proxies based on #DNU pattern that they guarantee a clear separation between meta-levels, i.e. even if you send #doesNotUnderstand: message to a proxy, this message will be trapped and handled by handler as any other. -- Best regards, Igor Stasenko AKA sig.

May 5, 2011

Re: [Pharo-users] Object>>copyTwoLevel

by Johan Fabry

On 04 May 2011, at 17:30, StÃ©phane Ducasse wrote: > > mariano got a really nice proxy model based on the stratified proxy of igor (which is in 1.3). Question: could this be used for PHANtom? We are now using Methodwrappers but there are some issues (fixed in PHANtom). Mariano, Igor, can you explain the advantages of your model wrt methodwrappers? -- Johan Fabry jfabry(a)dcc.uchile.cl - http://dcc.uchile.cl/~jfabry PLEIAD Lab - Computer Science Department (DCC) - University of Chile

May 4, 2011

Re: [Pharo-users] [Pharocasts] New comment on SandstoneDb, simple ActiveRecord style persistence....

by Ramon Leon

On 05/04/2011 02:37 PM, StÃ©phane Ducasse wrote: > something that would be good is to have SActiveRecord as a trait like in SqueakSave > this way no need to change the superclass definition of your domain to save your objects. As far as I'm aware, traits don't do instance variables, of which active record has a few, as well as a class instance variable and a class variable. It also actually uses the class hierarchy throughout to allow classes to do queries on all its instances and sub instances. SDActiveRecord is a bundle of both behavior and state; not suitable for a trait. Beyond that, it's an opinionated library, your superclass should be SDActiveRecord, because you shouldn't be trying to use an object that has some other superclass as the aggregate root of a persistent bundle of objects. All of your business objects don't need to subclass SDActiveRecord, just the aggregate roots. Member biz objects can use whatever superclass they like as long as it's serializable. -- Ramon Leon http://onsmalltalk.com

May 4, 2011