seaside August 2004

seaside@lists.squeakfoundation.org

30 participants
46 discussions

Nested form tags
by Bany, Michel 04 Nov '05

04 Nov '05

I came accross some weird situation when using a component that is a composition of other components. A subcomponent may in turn be a composition of some more components. To make things even more complex, some deeply nested subcomponents may call some other component. Some of the subcomponents may contain some input html tag and as such will require to be contained in a form html tag. In such a complex construct it may be difficult to decide exactly what component should have the responsibility to generate the form tag. Also, it happened frequently to me that form tags become nested especially when a subcomponent calls another component that happens to be a dialog. Therefore I am making the following suggestion : Seaside should prevent form tags from being nested. It should discard any form tag that occurs within a previously opened form tag. With this in place, I can design a component that needs a form with a form construct knowing that if my component happens to be used inside a form, then my form construct will be discarded. Otherwise my form construct will be preserved. If my component calls a dialog component, the form generated by the dialog component will be discarded if my component happens to be used from inside a form. Michel.

6 6

children & egg problem
by radoslav hodnicak 08 Sep '04

08 Sep '04

In 2.5, the complete root component with its children must be ready before authentication decoration kicks in - but I can't initialize my (non-root) components before I know who is logged in. In 2.3 this was not a problem because there was no #children. Suggestions? rado Btw, I don't know where this fascination with wrapping things into other things comes from, but do I have to remind you what happened to the current VisualWorks UI framework ("Wrapper")? Right, it's being rewritten ;-)

3 8

Re: Problems with expire in Seaside
by Avi Bryant 06 Sep '04

06 Sep '04

(Replying on-list so that others can benefit from the question) > Hi. > I've searched for this problem on the web but haven't found any > solutions so i was hoping that maybe you could give me a hint. > > I'm calling my Seaside application with a query parameter specifying > which company database it should access > (http://localhost:9090/seaside/app?site=1111111111). > The problem is that when a page expires, Seaside redirects to > 'http://localhost:9090/seaside/app' which makes it impossible to login > again. > Is there any way to alter the WAApplication's baseUrl from the code to > include the parameter? Or any way to override the default > expire-behaviour? That's a tricky one - the whole point of expiring a session is to free up the memory associated with it, and so there's nowhere to maintain the state of which particular site that session was for. And it does need to be session-specific - you don't want to alter the application's base URL because that will affect all sessions in that application at once. It seems like maybe we want a mechanism almost like finalization in garbage collection, where the session itself doesn't stick around but is optionally replaced by some other, smaller object that stores a few pieces of info about the session (like, in your case, the site number), and gets control if any further requests come in for that session. It can then display the right message, do the appropriate redirect, etc. Can someone think of an appropriate name for this object? A "forwarder", maybe? (as in, the session is gone but it left a forwarding address). The Forwarder should expire too, but after a much longer period of time than the session. This shouldn't be too hard to implement - it'll just take some messing around in #unregisterExpiredHandlers so that you send #forwarder to anything that's expired, and throw the result (if non-nil) into the new keysByHandler and handlersByKey dicts, using the same key as the expired handler its replacing. Then put a default #forwarder method that returns nil onto WARequestHandler, and override it on your session subclass to return a custom forwarder object. This will get sent #handleRequest: (and be expected to return a WAResponse) if any requests are sent to that session once its expired, and it can do whatever it likes from there. Just make sure it doesn't keep a reference to the whole session object, which would defeat the purpose. You'll also need an #isActive method (quick and dirty is just to return 'true' so that the forwarder never expires). Unless someone can point out why this is a bad idea, or unless someone else implements it first, I'll probably post a version of the above on Monday. Cheers, Avi

5 14

MVP on Seaside
by Sebastian Sastre 02 Sep '04

02 Sep '04

Hi all, do you ever think about making an MVP framework on top of seaside? I mean, there is a presenter, obviously there are model (your objects), the only thing is missing here is something that decouples the view from the application-model. Actually, if you subclassify a WAComponent, with your own, components (presenters) the application model is coupled with the view (rendering directives). So you have to 'draw a lot' to make your application go where you want to go. I make the propose that a WebView separated hierarchy, could be analogous to the View hierarchy on an ordinary ST. For instance: WebTextInputView, WebTextAreaView, WebDateView, WebListView, WebBatchSelecionView, etc, etc. In the other hand, we can have a WebPresenter hierarchy, with WebDatePresenter, WebTextPresenter, WebListPresenter, WebButton, WebShell, and so on. Note: Perhaps the batch selector, instead of being a view, could be a presenter, implemented using WebListView and buttons to control the batch in the own WebBatchPresenter (application-model). In that way, a set of WebPresenters could have the rememberance of the views created for each presenter. Ideally, you can make a subclass of WebShell, call it MyApplicationMainWebShell, create a beautiful set of web view components for it with some web view composer, lovely dedicated to web expert graphic designers. Then, developers can use those saved views, and choose one to tie each 'web view component' the main shell has, to the corresponding subpresenter it has. At that time events can be hooked if necessary and models can be fexibly asigned to subpresenters. As I see is very posible, and convenient. waht about this? anyone could see I'm wrong? regards, Sebastián Sastre HYPERLINK "mailto:ssastre@seaswork.com.ar"ssastre(a)seaswork.com.ar HYPERLINK "http://www.seaswork.com.ar/"www.seaswork.com.ar --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.744 / Virus Database: 496 - Release Date: 24/08/2004

2 4

RE: [Seaside] registerObjectForBacktracking, submit and execute b lock
by Bany, Michel 02 Sep '04

02 Sep '04

This subject was already discussed here. See http://lists.squeakfoundation.org/pipermail/seaside/2004-June/003107.html I have actually fixed the problem with a bit of refactoring of the snapshot logic by moving the snapshot behavior to Object. I defined the following methods in Object . snapshotCopy . restoreFromSnapshot: . isIdenticalToSnapshot: And I changed the following methods in WAStateRegistry accordingly . registerObject: . restoreObject:fromSnapshot: . snapshot . snapshot:isIdenticalTo: The default behaviour remains based on shallowCopy for the snapshot. You may however re-implement the three Object methods as needed in subclasses to use a deeper copy. Michel. > -----Original Message----- > From: seaside-bounces(a)lists.squeakfoundation.org > [mailto:seaside-bounces@lists.squeakfoundation.org]On Behalf > Of Andreas > Nilsson > Sent: Sunday, August 29, 2004 2:44 PM > To: The Squeak Enterprise Aubergines Server - general discussion. > Subject: Re: [Seaside] registerObjectForBacktracking, submit > and execute > block > > > Ok, that makes sense, didn't know that it used an array internally. > Any ideas of how to make backtracking work for adding / > removing items > from a collection? > Somebody must have tried before ;) > > /Adde > > On 2004-08-29, at 12.29, radoslav hodnicak wrote: > > > On Sun, 29 Aug 2004, radoslav hodnicak wrote: > > > >> Registering object for backtracking doesn't do what you probably > >> think it does. It's only sensitive to changes of objects, *not* > >> object's > >> contents. > >> > >> holder := WAStateHolder new. > >> > >> holder contents: (OrderedCollection with: 1 with: 2). > >> > >> holder contents add: 3. "NO CHANGES HERE, STILL THE SAME OBJECT IN > >> HOLDER" > >> > >> holder contents: (holder contents copyWith: 3). "NEW > OBJECT IN HOLDER, > >> WILL BACKTRACK" > > > > Ok I didn't explain this good enough - backtracking works > for changing > > object state too - but it works on shallow copy so the > changes it can > > detect are only one level deep. > > > > Take this example with OrderedCollection. The items are internally > > held in > > an array. > > > > anOrederedCollection (object1) > > --has anArray(items ...) (object2) > > > > when you register object1 for backtracking, it makes a > shallow copy of > > it, > > which means they share the same instance of object2. And on > backtrack, > > it > > looks whether object1 = object1 copy - and it does, because > adding and > > removing items changed the same array (object2). If you changed a > > simple > > instvar holding a string or whatever, then it would work as > expected. > > Clear enough? > > > > rado > > -- > > gentleman, n. - a man who knows how to blog, but doesn't > > > > _______________________________________________ > > Seaside mailing list > > Seaside(a)lists.squeakfoundation.org > > http://lists.squeakfoundation.org/listinfo/seaside > > _______________________________________________ > Seaside mailing list > Seaside(a)lists.squeakfoundation.org > http://lists.squeakfoundation.org/listinfo/seaside

5 10

current/latest version for VisualWorks
by Travis Griggs 01 Sep '04

01 Sep '04

I thought I'd play with Seaside this afternoon, for VisualWorks. The problem is, that if I open the published items on the OR, and enter "seaside" in the filter field, there are two scrollbars worth of packages/bundles to choose from. What should I load? Any chance that some of the old stuff could be cleaned out? TIA -- Travis Griggs Objologist One man's blue plane is another man's pink plane. ----------------------------------------- DISCLAIMER: This email is bound by the terms and conditions described at http://www.key.net/disclaimer.htm

2 3

Re: [Seaside] Performance!
by Jon Paynter 31 Aug '04

31 Aug '04

---------- Original Message ---------------------------------- From: Avi Bryant <avi(a)beta4.com> Date: Tue, 31 Aug 2004 19:36:44 +0200 > >On Aug 31, 2004, at 7:27 PM, Jon Paynter wrote: > >> But I found when working with friends to test remotely their response >> times were very slow. No timeout errors from IE, but typical >> click-response cycle was 3-10 seconds. > >3 to 10 *seconds*? I've never seen anything like that. Seaside does >sometimes feel sluggish when there's high latency, because by default >it redirects after every request, which doubles the round trips. But >it should certainly still be sub-second unless your responses are huge >or you're doing some heavy computation. > >Just out of curiosity, have you tried any other browsers, or just IE? yes 3-10 seconds. I had my brother (on the phone) tell me when he clicked a link and then when his page finally displayed. No heavy computations.. just simple forms & tables. Im wondering if it has to do with the VW web toolkit im running this thing under? I know there is a bunch of logging & such that it does by default, but shutting that off didnt seem to help any. And no -- only IE has been used as a browser. I suppose I could try Opera or something, but it doesnt exhibit any of the symtoms of being a browser issue like early versions of VisualWave with early versions of IE (30 seconds to _parse_ a page in some cases) Anyone have a VW Komanche port I could try? Jon.

1 0

Re: [Seaside] Performance!
by Jon Paynter 31 Aug '04

31 Aug '04

---------- Original Message ---------------------------------- Date: Tue, 31 Aug 2004 19:10:38 +0200 >I just checked out the counter example Borges (The seaside clone in >Ruby) with Webrick and thought it was a bit slow. >So i brought up the same counter example in Seaside in another browser >window and WOW it's really fast. >This is something that you get used to when using Seaside exclusively >(at least i did), you really need to compare it to something else to >appreciate the speed. >I just thought I'd remind you all ;) Yes on a local LAN it IS vast.. very fast. But I found when working with friends to test remotely their response times were very slow. No timeout errors from IE, but typical click-response cycle was 3-10 seconds. No visible CPU spikes on my end. and connections to my IIS webserver on the same machine gave normal sub-second response times back. Ideas anyone? Jon.

3 2

Performance!
by Andreas Nilsson 31 Aug '04

31 Aug '04

I just checked out the counter example Borges (The seaside clone in Ruby) with Webrick and thought it was a bit slow. So i brought up the same counter example in Seaside in another browser window and WOW it's really fast. This is something that you get used to when using Seaside exclusively (at least i did), you really need to compare it to something else to appreciate the speed. I just thought I'd remind you all ;) /Adde

1 0

Re: [Seaside] current/latest version for VisualWorks
by Jon Paynter 31 Aug '04

31 Aug '04

---------- Original Message ---------------------------------- Date: Mon, 30 Aug 2004 16:14:34 -0700 > >On Aug 30, 2004, at 14:35, Michel Bany wrote: > >> Travis Griggs a écrit : >> >> > I thought I'd play with Seaside this afternoon, for VisualWorks. The >> > problem is, that if I open the published items on the OR, and enter >> > "seaside" in the filter field, there are two scrollbars worth of >> > packages/bundles to choose from. What should I load? >> > >> You need to load a bundle : either SeasideForWebToolkit or >> SeasideForSwazoo. >> These two bundles are made of many bundles/packages that show up in the >> list. >> > >OK, I tried the SeasideForWebToolkit one, and got the stack dump >attached below the sig. This is in a 7.2.1 image, loaded with a couple >of tools, but otherwise, pretty pristine. Any help much appreciated. > >-- >Travis Griggs >Objologist >My Other Machine runs OSX. But then... so does this one. Travis, Go out the the cincom public repository. There are published parcels you can load. Make sure you read the comments on each of the versions. They say which versions to load and not to load. If that doesnt work for you, let us know and i'll send you the VW file out a made from the Seaside parcels. Jon.

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

seaside August 2004