[flexcoders] Re: TileList Loading shows duplicate ArrayCollection?

[jeffjvroom]

Hi Kevin,

(BTW, my flexcoders filter looks for various keywords in the subject 
like DataServices, Livecycle, etc. so I missed this one originally)

Are you using paging with your filled collection?   If you are not 
paging, we return the entire collection in one batch.  We do dispatch 
collection events as we are walking through the collection.  The UI 
controls should be watching these events and invalidating their 
display lists as necessary but the revalidation occurs during the 
next enter frame event.   I don't think it is possible for the 
redisplay to occur in the middle of us processing the fill result 
however so it seems like more likely you are paging in values if you 
are seeing visual artifacts as the data comes in.  Turning off paging 
might improve the display but obviously would have other side 
effects.  

If paging is on, we send the collection events as each page comes in 
(replacing NULL with the item for each index in the 
ArrayCollection).  So unfortunately the only way to listen for these 
events is the same way the TileList listens for them.  you could 
always listen for these events and use some trigger as the means to 
set the dataProvider for the TileList.  Or I think I sent you my 
NotPendingCollection.as class which can be used as a means to 
suppress pending errors altogether... putting that between the 
TileList and the ArrayCollection might yield better behavior.  

When you are processing pushed changes from the server, you do have 
more control.  You can turn off autoMerge.  The pushed changes then 
queue up on the client until you call dataStore.merge().  At that 
point all pending changes are applied.  Again, this method should 
invalidate the display list of any UI controls but the actual 
rendering will take place in the next enter frame event which 
should occur after all of the merged changes have been applied.  

Jeff 

--- In flexcoders@yahoogroups.com, Kevin [EMAIL PROTECTED] wrote:

 I am getting some odd behavior from the TileList when loading and
 ArrayCollection from the server.  I am not sure if this is an
 actionscript issue OR the way Data Services serializes the 
collections.
 
 I have bound the DataProvider of my TileList to a paged collection
 that get's automatically lazy loaded from the server.  When the
 collection comes in, you can see the TileList load with double the
 amount of records and then eventually redraw with the correct 
list.  
 
 1. Can I listen for the completed load from the server and then draw
 my TileList then?  It seems in efficient (if not dangerous) to keep
 drawing the TileList as the collection update.  I also would want to
 disable the TileList when new loads are getting pushed to the 
client.
 (These seem to be causing redraw issues.)
 
 2. Can I monitor the progress of the collection as it fills from the
 server and create a progress monitor.  Presumably we know how many 
we
 are expecting because data services pulls in the id's.  Is there a 
way
 to then monitor the load in realtime.
 
 OR
 
 3. Is this a problem in TileList and how the TileList lays out it's
 children?  Is it best practice to hide the component until all the
 children are laid out to avoid this.
 
 Thanks for the input on this one.
 
 - Kevin

[flexcoders] Re: FDS ManualSync and AS [Managed] Metadata

[jeffjvroom]

--- In flexcoders@yahoogroups.com, Jeff Glass [EMAIL PROTECTED] wrote:

 I have some questions about manual sync with LCDS 2.5.1.
 
 --- In flexcoders@yahoogroups.com, Jeff Vroom jvroom@ wrote:
 
  One other difference is that the refresh fill process - i.e.
  where the server refreshes fill methods after an insert or update
  or explicit refreshFill call is made on the server does not work
  with manual sync. Since the server has no knowledge of what fills
  each client is actively managing and does not cache its current
  set of ids, it can't refresh it.
 
 But the assembler doesn't know what fills are active, either. Or am 
I
 missing something? The assembler can keep track of a client's 
initial
 use of a fill (by saving the fillParams argument to the assembler's
 fill method), but can the assembler get notification when the fill 
is
 no longer being used (e.g., when the client goes away)?
 

Yes, this is the case.  The auto sync mechanism is what keeps track 
of which fills clients are looking at specifically.  You would have 
to duplicate this logic and it is not trivial.  To solve this 
generally is a bit tricky - you need to watch when fills are made, 
associate them with a session and remove that when the session 
expires.  You might also need to add a remote object call to release 
the collection if the client stops looking at the results of that 
query.  There could be short cuts based on how your app uses queries 
and what changes might affect those queries. 

Clearly this is something we need to address by supporting the paged 
fill functionality with auto-sync=true. 

  You can still update queries by using the
  DataServiceTransaction.addItemToFill and removeItemFromFill 
methods.
 
 In my experimentation, you have to call addItemToFill and
 removeItemFromFill with the *exact* fill parameters (unlike the
 matching that is done by refreshFill). Hence the need to keep track 
of
 which fills are active.
 
 Also, in my experimentation, the client only seems to get the
 collection-change event if the addItemToFill call has position=-1. 
Is
 that expected? In my application, the fills have a sort value, 
meaning
 that the new item could be in different positions in different fills
 anyway, so what I think the client should do is to just start the 
fill
 over again when an item is created. How would the assembler notify 
the
 client that it should do that? (That is, how do I get something
 similar to a refreshFill method returning EXECUTE_FILL in the
 auto-sync case?)
 

This sounds like a bug which is fixed in a hotfix we have. I think 
that the addItemToFill method would fail if the index was not -1 and 
there was no sequence.  

Again, we just don't have a great way to refresh an entire sequence 
when auto sync is false.  It sounds like you really need the auto-
sync=true with the paged fill support. 

  Of course the other server side methods updateItem, createItem, 
and
  deleteItem also still work though make sure that any use of these
  apis also supply the producerSubtopics or producerDefaultHeaders
  needed to route these changes to the right clients.
 
 What exactly are createItem and deleteItem good for in the manual-
sync
 case? They're not pushing anything to the clients.
 

You can route createItem messages to other clients but you also need 
to push some updateCollection or updateItem to add a reference to 
that item for it to show up someplace.  The deleteItem should just be 
handled automatically... when a delete is pushed to a client, it is 
automatically removed from any refernces.  

[flexcoders] Re: Urgent call for Flex Portlet (Flex Taglibs) specialists (James Ward, help!)

[jeffjvroom]

I'm not sure why this is happening unfortunately.  In both of those 
stack traces, the line of code looks like:

   String uri = request.getRequestURL().toString();

So either request is null or the request.getRequestURL() method is 
returning null?  I'm not sure why either of those would happen.  

Jeff  

 
 --- In flexcoders@yahoogroups.com, johan.wasserman 
 johan.wasserman@ wrote:
 
  I have tried and tried, fried my brain, and I am desperate.
  
  I firstly wrote a simple portlet and tried to present a flex page 
 on 
  JBoss portal, unsuccessfull because of the -
  java.lang.NullPointerException
  at flex.server.j2ee.cache.CacheKeyUtils.generateJspCacheKe 
  (CacheKeyUtils.java:68)
  
  OR
  generateSwfCacheKey(CacheKeyUtils.java:54) if I try to show the 
 mxml 
  file directly, without going through the taglib.
  
  - error.
  
  Then, with effort, learned how another portal server works, I 
  inastalled and configured Jetspeed 2, and after trying just about 
  every trick in the book, had to give up and realised the same -
  java.lang.NullPointerException
  at flex.server.j2ee.cache.CacheKeyUtils.generateJspCacheKey
  (CacheKeyUtils.java:68)
  
  - error haunting me.
  
  I have a simple, yet urgent requirement, I need to make Flex 
  applications available as portlets, on (prefferably) an Appache 
  opensource portal server.
  
  PLEASE will SOMEONE help me?!
  
  I'll send my JBoss and / or Jetspeed 2 installation, zipped, so 
you 
  can have the full config and everything I have done, I'll send 
the 
  MySQL databases for these portal servers, ANYTHING, just say what 
 you 
  need.
  
  The only common denominator is that both these portal servers are 
  based on Apache Tomcat
  
  PLEASE PLEASE ASSIST.
 

[flexcoders] Re: Lost commit on data source with polling AMF

[jeffjvroom]

What is happening here is that the latency introduced by the poll is 
causing the client to receive the response from its commit before a 
previously queued pushed message to that client.  By the time it 
receives the pushed message, it is stale - i.e. the previous 
version is not accurate anymore.  We should be sending those queued 
messages to the client before the client processes the response to 
its operation.  Thus the client would receive notification of the 
conflict two times in this case but it would be the same conflict so 
you would only have to resolve it once.  

The only workaround I can think of right now is to resolve that 
second (stale) conflict using acceptClient without reporting it to 
the user.  The problem though is knowing when this is a stale 
conflict versus a real new conflict.   There is a timestamp in the 
message which theoretically could be used to detect that but it is 
pretty tricky so I'm not sure I'd recommend trying this workaround.  
I am working on a fix now... once it is ready we could investigate 
the possibility of turning it into a hotfix.  

Sorry for the problems and thanks for reporting this issue.

Jeff





--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com 
Yahoo! Groups Links

* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/

* Your email settings:
Individual Email | Traditional

* To change settings online go to:
http://groups.yahoo.com/group/flexcoders/join
(Yahoo! ID required)

* To change settings via email:
mailto:[EMAIL PROTECTED] 
mailto:[EMAIL PROTECTED]

* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]

* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
 

[flexcoders] Re: fill() with multiple ArrayCollections

[jeffjvroom]

The deleteItem method implies you want to delete the method from all 
collections it lives in so I would not use that at all to remove the 
item from a group.  Usually, there is a property on the item itself 
which indicates which group it is in.  You might just set that 
property to null and commit that change.  If you have autoRefresh 
enabled for your fills, they would be reinvoked to refresh their 
contents.  The first one would still return the item, but the second 
one would no longer return the item because the item's group property 
was not set to that value anymore.  This assumes that the fill to 
retrieve a group's members is made on a query against properties on 
the item in the group.

It is also possible that the group membership is maintained by some 
separate table which is not part of the state of the item in the 
group.  In this case, the other way you can do this is to implement 
the flex.data.assemblers.Assembler interface (if you are not already) 
and override the removeItemFromFill method.  Then on the client, 
you just remove the item from the second collection using one of the 
ArrayCollection.removeItem or removeItemAt methods.  When you commit, 
your Assembler.removeItemFromFill method should be called which you 
can use to update that other table.  

Jeff

--- In flexcoders@yahoogroups.com, dreuimar [EMAIL PROTECTED] wrote:

 Hi everyone,
 
 In FDS I have a Java class that with no parameters fills an
 ArrayCollection with all possible items, and by passing a parameter
 (such as a group number) another ArrayCollection is filled with what
 items have been 'added' to a particular group. Now this assembler
 obviously has a deleteItem method. When an item is deleted from the
 master list it should delete from both with FDS, and it is. The
 problem I'm having is that when someone deletes something from one 
of
 these 'groups' I just want it deleted from that group, rather than
 from the whole database.
 
 Any ideas on how to pass parameters to deleteItem() without putting
 some property in the object or what I could do about this? (I think
 deleteItem only supports an object as a param.








--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com 
Yahoo! Groups Links

* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/

* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]

* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
 

[flexcoders] Re: FDS - how do I manage a single item?

[jeffjvroom]

Either with dataService.getItem, if the item exists and you have its 
id or dataService.createItem if the item does not exist and you want 
to create a new one.  

If you need to actually execute a query to retrieve the item (e.g. 
you are trying to retrieve a User by the lastName), you have to use 
the fill method to pass in the parameters and just have that fill 
method return a Collection with just one item.  In this case, the 
fill method returns an AsyncToken whose responder will be called when 
the fill results are returned to the client.  At that point, you 
could look at the filled array collection and call getItemAt(0) to 
get the first item in the list.

Jeff

--- In flexcoders@yahoogroups.com, Dmitry Miller [EMAIL PROTECTED] 
wrote:

 I have a managed class.
 
 [Managed]
 class User
 {
   
 }
 
 
 and in the main application .mxml file I have a reference of this 
type
 that I want to manage by a data service
 
 
 i.e.
 
 mx:DataService id=ds destination=userSvc /
 mx:Script
  [Bindable]
  public var me:User;
 /mx:Script
 
 Now, how do I get to populate variable me and get it maneged by 
userSvc?
 
 Thanks,
 
 --- Dmitry








--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com 
Yahoo! Groups Links

* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/

* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]

* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
 

[flexcoders] Re: FDS: Can I disconnect a specific client from the server?

[jeffjvroom]

It looks like this is something we (Adobe) are going to need to 
improve upon though there is a workaround for now.

First off a little background info.  There are two different ways you 
can manage subscription timeouts with FDS.  If you set the session-
timeout on the destination to 0, FDS ties the subscription for a 
given client to the flex session's lifecycle.  If your destination is 
setup this way, invalidating the session will remove the subscription 
for that consumer (this is the same as calling consumer.disconnect on 
the client).  This has the effect of disconnecting the consumer, and 
if that consumer is the last service using a given channel it will 
drop the connection.  Unfortunately this will not disconnect the 
producer and usually a client will have both a producer and consumer 
using the same channel.  So you'd still need to send a message to the 
client to tell it to disconnect the producer (as well as any other 
services using the channel).  Also, this assumes a cooperative SWF 
and does not address the hacked swf which can ignore this message.

Note however that for now at least there is a workaround for this 
hacked SWF case.  If you throw a runtime exception from the code 
which is handling a message, the RTMP channel is closed which will 
prevent the malicious SWF from continuing to use that connection.  
This is a short-term workaround at best and that behavior might 
change in the future (since it is not clear you always want an RTE to 
abort the connection).  I think we need to provide a more reliable 
way for the server to unconditionally close the RTMP connection for a 
given client and this is likely serious enough to warrant a hotfix in 
the near future.  

If you set an explicit session-timeout for the destination or leave 
it as the default (20 minutes), the subscription is not tied to the 
session and invalidating the session does not affect the consumer's 
subscription or connection state.   

Jeff Vroom

--- In flexcoders@yahoogroups.com, Dave Wolf [EMAIL PROTECTED] wrote:

 More reason to simply destroy their J2EE session which would let the
 container enforce the ban.
 
 -- 
 Dave Wolf
 Cynergy Systems, Inc.
 Adobe Flex Alliance Partner
 http://www.cynergysystems.com
 http://www.cynergysystems.com/blogs
 
 Email:  [EMAIL PROTECTED]
 Office: 866-CYNERGY
 
 --- In flexcoders@yahoogroups.com, Tom Bray tombray@ wrote:
 
  That would work if we could trust that our SWF hasn't been 
hacked. 
 We have
  people trying to hack our clients all the time and they're often 
the
 ones we
  want to disconnect in the first place.  If we relied on the 
client to
  disconnect itself, these guys would prevent that message from 
getting
  through.
  
  -Tom
  
  On 7/23/06, James Ward jamesw@ wrote:
  
 I can't find anything in the api:
   http://livedocs.macromedia.com/flex/2/fds2javadoc/index.html
  
   However, this could be easily achived my sending a message to 
the flex
   client (via messaging) which told the flex app to disconnect.
  
   -James
  
  
   On Sat, 2006-07-22 at 14:28 -0700, Tom Bray wrote:
No, I'm talking about disconnecting on the server-side with 
Java. In
other words, getting a reference to the server-side client 
object
using the FDS API and disconnecting it.
   
-Tom
   
   
   
On 7/22/06, James Ward jamesw@ jamesw%40cayambe.com
   wrote:
Yup. Look into the disconnect() method on Consumer and
DataService
objects.
   
  
 
http://livedocs.macromedia.com/flex/2/langref/mx/messaging/Consumer.ht
ml#disconnect
   ()
   
  
 
http://livedocs.macromedia.com/flex/2/langref/mx/data/DataService.html
#disconnect
   ()
   
-James
   
   
   
   
On Fri, 2006-07-21 at 14:07 -0700, Tom Bray wrote:
 I'm coming to FDS from FMS2 where I have the ability to
disconnect any
 client for any reason -- say after a period of inactivity or
because
 an administrator of the chat has banned a user. Can I do
this in FDS?

 Thanks,

 Tom




   
   
   
   
   
   
  

  
 








--
Flexcoders Mailing List
FAQ: http://groups.yahoo.com/group/flexcoders/files/flexcodersFAQ.txt
Search Archives: http://www.mail-archive.com/flexcoders%40yahoogroups.com 
Yahoo! Groups Links

* To visit your group on the web, go to:
http://groups.yahoo.com/group/flexcoders/

* To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]

* Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/