to answer your question,
jfv
> On Dec 12, 2017, at 10:49 AM, Jean-François Veillette
> <jean_francois_veille...@yahoo.ca> wrote:
>
> You might need to create a new jdbc connection for every user.
> If I understand your situation correctly, I would try to use a unique E
You might need to create a new jdbc connection for every user.
If I understand your situation correctly, I would try to use a unique EOF stack
for every session (user). In practice this can work well as long as you don’t
have too many sessions (exhausting DB connection).
You will have to change
Hi Remy,
I used polymorphic relationship, in my case I used ERXLongPrimaryKeyFactory
(splitting Long primary key numbers into 2 parts, 1 sequence and 1 entity
index), to help EOF a bit.
In the model you describe, I really think you should change it a bit.
- an object should not change class ‘on
Hi André,
You are doing great learning WebObject by yourself! I encourage you to continue
and learn about WOnder too for many reasons.
While you continue learning on your own, you will find out that there is a
pitfall in setting connection dictionary like you did (once in the early
startup of
+1
jfv
> Le 6 févr. 2017 à 08:11, Jérémy DE ROYER a
> écrit :
>
> Is there anyone alive ?
>
> Jérémy
>
> ___
> Do not post admin requests to the list. They will be ignored.
> Webobjects-dev mailing list
<b...@salient-doremus.de> wrote:
>
> Hi Jean-François,
>
> so you suggest using RPC to circumvent takeValuesFromRequest and invokeAction
> triggered by the AjaxSubmitButton?
>
>
>> Am 20.01.2017 um 14:46 schrieb Jean-François Veillette
>> <jean_franco
Thinking about it …
It seems the problem partly come from having a repetition or some kind or a
form of sequential access to your textfield bindings.
Maybe you could avoid that by pointing directly to it by means of RPC? So use
an AjaxProxy object, that talk to your server-side controller
> On Jul 18, 2016, at 6:33 AM, Jean-François Veillette
> <jean_francois_veille...@yahoo.ca> wrote:
>
>
>> No other (reasonably simple) way you would know of to hook into EOF to
>> obtain the desired effect, i.e., to be able to determine the target entity
>>
> No other (reasonably simple) way you would know of to hook into EOF to obtain
> the desired effect, i.e., to be able to determine the target entity _before_
> a fetch?
I have used such mechanism to good results. It was largely inspired by
ERXLongPrimaryKeyFactory, it worked well.
I could
On Jul 12, 2016, at 8:52 AM, Paul Hoadley wrote:Hello,I’m working on an app where some page-level informational messages (such as validation problems) are stored in ERXThreadStorage, and displayed by the page wrapper. This works great for full page loads.In a few places, I’m
it be that
fr.univlr.colloque.components.admin.ajax.ListeUtilisateursGroupe is returning
an AjaxResponse but the handler expect a WOComponent?
We need more stack frame to be able to identify where you are in the RRL and
what could be the cause of this.
jfv
> On Apr 25, 2016, at 9:13 AM, Jean-François Veillette
> <jean_francois_veille...
Bonjour Raymond,
La stack manque plusieurs niveaux.
Selon le peu que l’on a comme information, on en comprend que la requête a bien
été exécuté, a retourné une AjaxResponse, mais que le « handler » s’attendais à
recevoir un WOComponent.
There is an userInfo dictionary in the editing context, you can store stuff in
there as well. At page level, when you have access to all contextual
information (local, user, etc.) it’s time to set that in the editingContext
userInfo dictionary before you trigger your core methods.
Depending
23:12 GMT+01:00 Jean-François Veillette
jean_francois_veille...@yahoo.ca mailto:jean_francois_veille...@yahoo.ca:
WOSwitchComponent?
https://wiki.wocommunity.org/display/documentation/WOSwitchComponent
https://wiki.wocommunity.org/display/documentation/WOSwitchComponent
I used it with great
Hi Ronald,
is your B component syncronizing bindings?
If so, your B instance should receive a call equivalent to «
b.takeValueForKey(session().authenticatedUser(), “userAccount”); » so you
should comply to KVC for this key (public ivar, get/set method .. your choice)
to get it assigned and
area ofmy central page.
Thanks for your help.
2015-05-29 16:57 GMT+01:00 Jean-François Veillette
jean_francois_veille...@yahoo.ca mailto:jean_francois_veille...@yahoo.ca:
Hi Ronald,
is your B component syncronizing bindings?
If so, your B instance should receive a call equivalent
At my previous workplace, we did the switch to Maven. Luckily we had a real
maven expert to drive the move.
We started with around 50+ projects, all ant based, using the ‘standard’
fluffy-bunny layout. He added pom.xml here and there, and everything just
started working with maven. We had
Look at « autorevision »
https://github.com/Autorevision/autorevision
it's a little script that output detailed version information even when there
is no apparent version scheme.
We are looking at it to build an « About this app » page (direct-action) that
will display the version of the .woa
On first sight, it doesn't seem like I set another delegate.
I'll take time to investigate it further in the next few days.
Thanks!
Le 2014-05-08 à 11:06, Ralf Schuchardt r...@gmx.de a écrit :
Am 08.05.2014 um 05:36 schrieb Jean-François Veillette
jean_francois_veille...@yahoo.ca
PrimaryKey obtained by permanentGlobalID() are not saved as the primaryKey on
saveChanges().
I'm setting a global id as a reference for « soft relationship » (since it
point to undefined object) instead of a « hard relationship » (usual when we
store the pk).
The problem is that I get 2
Here is how I implemented such attribute:
{
adaptorValueConversionClassName =
er.extensions.foundation.ERXPropertyListSerialization;
adaptorValueConversionMethodName = jsonStringFromPropertyList;
allowsNull = Y;
columnName = dictionary;
Fix:
So when you need to deal with an immutable version, do it with
mutableDictionary().
So when you need to deal with a MUTABLE version, do it with mutableDictionary().
jfv ___
Do not post admin requests to the list. They will be ignored.
Le 2014-03-18 à 12:08, Altera WO Team webobje...@altera.it a écrit :
I can quote every single line of Maik's email.
We have a lot of happy clients with running WO apps and we use WO all the
time. The problem of finding good new developers is the same, we need good
developers and that is
I do what Lars recommended and it work!
I did a compare of the file before/after and the only diff is the number of
that properties:
org.objectstyle.wolips.componenteditor.sashWeights=805,194
so you can simply remove that properties and it will work again.
From what I looked at, the fix from
You might consider using a brand new EOF stack (create a new Object Store
Coordinator) for the background task. The advantage would be that you are the
only on using the database connection, no jdbc locking problem with the main
thread.
The disadvantage is that you do not share the snapshot,
In fact the problem with to-many relationship traversal come from the fact that
we are used to expect sql-like behavior from the memory evaluation when in fact
EOF mangled the qualifier for us.
We forget that the database behaviour is obtained by __transforming first__ the
qualifier. ...
It look like a RPC communication stream between the client (?GWT?) and server.
Le 2013-07-22 à 04:06, Helmut Tschemernjak hel...@helios.de a écrit :
I just watched this video and can see some data structures related to Java.
http://www.youtube.com/watch?v=q000_EOWy80
So far I don't know if
It can be done by using « git filter-branch ». This is what I used when I
converted from cvs.
Beware that since the author is part of the hash calculation, changing it would
make a different a different history. So if you have references outside of
your repo that refers directly to hash code,
), an Application.java that works and Henrique's
JarResourceRequestHandler. See if that helps...
On 25/09/2012, at 6:20 AM, Jean-François Veillette
jean_francois_veille...@yahoo.ca wrote:
I need some help to debug maven issues from inside Eclipse [*a*].
I converted some existing (and working) projects
I need some help to debug maven issues from inside Eclipse [*a*].
I converted some existing (and working) projects, both frameworks and
applications.
1- When I run the application, I now (after the switch to maven) need to
explicitely set the class to be used as Session and some components
Possible cause would be binding synchronisation.
Try to disable automatic sync binding on TestsPascal, or trace and debug those
sync calls to see if they are causing this.
--
jfv
Le 2012-09-18 à 14:27, Pascal Robert a écrit :
Hi guys,
I have a table with a chunk of AjaxUpdateContainer and
The first page is a direct action login page. The user logs in with a key,
the direct action class verifies the key and if the key is ok redirects
automatically to the Main page. The key has to be passed from the URL using
appendToResponse.
1. Login DA
2. Authenticate
3. Is ok show Main
It sound as if you know which page should be the next page a bit too late.
Are you in the default RRLoop (.../wo/...)? Or maybe in an Ajax call?
The default Request-Response-Loop is designed so that such decision are made in
invokeAction(). And it's in that phase / method that you are given
Le 2012-08-02 à 14:41, Pascal Robert a écrit :
Well, it's quite rare that I ask question instead of answering questions, but
anyway I have one. I just started using AjaxProxy because I need to take data
coming from JavaScript and send that data to a NSDictionary.
Problem is: when I call
I'm about to migrate a large project which is currently managed in subversion
to git. Git experience is still limited. Besides wonder frameworks, the
project consists of
- one application
- several frameworks that are more or less specific to the application
- several more generic
I'm trying to implement autocomplete based on my database and I want to use
AjaxProxy for that. I have two different fields on the same page that need to
be auto completed. I create two AjaxProxy objects:
FirstAutocompleteProxy : PMAjaxProxy {
name = autocomplete;
proxy = firstProxy;
Upgrading an old Window's machine, I decided to give java 1.6 a try.
I was careful to pick the Multi-Language (the only available)
version, but still I got the NSMacOSRomanEncoding not recognized error.
Is macosroman encoding removed from the win* jvm in 1.6 ? Is it
expected ?
- jfv
manually change few installed files as well in order to make
womonitor working.
- jfv
Le 08-09-12 à 15:37, David Avendasora a écrit :
On Sep 12, 2008, at 3:03 PM, Jean-François Veillette wrote:
I decided to simply change the .api files of WO (and Wonder) since
they are ascii anyway.
Hi
Le 08-08-28 à 08:12, Dov Rosenberg a écrit :
...
We replace the connection dictionary at startup. Could the
difference in EOEditingContexts cause a second
EOCooperatingObjectStore to be added?
If by replace the connection dictionary at startup you mean in
'Application' or something
Le 08-08-28 à 09:51, Jonathan Ricker a écrit :
On Aug 28, 2008, at 8:25 AM, Jean-François Veillette wrote:
If by replace the connection dictionary at startup you mean in
'Application' or something like that, then yes this will cause
problem. I've seen some methods in EOUtilities
Le 08-07-22 à 08:36, Steve Quirk a écrit :
On Mon, 21 Jul 2008, Jean-Fran?ois Veillette wrote:
I use this one to reformat java source :
/path/to/eclipse -application
org.eclipse.jdt.core.JavaCodeFormatter -verbose -config /Path/To/
Perfect/Setting/.settings/org.eclipse.jdt.core.prefs
I use this one to reformat java source :
/path/to/eclipse -application org.eclipse.jdt.core.JavaCodeFormatter
-verbose -config /Path/To/Perfect/Setting/.settings/
org.eclipse.jdt.core.prefs /Path/To/Source/To/Reformat/
What would be the invocation to reformat .wo/*.{html,wod} files ?
I
backup and your fixed models, the changes you see are
the one you will have to make at runtime, in that notification.
- jfv
On Jul 16, 2008, at 5:09 PM, Jean-François Veillette wrote:
Whenever used, we use the notification of model loaded, and at
runtime change whatever values needed
Whenever used, we use the notification of model loaded, and at
runtime change whatever values needed to be tweeked for the current
project. Like changing table name, setting isAbstract (or not),
playing with relationship, etc.
Here is a simple case where we define the isAbstract attribute
Le 08-06-10 à 14:54, Gennady Kushnir a écrit :
Hi' all
I've started experimenting with git recently.
It required compiling as I'm not on Leopard yet.
I'm still on tiger, just './configure ; make ; sudo make install '
and you are done.
First I was confused, how to use it with Eclipse -
I have 2 models, one which contain a set of abstract eoentity,
another one which contain a set of concret eoentity (subclasses of
the abstract set).
It was working just right, but suddently after reinstalling one of
the framework, it just stopped working. I can't make it work again.
Le 08-05-11 à 15:50, Gennady Kushnir a écrit :
Thank you all for your opinions.
I see all hands are for SVN except built in Eclipse integration.
Count me as one hand down for cvs/svn ... mainly for its tag/
branching model.
Count me as a Git happy user !
Maybe I should have used google,
I do agree, though, that your partial disconnectedness makes CVS/
SVN a bit of a tougher workflow. However, I would be nervous of
turning you off to version control by going to git out of the gates
(Brought to you by the Makers of the Linux Kernel :) ).
It matters far more that you use
As other mentioned, WO is designed around multiple design patterns.
Implement solution from the use cases ... without any WO connection
at all, only pure java and EOF. At that point, the business layer is
implemented and self contained, including editing context management
(use cases
.. about one a month, so haven't worried
too much about it. But is there a rule of thumb about which
attributes should have it turned on, and, why when you create an
attribute it is always on?
Thanks
James Cicenia
On Mar 22, 2008, at 8:51 AM, Jean-François Veillette wrote:
In addition
In addition to the good comments from Guido, I would also point out
that all your attributes are set for 'optimisting locking' (the
padlock). You probably do not want that. In fact, considering that
you might not have the time to deal with locking problems, you
probably want to remove
Assuming the relation between a Person and a Convocation is a many-to-
many with a qualified state, you already have your pk and fk, just
not marked appropriately.
In PersonXConvoc, both id_person _and_ id_convocation should be
primary key (as in ExpertiseXPerson).
In EOModeler, you probably
21, 2008, at 12:46 PM, Jean-François Veillette wrote:
Assuming the relation between a Person and a Convocation is a many-
to-many with a qualified state, you already have your pk and fk,
just not marked appropriately.
In PersonXConvoc, both id_person _and_ id_convocation should be
primary key
subversion is a nice improvement over cvs, but if you are going to
switch, now is the time to take a look at the other options ...
Take a look at git (and/or mercurial) :
http://www.kernel.org/pub/software/scm/git/docs/cvs-migration.html
I think git's branching model is way better than the
The only solution I see right now is to modify the ajax request to
return in a structured way what are the needed resources and build a
js library that will interpret this and adjust the dom to include
those needed resources.
- jfv
random thought ...
I think the problem come from the
I'm relatively new to unit testing, junit, selenium and all that stuff.
I've built simple test so far for logic that didn't involve database
operation and didn't invole much of eof in fact.
I have to modify an existing project and would like to start building
test cases for it, slowly but
add the '@' sign, try [EMAIL PROTECTED] instead of aa.count.
- jfv
Le 07-09-24 à 08:49, Mike Schrag a écrit :
It sounds like you are calling count on a relationship in your
template, which is not a method on eorelationship.
ms
On Sep 24, 2007, at 5:04 AM, Jan Grathwohl
[EMAIL PROTECTED]
Last time I remember, .gif did not support alpha values. Gif only
have a concept of mask, where one color is treated 'special',
absolutely invisible.
If you want real alpha value, look at png (and IE workaround).
- jfv
Le 07-08-03 à 21:12, Anthony Arthur a écrit :
How can I preserve the
We have solved this problem by enhancing kvc at the component level.
Let's say you have those attributes : name_FR, name_EN, name_DU, etc.
which holds localized values for the name attribute.
what we do is we alter the keypath that ask for 'name'.
instead of wostring: value= myEO.name;
and let
from your description, it look like 'formValues()' wrap all values in
an array (from the doc).
formValues()
Returns an NSDictionary containing all of the form data
with names for keys and NSArrays containing the value(s)
associated with
Le 07-06-30 à 09:16, Paul Marvine a écrit :
Error Message
/Application Server.build/JavaFileList'
myTestPage.java:33: setXPage(xPage) in EditPage2 cannot be applied
to (com.webobjects.eocontrol.EOEnterpriseObject)
nextPage.setXPage( xPage );
^
1
The component TextFieldCalendar should not synchronize it's bindings.
- jfv
Le 07-06-29 à 10:50, Steven Mark McCraw a écrit :
Hi all,
I've always thought that using caret notation was a good idea when
creating reuseable components, but it only seems to work for me
about half the time, so
Le 07-06-10 à 04:18, Thierry Kramis a écrit :
this might be a nice solution. What you're asking to do is possible,
but certainly more work...
How would I realize this then?
Consider that all the comments bellow have not been tried by me, they
are pure _speculation_ based on what I know
I would expect (didn't try exactly this case) that by having each
files in a different directory (Classes/com/abc/Customer.java and
Classes/com/xyz/Customer.java) would work correctly.
- jfv
Le 07-06-08 à 15:22, Joe Moreno a écrit :
Is there a trick to adding two classes, with the same
Before sending the xsl-fo stream to FOP, can you save it and analyse
it independently ?
I would be interested to see the fo stream that is causing problem.
Once we can identify the xsl-fo problem, we will know what we should
generate from webobjets.
- jfv
Le 07-06-05 à 08:35, WO Dev a
Here is a method that 'escape' some chars.
We call this method before inserting a string (comming from an eo
attribute for example) into a fo-xml stream.
static public String transformeCaracteresSpeciaux(String chaine) {
if (chaine != null) {
another javascript library: http://jquery.com
I only looked at it briefly, done some demo, looked at the plugin
architecture, impressive and interesting.
- jfv
Le 07-05-23 à 09:37, Mike Schrag a écrit :
A short version of the question: which toolkit? why?
Prototype/Scriptaculous are sort
James, stop it, It can't be done like that. Think about another way
to do it.
There are many options to solve this problem, but not knowing anymore
on how your app is organized, I would try to give extra argument to
the generated hyperlink (http://www.aa.aa/cgi-bin/WebObjects/xxx.woa/
Le 07-05-07 à 09:32, James Cicenia a écrit :
So after refactoring my code, I figured out how to use that
component, get the url from it, and embed
it in my javascript YUI Treeview. Everything seems to work great
except for one big confusing problem.
My link's action binding isn't being
that doesn't
work... yet its appendToResponse is always firing so I know it is
getting rendered
and of course the link show up too so the it does its Hyperlink
correctly.
So hard to debug this type of problem.
- James
On May 7, 2007, at 12:23 PM, Jean-François Veillette wrote:
The code
Carter, note that you most likely do not want to change your list
before the invokeAction() phase.
Think about what would happend if you ask to delete the 3rd entry in
the list you see on screen, but once submitted and before the action
is taken, the liste is reordered/refetched/etc.
The
or store it in either context().request().userInfo() or in
context().response().userInfo() dictionary. look at the doc, for a
which one fit better your needs (most likely the context().request()
one).
- jfv
Le 07-04-20 à 21:52, Kieran Kelleher a écrit :
For stateless subcomponents
You might consider passing informations through worequest.userInfo
dictionary as an alternative to bindings.
From a sub component, you can still do bindings like value =
context.request.userInfo.theKeyToSomethingINeed; .
- jfv
Le 07-04-19 à 13:37, James Cicenia a écrit :
A navigation
Thanks Mike, I have a better picture now of what/where/how
ERXThreadStorage is helping.
I still think that I don't want per thread storage, we already have
request.userInfo for that and I use it only for web/request related
informations (used at the interface layer only).
I still feel
Fred, consider that having an EO access a Session objet can be seen
as a bad practice (by me at least).
That being said, an EO frequently need to access the current logged
in user, either for validation or for business processing.
Then you realize that the concept of a current logged in user
:47 PM, Jean-François Veillette wrote:
...
My next implementation at solving this problem will turn around a
common 'context' class that will have static methods to return the
current 'consultation' object given an EO (internally using the
eo.editingContext() has a key to map
Since I do not use Eclipse (yet), I can't try for myself. I'll
eventually do the switch, but meanwhile, i'm wondering if anyone have
any comments on this tool : http://labs.adobe.com/technologies/jseclipse
- jfv
___
Do not post admin requests
I have never tried but, look at this demo, this look promising !
http://www.terracottatech.com/TC_container.html
- jfv
Le 07-03-01 à 00:55, Mike Schrag a écrit :
In Java, static variables are unique per-class-per-class-loader.
That is, if you have the same class loaded in two separate
Le 07-01-31 à 12:58, Thierry Kramis a écrit :
Depending on what exactly your goal is you might look at these:
1. Apache FOP ( http://xmlgraphics.apache.org/fop/)
2. iText ( http://www.lowagie.com/iText/)
3. jFreeReport (http://www.jfree.org)
4. UJAC ( http://ujac.sf.net ) (similar to fop
look at NSBundle, it's the foundation equivalent of WOResourceManager.
- jfv
Le 07-01-26 à 05:57, Logi Helgu a écrit :
I have a unit test that needs to get data from a text file in my
languge(
lproj ) folder and I'd like to set it up without an absolute path
( like
I've seen this as well. It occurred when I use the browser 'back'
button.
But I was also able to replicate the problem when in a form, I
download a file (which doesn't refresh the page, the response get
written to disk), after that, when the form get submitted, all values
are null.
In our
Look at http://www.aflax.org
from the first page :
AFLAXtm (Wikipedia Entry) stands for Asynchronous
Flash and XML. Defined simply, AFLAX is a development
methodology which combines Ajax and Flash to create
more dynamic web based applications.
Developed by
Make sure that everytime you convert from binary to string you
specify the encoding.
Same thing when you convert from string to binary, explicitely set
the encoding.
If you do not specify the encoding, java will use the platform
default (MacRoman on OSX) to do the conversion.
I haven't
Depending on what you really need, JooReport (
http://jooreports.sourceforge.net ) might be enough.
- jfv
Le 06-08-11, à 11:04, Pascal Robert a écrit :
On 06-08-11, at 08:00, Inessa Weiner wrote:
Hello,
Our client would like an ability to generate a letter - MS Word
document from Web
If you are not familiar with WOnder enough, here is another solution to
the problem.
We use .strings files (see WOResourceManager for localization) and
prefix the number with a constant.
It end up like :
{
Status_0 = stopped;
Status_1 = running;
...
}
We overwrite KVC at the
Chip,
try to capture selected/unselected as part of 'takeValueFromRequest' of
your component. One way to do is to use a selection method like this :
Object myItem; // this is your item of your repetition
NSMutableDictionary itemSelected; // probably instantiated in your
constructor, cleard as
use itemSelected as the
checked value.
On your action method, you know which item was checked by looking at
the itemSelection dictionary and looking for a true value.
- jfv
Le 06-07-19, à 13:34, Jean-François Veillette a écrit :
Chip,
try to capture selected/unselected as part
I see you really need a bidirectionnal mapping.
A dictionary won't help you in that case (uni-directional).
Hmmm ...
What about keeping the configuration in a .strings like format (old
style .plist), but when you load it, you create bidirectional access to
it.
So you load it as usual in a
It seem that you will have an web-server rule to mangle the url.
Whatever the rule, once you get in WO, and you hit the DA, set the
current language for the rest of the request, then in
performActionNamed(), you will first look in the localized .strings
file dictionary for that key (usual
Why not reverse the problem and use an alias mapping.
it would look like this :
{
// english alias
search = search;
find = search;
// french alias
recherche = search;
trouve = search;
}
then in you DA, you would overwrite :
public WOActionResults
I was hoping to handle the View entirely in .html + .wod, so if the
.css changes I just change the references in there. However, the .wod
logic won't cope with that, so I have to extend the View to include a
chunk of Java code. Still MVC as long as I separate it from the
Controller code.
You can add an entry in the userInfo dictionary of your entity,
defining the super class.
Then in the eogenerator template, you will be able to set the
superclass based on that key-value.
That way, you may not tell EOModeler the exact details of java
implementation (which can be all fine in
From a quick read, you have to replace the php/asp handling with WO
handling. A quick and easy way to do it is by using DirectAction, just
so you can easily write the url directly in the html file.
This will give you the bare essential just so it run.
This is not how you would like a WO
I have a somewhat similar setup.
I have a sub-component (call it Section), that have menu-tab and is
responsible to show the proper content based on the selected tab.
In section, I have an ivar which track which menu-tab is selected. I
have a method that return the content sub-component name,
This inverse the role-responsabilities of the component as proposed by
James.
This model work great when the wrapper is mostly static, or doesn't
interact directly with the actual content shown.
The other model proposed by James, is more appropriate (at least for
me) when your 'wrapper'
From what I see (and how I understand), this is not a vain intellectual
exercise ...
If you think of the content as a sub-view of it's parent wrapper : use
woswitchcomponent.
If you think of the wrapper as a decoration around the content : use
wrapper technique with wocomponentcontent.
Le 06-04-25, à 00:47, Anjo Krank a écrit :
You need to get a servlet.jar and add it to your
/Library/WebObjects/Extensions.
And you should be careful with that component. I haven't yet tested
it, but it seems to me that if you bind up your page as the proxy
object, you could call sth
14:33 schrieb Jean-François Veillette:
Le 06-04-25, à 00:47, Anjo Krank a écrit :
You need to get a servlet.jar and add it to your
/Library/WebObjects/Extensions.
And you should be careful with that component. I haven't yet tested
it, but it seems to me that if you bind up your page
, 2006, at 1:23 PM, Jean-François Veillette wrote:
Beware that the wonder implementation distribute changes snapshot and
tries to apply them to the database cache.
Unfortunately, under load, you will realise that this is causing a
deadlock.
I haven't got to the point where I was able to debug
Anyone experienced cougaar or jmemprof ? ( http://profiler.cougaar.org
and http://oss.metaparadigm.com/jmemprof )
They both offer a web interface, look at one example here :
http://oss.metaparadigm.com/jmemprof/examples/class-summary.html
I wonder if anyone got either product integrated in a
1 - 100 of 119 matches
Mail list logo