IDE Survey by Server Side

Please go to:
http://go.techtarget.com/r/1050598/2794892

and take this survey.

Wade

==================
Wade Chandler
Software Engineer and Developer

Netbeans Community and Dream Team Member:
http://wiki.netbeans.org/wiki/view/NetBeansDreamTeam

Check out Netbeans at:
http://www.netbeans.org

Re: VWP and Netbeans 5.5.1

Re: too many plugins, too many places?

> --- Coolboykl <khoo.james <at> gmail.com> wrote:
> >
> > Hi Jirka
> >
> > This is good news!
> > I would like to suggest the following for the
> > upcoming portal.
> > 1. Although we like to encourage developer to submit
> > more plugins to portal,
> > at the same time we don't wants too many similar
> > plugins exists, which will
> > confuse new Netbean user. or worse  "inherit"
> > eclipse famous plugins
> > incompatible problems. I would suggest for each new
> > plugins
> > a) The portal needs to verify  each submitted
> > plugins,
> > b) Ensure each new plugins pass certain test suites.
> > c) Put the plugin under review folder, till selected
> > communities member
> > agree to publish the plugin.
> >
> Again, you get into limiting contributors.  Are you a
> module developer and do you give your code away for
> free?  I am, and I would take offense.  My feelings
> are this is not much different than the way contrib
> works now.  Too much validation == not many plugins
> will make it out.  Some will be fine plugins which
> just work.  Users should use modules based on feedback
> or if they are into cutting edge try them out.  A
> rating system can help those who worry about things.
> Limiting people just limits the choices.  This is how
> Mozilla works...many plugins perform similar tasks,
> yet they are able to be included for people to try
> out.  Different developers have different coding and
> usability styles.

> > 2. Use Java.net to host ongoing netbeans plugins
> > projects, and provide link
> > to access the ongoing plugin development projects.
> > Developer could
> > collaborative participate any of those project.
> >
> Each developer needs to be able to make their own
> choice about hosting.  I don't believe modules for
> free use should be forced to be open-sourced either.
> Module developers shouldn't be limited in what they
> want to give to the community.  Doing so doesn't just
> limit them, but it limits users.  Not to mention the
> fact you are getting something for free when someone
> contributes their module.  Again, limitations are not
> a great thing.  This has been part of the
> problem....too many modules is like too much
> money...just doesn't happen....now too few
> modules....who wins?  The more open and loose this
> process is the better.  You're going to have some bad
> modules and good modules, but bad module developers
> can become good module developers if given
> opportunities, but if we run them off before they
> become good we'll never see what potential they have.

> Wade

I agree on the point that "Too much validation == not many plugins".
But, I also think some kind of validation system is necessary.
We may have a recommendation system which is strict enough to be
authoritative.
The recommendation system should be independent of rating system,
For example, recommended plugins are not always those high rated ones.
The recommendation system may be hold by a group of experts, the system
can do:
1. Give a passing line to these new born plugins.
2. Give some advices to the developer, For example, two plugins
which have the similar features can be merge together.
3. Lead or adjust the development direction.

Tetsu.

MIME WYSYWIG editor for Swing?

Hi there,
do you know any free and easy to embed editor for emails? In my
application, there is template service that produces ready-to-send
(multipart) MIME messages, but before I let my service send it, I have
to give application user a chance to see and/or modify that email
(before accepting).

I was trying to embed editor used in Colomba Java Email Client (looks
almost like Thunderbird), but after few hours of digging in code (it
is  1100 java files) I gave up. It was too hard for me to tear it off
that application :(

Do you know about something else I could use? I would like it to
accept either raw MIME message or javax.mail.internet.MimeMessage...
or whatever else I can possibly do with generated email content.

Thanks,
Witek

Re: OC4J and VisualWeb

Hi Michal,

Thank you very much, that definitely put me on the right path (although now
I'm getting a connection error when I deploy talking about "'Example
Connection Pool_default'", but that's a different problem altogether).

For anyone curious, here's how I got the visual editor to work:
*I downloaded the ADF Faces stand alone from Oracle
(adf-faces-10_1_3_0_4.zip)
*I then looked inside their example WAR (adf-faces-demo.war) and traversed
to their WEB-INF/lib directory. 
In there you'll find all the libraries you'll need (including the JSF base
implementation that the installation instructions for the ADF faces
requests, and the jstl libraries that Michal talked about).  
*Copy all those libraries to your projects WEB-INF/lib folder
(adf-faces-api-SNAPSHOT.jar, adf-faces-impl-SNAPSHOT.jar,
adfshare-3549S.jar, commons-beanutils-1.6.jar, commons-collections-2.1.jar,
commons-digester-1.5.jar, commons-logging-1.0.3.jar, jsf-api-1.1_01.jar,
jsf-impl-1.1_01.jar, jstl.jar)
*Close all your open JSP files (if I didn't do this for some reason I kept
gettting errors).
*Do a clean and build
*Voila should work

-jeff

Michal Mocnak wrote:
> 
> Hi Jeff,
> 
> just for clarification. OC4j 10g is java ee 5 compliant (supports ejb3, 
> jpa and so on) and there is no problem with these projects.
> 
> Actually your problem with Visual Web Project (with JSF) is simple. OC4J 
> doesn't contain JSF and JSTL libraries and due to you don't have these 
> libraries on classpath -> there is a few ways : include these libraries 
> into project and package them or include all these libraries on a server 
> classapth -> <ORACLE_HOME>/j2ee/home/applib
> 
> Thanks
> 
> -mcihal
> 
> jbethune wrote:
>> Hi everyone,
>>
>> I'm new to the Application Server model of programming, and for the first
>> time decided to give the Visual Web and Enterprise Packs a go. 
>> Everything
>> is pretty straightforward using the "Sun Java System Application Server"
>> as
>> the J2EE server, but unfortunately I need to deploy to an Oracle 10g
>> server. 
>> So I picked up the OC4J module, and attempted to create a Visual Web
>> project
>> based on that.  This doesn't work giving me the following exception:
>>
>> Page1.java:16:19: package javax.faces does not exist
>> Page1.java:37:26: cannot access javax.faces.component.UIComponentBase
>> file javax\faces\component\UIComponentBase.class not found
>> etc...
>>
>> Now is this because the OC4J doesn't fully support Java EE 5, and instead
>> supports J2EE 1.4 with some stuff from 5 thrown in?  If so, is there
>> anyway
>> around this?  I'd really like to use the VisualWeb editor and not use
>> Oracles Development suite, as all our coding for the past 3 years has
>> been
>> using Netbeans.
>>
>> The interesting thing, is the "New Visual Web Application" Wizard
>> autodetects the Java EE Version to be 5 even though technically it's not
>> (I
>> think).  
>>
>> Thanks, any help would be much appreciated.
>>
>> -Jeff
>>
>>   
> 
> 

--

-- 
View this message in context: http://www.nabble.com/OC4J-and-VisualWeb-tf3307071.html#a9242753
Sent from the Netbeans - Users mailing list archive at Nabble.com.

Re: help - ojdbc14.jar only works partially in my NB DB visualizer ???

Thanks for letting me know this.  Me, as a dumb end user of their JDBC 
drivers simply went by the Oracle site's recommended driver for the DB 
we have: 9.2.0.5 - but that one didn't work?!  The ojdbc14.jar that 
worked wasn't even recommended by Oracle for use with 9.2.05 (it is for 
10g rel2.)
Ah well....doesn't matter, I suppose - as long as it works.

Thnx again,
tom

Jim Davidson wrote:
> Glad to hear that it worked for you.
>
> Oracle's drivers are generally backward-compatible -- that is, they 
> will work with previous versions of the DBMS.
>
> Their support for JDBC3 has been improving over time, so it's not 
> surprising that you noticed a difference when you upgrade to the 
> newest one.
>
> -Jim
>
> On 2/27/07 9:06 AM, Thomas Wolf wrote:
>> Michel - thanks a bunch.  That was it - once I upgraded to ojdb14.jar 
>> for 10g, it worked.  Hm, I was too good a citizen - i was using the 
>> ojdbc14.jar that Oracle tells me to use for the 9.2.0.5 DB we're using.
>>
>> again, thanks.
>> tom
>>
>> Thomas Wolf wrote:
>>> Hi Michel,
>>> As far as I know, I have the latest ojdbc14.jar from Oracle for our 
>>> version of the DB 9.2.0.5.  But I will make sure of that.  Thanks.
>>> tom
>>>
>>>
>>> Michel Graciano wrote:
>>>> You need driver for jdbc 4. Maybe geting diver from oracle 10 can 
>>>> solve your problem. This, if I am not wrong, happen since 5.5 RC.
>>>>
>>>> Regards.
>>>>
>>>> On 2/26/07, * Thomas Wolf* <twolf <at> netforensics.com 
>>>> <mailto:twolf <at> netforensics.com>> wrote:
>>>>
>>>>     I've been having this weird problem for several months now (maybe
>>>>     since
>>>>     going to 6.0?)  I can't see the tables in my DB!!!
>>>>
>>>>     I am going to Runtime->Databases->Drivers and add my odbc14.jar
>>>>     driver.
>>>>     After that, I establish a connection as I usually do:
>>>>                     jdbc:oracle:thin:@neptune.netforensics.com:1521
>>>>     <http://jdbc:oracle:thin: <at> neptune.netforensics.com:1521>:nfdb
>>>>     And NB seems content with everything:  I see a DB connection tree
>>>>     that
>>>>     expands to have "Tables", "Views", and "Procedures".  If I expand
>>>>     "Views" or "Procedures", I indeed see the views and stored
>>>>     procedures on
>>>>     my database server.  But when I select "Tables", I get the error:
>>>>                     Unable to read the database structure, connection
>>>>     error;
>>>>     Unsupported feature
>>>>
>>>>     Can anyone help?  Colleagues using DBVisualizer with the same
>>>>     ojdbc14.jar have no problem accessing the tables in the DB.
>>>>     thnx,
>>>>     Tom
>>>>
>>>>     p.s.
>>>>
>>>>     --
>>>>     Thomas Wolf
>>>>     twolf <at> netforensics.com <mailto:twolf <at> netforensics.com>
>>>>     732-393-6074
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> -- 
>>>> Atenciosamente,
>>>>
>>>> Michel Graciano
>>>> KSI Soluções em Informática Ltda.
>>>> http://www.jroller.com/page/hmichel
>>>> http://translatedfiles.netbeans.org/index_pt_BR.html 
>>>> <http://translatedfiles.netbeans.org/index_pt_BR.html>
>>>> https://copypastehistory.dev.java.net/
>>>> https://jae.dev.java.net/ 
>>>
>>
>

--

-- 
Thomas Wolf
twolf <at> netforensics.com
732-393-6074

Re: Re: Producing classes compatible with JRE 1.1

"Wade Chandler" <hwadechandler-nb <at> yahoo.com> wrote in message 
news:20070228165155.70689.qmail <at> web33805.mail.mud.yahoo.com...
> Slow UI:
> The example for a UI your provided isn't very common
> place.  A UI while not static usually isn't running
> through and creating a bunch of buttons (or tiny
> canvases) over and over again.  I have see many
> calendars which do basically what your example applet
> was doing and they have good speed and work fine.  If
> you need graphics and drawing use Java2D and draw
> instead of trying to produce many components.  See
> some of the common graphing components.  No speed
> issues.  When a lot of screen redraws take place if
> not all over the screen then you can just redraw
> pieces of the screen.  If you are doing a lot on user
> events then move code out into a separate thread.  A
> lot of times the perceived UI issues come from the
> threading model.  I'm not sure, but maybe for what
> ever it is  you are specifically doing maybe the MS
> JVM uses a different threaded model so what is flow
> related issues seems like performance issues to you.
> There have certainly been UI performance improvements
> in the Sun JVMs...there is no question especially
> related to graphics redrawing.

It is true that there has been some improvement in GUI functioning since the 
early Sun JVMs, but Microsoft's old VM is an order of magnitude faster on 
GUI operations than any of Sun's JVMs.  The applet at 
www.segal.org/java/CanvasTable3/ was meant as an extreme and reductionistic 
example, but it illustrates a problem that occurs in many Java programs, 
except for simple programs.

> Serious crash bugs until 7:
> I really have not seen this at all.  I have been using
> Netbeans IDE for many years, and it doesn't crash
> often.  The only times I have seen crashes it has had
> to do with bugs in graphics drivers and MS direct
> draw.  Once driver updates were installed or direct
> draw turned off I didn't see issues.  I even created
> infrastructure to tie 1.3, 1.4, and 1.5 JVMs into the
> Palm Pilot hotsync capabilities and used the JVMs for
> user interfaces which I tied to backend web services
> and local Java logic...I never saw crashes which were
> not related to programming errors, and these
> applications were handling hundreds of thousands of
> data records.  Which specific crashes are your
> referring?  I use Swing for my UIs.

In 1.5: http://www.segal.org/java/sun_jit/index.html, fixed in 1.6
In 1.6: http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6368342, fixed in 
one of the early 1.7 builds.
To its credit, Sun took these bugs seriously, with involvement at the 
highest levels of management, and they are now fixed.

> JAR file compactness:
> I haven't had issues with this really.  Not enough for
> there to be noticeable issues.  I think using a
> standard and simple format as PKZIP has been better
> than a proprietary format for not that much gain.
> They could have used something like bzip2 (much better
> compression than anything) or even gzip and achieved
> better results, but something tells me the trade off
> with PKZIP is partially a speed issue when dealing
> with decompression.

Our large applet and database is 532 KB as a signed Microsoft CAB file and 
697 KB as a signed Sun JAR.  Both contain the same large database file and 
many class files, presumably most of the difference is in handling the many 
small class files, and if about 375 KB of each is the compressed database, 
the class files are about 160 KB with the MS CAB and 320 KB with the Sun 
JAR.  That is a big difference, but it no longer matters now that few people 
use dial-up. 

Re: Re: Re: Producing classes compatible with JRE 1.1

Mickey Segal wrote:
> It is true that there has been some improvement in GUI functioning since the 
> early Sun JVMs, but Microsoft's old VM is an order of magnitude faster on 
> GUI operations than any of Sun's JVMs.  The applet at 
> www.segal.org/java/CanvasTable3/ was meant as an extreme and reductionistic 
> example, but it illustrates a problem that occurs in many Java programs, 
> except for simple programs.

It illustrates that heavy weight components are very expensive to render because 
of the interfaces between Java and the native environment.  If you change this 
applet to be a JApplet, change the Canvas to a JPanel, add a setLayout(new 
FlowLayout()) to the JApplet init(), and replace LabelCanvas with JLabel, you'll 
see that the refresh creates minimal flicker on the display.

Native components are not where Java excels and they are not what Sun has been 
paying attention to.  Swing, when used on reasonable graphics hardware, will 
provide perfectly acceptable rendering in most cases.  In JDK 1.6 things got 
faster still with the rework of the rendering stack for Windows to use the 
underlying systems correctly.

It might seem like a good thing to just send them an applet and hope that they 
have a reasonable JVM.  That's really not to your advantage.  Using some of the 
techniques already suggested through JNLP and otherwise, you can "demand" that 
they have an appropriate JVM.

If no one uses dialup any longer and a 340kb vs 500kb download doesn't matter, 
then I'd say that it probably doesn't matter that you're application requires 
them to install a JVM that supports Swing.  Doing that will make you life easier 
in the long run because you'll take advantage of working within classes and 
toolkits that everyone else is testing and exercising, and you'll get bug fixes 
sooner most likely.

Gregg Wonderly

Re: Producing classes compatible with JRE 1.1

"Gregg Wonderly" <gregg <at> cytetech.com> wrote in message 
news:45E65D22.4080705 <at> cytetech.com...
> If no one uses dialup any longer and a 340kb vs 500kb download doesn't 
> matter, then I'd say that it probably doesn't matter that you're 
> application requires them to install a JVM that supports Swing.  Doing 
> that will make you life easier in the long run because you'll take 
> advantage of working within classes and toolkits that everyone else is 
> testing and exercising, and you'll get bug fixes sooner most likely.

Our users, though highly valued knowldge workers, have neither the freedom 
nor the knowledge to change institutional policies about JVMs, and we are 
stuck supporting Java 1.1 for the next few years.  Once we can end Java 1.1 
support we will be able to make various changes.  In the meantime, hardware 
is getting fast enough to make the problems of the Sun JVM insignificant. 

Re: How do I do the following: