Examples of large, algorithmically complex applications

Does anyone have an example of a large, algorithmically complex Java Application (something on the order of 250K-1M lines of code)? I'm doing a language study to see how well mature Java applications scale over the life of long projects.
If anyone knows of any examples or could point me to someone who can, it would be appreciated.
Thanks.

Two ideas come to mind:
1) OpenCyc: http://www.opencyc.com/ A java based platform derived from an attempt to breed intelligence into machines.
2) ANY examples of neural net software you can find that are java based.
http://www.google.com/search?hl=en&ie=UTF-8&oe=UTF-8&q=java+neural+net

Similar Messages

Large and Complex Applicaton development process in Apex

Hi,
Apex can scale well in terms of Data size, number of users etc. We are trying to decide if Apex is suitable for our next large and complex application (Large and complex in terms of quantity and efforts of work and business logic). We have already identified that Apex can handle complexity of our transaction/business logic management. My question is about scaling application development process in Apex. I am curious to know about what other people are doing with Apex in terms of:
* What is the largest Number of Pages, Forms, Reports etc. within a single application one has developed
* Assuming with so many pages (objects) to be developed, more than one developers must have worked on the project and the project is divided into multiple phases release. What was/is the max number of developers in your project?
* Did you run into "Code Conflct" issues for Apex Shared Objects when multiple people had to work on them? How did you manage it and merge the changes?
* How did you identify, track and manage changes in Apex objects between various project phases and between various environments (DEV, TEST, UAT, PROD).
* What else you would like to add (in terms of development life cycle only) to my thoughts above that we need to consider while choosing Apex for our next project.
* Based on your experience would you recommend Apex for our large and complex application?
Many thanks in advance for your responses.
Paresh (yparesh at gmail dot com)

Hi Przemek, Jerry and Roel,
Thanks to all of you for your helpful responses. If we follow best practices mentioned like trying to centralize code in Page 0, db packages/procedures, JS libs shared components (especially "Shared App" having Shared components between multiple apps) we can minimize code management efforts and use existing technologies to identify, merge, promote code changes. This still leaves with some code and changes that will localized to individual pages (and we still need to manage page 0 and Shared object code changes. I will explain what I am trying to say with an example below about managing changes in a page.
Przemek, and Jerry,
How are you identifying, managing and promoting changes between applications environments say TEST to PROD. For example we have 2 separate dbs (and hence workspaces) to host Apex TEST and Apex PROD environment. If I have modified page 15 in TEST and now need to promote changes made to page 15 (deltas) to PROD how do I compare and identify the deltas? Apex can not migrate the whole page across apps in different workspaces and even if Apex can I don't think it is a good idea to overwrite page 15 in PROD with page 15 in TEST unless we can compare them, identify the deltas and make sure the changes are what we intended to promote. Exporting pages and comparing using a text editor is time consuming because of various ids and the way New Line characters can get embedded in page export. Logic that the TEST env should have identical copy of prod due to a refresh before promoting changes from DEV won't help because then whole of above will apply between DEV and TEST (albeit with little less risk as the target is TEST and hopefully any unintended changes will be caught during testing).
Apex 3.1 has Apex Applications Compare facility? If you have used it, what is your take on it? I think it is too basic but a good start. Won't it be good if there was a nice full fledged compare tool for Apex objects?
Thanks
Paresh

Large data-centric applications in flex?

Has anyone built a large data-centric application in flex? Are there code examples online?
I have read several articles about how to structure large applications in Flex but as yet have not seen any code examples of how this is achieved in practice.
Any help/links would be greatly appreciated, google hasn't helped me too much.
M

Tim, here is a bit of feedback about DCD
data management only supporting int as returntype for method createItem ( signature in the docs is createItem(item:myDatatype):int ) seems a bit limitative. We should be able to return the PK datatype or eventually the bean itself. When we are configuring data management the first thing we pick is the primary key so the wizard should be able to validate if the returntype for createItem match de PK. UpdateItem is also problematic, we are assuming that the bean is saved as is, but if we want to update some additional fields server-side, we have to implement some kind of push so the client is updated. The update signature should be updateItem((item: myDatatype):myDatatype or updateItem((item: myDatatype):*
Really don't like the way the services are generated inline and with the endpoint hardcoded. In the past, no endpoint was necessary, even less a full hardcoded one. Since we point it to the services-config, why is there a need to write it in the code? Removing it from the services generated we are still able to run it successfully.
Pagination should support a single count method for multiple paged queries as I explained in another thread. A ER has been made https://bugs.adobe.com/jira/browse/FB-20809.
Lazy-loaded related objects should be supported, the basic support lazy-loading just for the first level (object itself) is quite limitative because complex objects (with many associations ) will take a lot of time to load every single related object. For small projects it can ok, but not for medium / large ones.
The invocation process to service call should be based on events like the rest of the frameworks, flex included, like that we would be able to easily replace the generated code with our own implementation.
Forms generation is quite useful, just think it should be more advanced like enabling formatting and validation process.

Warning:The EXPORT data cluster is too large for the application buffer.

Hi Friends,
I am getting following warning messages whenever I click on costing in "Accounting" tab.
1. Costing data may not be up to date Display Help
2. No costing variant exists in controlling scenario CPR0001
3. An error occurred in Accounting (system ID3DEV310)
4. The EXPORT data cluster is too large for the application buffer.
I can create project automatically from cprojects. PLan costs, budget and actuals maintain in WBS elements can be visible in cproject Object Links.
Your reply is highly appreciated.
Regards,
Aryan

Hi;
Please, check the Note --> 1166365
We are facing the same problem in R3, but apply this no fix it.
Best regards.
Mariano

HR_INFOTYPE_OPERATION-- Complex Application Error

Hi:
I am using the FM "HR_INFOTYPE_OPERATION" for uploading data from flat file to IT0008 infotype.It's giving some error "A complex application error has occurred".I have checked SAP note 710779 but it does not help very much because we are already in the release and the correction is already there.Anybody with any idea is highly appriciated.Award points confirmed.
~Sinchan

try putting DIALOG_MODE = '2' and execute the program.. you cn check if the values get passed to the function module ocrrectly..
~Suresh

Most complex application

What is the most complex application that users have developed using HTMLDB? Is it available for preview (read-only) on htmldb.oracle.com?
I am having difficulty "selling" HTMLDB in our predominantly client-server/VB/.NET/ASP, etc shop. "This is a configurator, not a development environment, you cant develop really complex apps using this" "You dont even have access to the raw HTML to tweak it". "Our typical apps have loads of business logic in it, no way this toy can do it", "The (MS) tools we are currently using can do all of this stuff, why would I want to consider learning a new tool and switching", etc.
Thanks

to be derived from zero experience or knowledge of
HTML DB.Right, these were the comments I received after a 1-hour demo of HTMLDB to a group of MS-based developers who had never seen HTMLDB before.
I do not understand the comparison to a configurator,
that one is confusing. HTMLDB is a "configurator". You guys call it a "declarative programming environment". If anyone has seen Siebel Tools, that is also a configurator. In other words, a tool/environment which generates code for you and provides hooks to override/customize the generated code. The code in this case is HTML.
Their thought was that any point-n-click, wizard-based tool is for sissies. "Real men" use Visual Studio or the other MS offerings. HTMLDB was compared to MS Front Page or even MS Access to build simple data-manipulation forms.
I guess I would ask what is the definition of a "really > complex app"?An app with a lot of complex business logic, complex navigation, validations, high degree of client interaction (Javascript/DHTML with the browser paradigm), etc.
I also like the conclusion of "the MS tools...can do
all of this stuff". Do what? Write all of your
business logic for you automagically? Hardly. TheBy "all this stuff", they meant the application development environment that HTMLDB provides, the widgets, the branching, the validations, etc. That is not exactly ground-breaking, a number of development environments out there offer similar things.
Needless to say, I didnt agree. IMHO, HTMLDB is not a revolution, its a natural evolution for Oracle as a database company to provide tools to build apps using that data. But I can see how it does not offer anything so compelling as to persuade a 10-year MSCE veteran to consider switching
Thanks

Complex application dependencies

I have a somewhat complex application dependency tree.
I have four applications A, B, C, and D. Each application is deployed in its own EAR file. B depends on EJB's from application A. C also depends on A. D depends on A, B, and C. In addition, A depends on B and C. So it makes a diamond, with A at the top, B and C side-by-side on the middle level, and D at the bottom. Each application is deployed in its own EAR file.
The parent attrib in server.xml doesn't handle this at all. First, you can only specify one parent for each application, and I have one app (D) that depends on three others. Also, the parent relationship only flows one direction. I have three apps that have bidirectional dependencies (A<->B, A<->C).
Putting all the EJB's in a single EAR file would be very clunky. A couple of my apps already take serveral minutes to deploy, putting them all together would seriously dent development productivity.
Has anyone else had complex dependencies like this? If so, how did you deal with them?
Thanks,
Lance

On 19/7/09 16:04, Ralf Ebert wrote:
> I have to agree, setting up a run configuration and dependencies for
> SWTBot for a complex application is tedious.
>
> What do you think about reducing the runtime dependencies of SWTBot?
> Something like a single plug-in "org.eclipse.swtbot.runtime" containing
> all the classes required at runtime (maybe complemented by separate
> finder plug-ins)?
Could you elaborate more on what you mean by a plugin containing runtime
classes and how you propose to split them ?
Ketan
http://studios.thoughtworks.com/twist | http://twitter.com/ketanpkr

Need Example on Integartion of BEx application into EP

Hi
I need example on Integration of BEx application into EP and Display Data on EP from BW without using BEx web application.
Best Regards
Ravi Shankar

Hi Ravi,
have a look into the <a href="http://help.sap.com/saphelp_nw2004s/helpdata/en/88/3c1640d4642402e10000000a1550b0/frameset.htm">SAP Library</a>. There you will find descriptions of various options how to display BI content in the SAP NetWeaver Portal.
Best regards,
Martin

If a pdf file is to large or complex to convert ,what do i do ?

if a pdf file is to large or complex to convert ,what is my next step?@

Hello,
an interesting instruction provides exporting pdf file per Indesign(!), Please have a look at:
https://helpx.adobe.com/indesign/using/exporting-publishing-pdf.html >>> Reducing the size of PDFs
Hans-Günter

Help in profiling a complex application needed

Hi,
I have a rather complex, general, problem here I would like to solve using
dtrace so any hint, advise, similar experience etc would be more than
welcome.
Basically, my company has created a complex database system for acquisition
and management of reference stock market data and the process has several
phases (this is my understanding of it): acquire the data from
e.g. Bloomberg in their native format, transforms it in some universal
format and then use that format to load it into a database where it can be
analysed, processed etc using more or less (in house developed) standard
SQL with calendar extensions (e.g every Friday, last Monday in month etc
etc). There are several processes (servers) involved: data server,
scheduler, database server etc etc.
Now, when we have installed the product at one of our customers (running
E6500 with 24CPUs and 20GB of RAM, all that on Solaris 8 --I think), our
product was given a two hours window every day to do a certain amount of
work and to process several GBs of data (don't have the figures) and
everything was fine, since we were able to finish that in 90 minutes. But
now, as the amount of data within database grows (although still within
preallocated 80 or 100GB size -- split into 2GB files to work around (ex)
large file problem), for the same amount of work the product in question
needs something like 180 minutes -- which is 1 hour too much.
My task is to try to find out where the bottleneck is and to give some
suggestions for speeding things up. The customer claims that the
application is I/O bound and prstat -m more or less confirms that (almost
no CPU latency, large amount of voluntary context switches etc etc).
How do I confirm this using dtrace? I can't run dtrace on their production
system since it's Solaris 8 but I have set up a Solaris 10 (s10_60)
machine that has all the components installed and will try to run similar
kind of work there.
Where do I start? What shall I look for? The problem is also that there is
a lot of programs involved, a lot of IPC (shared memory, semaphores, pipes,
sockets) and in general when the system is busy, it is quite busy.
I guess in ideal world, we would call Sun and ask them to figure the
problem themselves, but it is not an ideal world and the golden age of .com
is gone :-(
Thanks and bye, Dragan

Where do I start? What shall I look for? The problem
is also that there is
a lot of programs involved, a lot of IPC (shared
memory, semaphores, pipes,
sockets) and in general when the system is busy, it is
quite busy.Well, this is definitely the hard part. DTrace makes it much easier to ask questions of the system, but you're asking "what questions do I ask?" There are lots of valid ways to go about this, but I often start by by looking at how the physical resources on the system are being used: CPU, I/O, networking and so on. For starters, squeeze all of the idle out of the system: as long as there is idle in the system, you're obviously not going as fast as you could. You want to start asking questions like "why am I doing I/O?" or "why am I coming off CPU?" If you could post some mpstat/vmstat/iostat data, perhaps we could provide some more concrete directions...
- Bryan
Bryan Cantrill, Solaris Kernel Development. http://blogs.sun.com/bmc

[JS CS5.5] Complex Application, want context that current script was called from

I'm trying to create a slightly sophisticated application that is spread across several script files. Some functions in these files call either other. Here is what I want to emulate:
A.jsx
if(mainScript=="A.jsx"){
doA();
function A(){}
B.jsx
#include A.jsx
if(mainScript=="B.jsx"){
doB();
function B(){
// some stuff
A();
//some more stuff
C.jsx
#include A.jsx
#include B.jsx
#include D.jsx
main(){
D();
A();
etc...
Basically, I want a script to act differently depending on if it is the one being called or if it is being called from elsewhere. I thought I could do this pretty easily by using global variables and something like:
if(mainScript==undefined){
var mainScript=$.fileName;
But global variables persist between script calls so this doesn't work. I'm a bit stuck. Anyone know how to do this? app.activeScript doesn't work because I am executing through the extendScript toolkit.
I'm vexed.

You're definitely not wrong John, this will resolve the problem. Sometimes, especially when doing development in a large project with lots of interlocking pieces it's convenient to be able to have some test code within each file that tests the functionality of that specific code. Even moreso when you are new to a language, so when you think you find a better way to do something you can quicky try it out and see what happens without invoking your monolithic program. It's not difficult to achieve this approach by splitting into two files, it just means that you have to keep twice as many file tabs open and remember which one to switch to "Go" instead of doing it with the code you are working on. It seems like an extra step for no good reason.
I do this in Java all the time, most of my complex classes have a main() method so they can be executed standalone to try out a quick snippet or test some corner cases. Comprehensive unit testing would probably be better but the time investment is much greater plus I doubt something like that would be well supported with ExtendScript.
Is there any different between a jsx and jsxinc file or is that just a common naming convention that you or a community of people use to differentiate files?

Large (1gb+) iphone application development

I'm looking for experience in developing iphone applications with large data sets. I have a sqlite database that is 1gb+ and would like to include it with my application. Unfortunately, I cannot even get Xcode to run it on the simulator unless I reduce the database size substantially.
I'm wondering if there is a timeout setting somewhere that is getting triggered and killing the upload process?
When I try to run the large application on my device, I get the following error message:
"Operation could not be completed. (com.apple.dtdevicekit error 49164.)
I've thought of other alternatives, such as downloading the database inside the application, but am unclear if this is a desired workaround.
Anyone have any experience with developing large applications?

I've never built an app that was 1% of that size, but hope my opinion might be helpful to you anyway.
As you may know, the hard limit for an iPhone app bundle is 2 Gb. But just from basic engineering experience I'd say 1Gb+ is asking for trouble. In my opinion, the day when you succeed in getting that bundle loaded and tested is when your real problems will begin.
Is there any way to modularize that database? For example, could you put a small, starter database in the bundle and then let the user download additional modules from your website? I think such a design would not only be technically superior, it would also be better commercially.
I've read that most users don't look at the size of an app before downloading. I've never met that kind of user. When my kids download music, they count every Mb, and know exactly how much room they have left. If I saw Gb in an app store description I'd have to be very very motivated to read the rest of the page.
On the other hand, if the app was under 10 Mb, and the description explained that the app could be expanded one package at a time... Well, I'm sure you get the point. Of course making an extensible database out of SQLlite is going to be a nice project.. It's not called Lite for nothing. But that would be my goal.
Hope that helps!
\- Ray

Deploying a large Oracle BPM Application with multiple UI projects

Hi everyone
I'm working on a large application with Oracle BPM 11.1.1.5.0.
I'm using JDeveoper 11.1.1.5.0 to build the application.
The application has lots of user tasks with lots of UI projects (one for each user task).
My problem is, each time I make a small change in the project, I have to deploy the whole thing, which includes one main project and five or six UI projects (for now) for the user tasks.
As the application gets bigger I'm sure this will definitely become an issue.
So my question is if I'm doing this the right way? or should there be a more efficient way?
Thanks in advance
Edited by: luke on Apr 17, 2012 12:16 AM

Hi Luke
1. Generally, there is NO need to have One UI Project for one individual Task. Say you have like 10 Tasks (.task files) like Initiator, Approver, Reviewer, LegalApprover etc etc. You can have just One UI TaskForms Project and have taskDetails.jspx generated for each one of these Tasks. This is the preferred way.
2. Any Workflow Application should have just 2 deployers. Means one deploy is for actual Workflow stuff that will have the main BPEL or BPM Process, human Tasks etc etc. Basically all that goes into workflow project. And another deployer is the REST of the things means the TaskForms UI Project and any other supporting projects like EJB projects, utils projects, any WebServices projects (if you have WebServices). All these will be deployed as one SINGLE EAR.
Lets take an Example from JDeveloper IDE point of view.
1. Application Name: SalesOrderApp
2. Workflow Project Name: SalesProcess (This is a a jdeveloper project of type BPM Project or SOA Project and has all human tasks, bpels, bpm process, business rules etc etc).
3. UI Project: SalesTaskForms (This is a auto generated form for the very first Task that you choose. Then for rest of the Tasks use this same Project, to generate jspx for all other .TASK files. REFER online docs on how to do this. I will see if I can compile this list of urls for you).
3. Supporting EJB Projects: MyEJBProject1 (Some EJB Project that connects to some backend database specific to your project)
4. Supporting another EJB Service Project: MyExtServiceProject1 (Like another EJB Project, that pulls data from some external source)
5. Some WebService Project: SalesCreditCardValidationWebService (Some WebService Project to be deployed as .WAR file that does some service).
Try to apply SOA Concepts for each Service that your application needs. Instead of putting all the EJBs and everying in one single giant project, try to split them logically/business need wise and geenrate separate projects so that they can service as it is to other projects and other apps.
Anyhow, coming back, in above Example, you should have jsut 2 Deployement: 1 is Workflow SCA JAR File and other is full EAR file.
a) For the Workflow Project, right click and generate a Deployment profile to deploy just this workflow and NOT any other UI Projects.
b) For each individual project genereate a deployment profile like MyEJBProj1, MyEJBProj2, MyWebService1 etc etc.
c) For TaskForms project also, generate a deployment profile like SalesTaskForms. This will be a WAR file.
d) At Application Level, create a new Deployment profile of type EAR and now assemble all the projects under this EAR except for the Workflow project.
First you deploy Workflow SCA JAR file. You need to redeploy this again only when you make changes to workflow, humantasks etc etc.
Deploy full blown EAR file that has all the stuff (EJB JARs, TaskForm WARs, WebService WARs etc).
Anytime you change taskforms, or any java or ejb code, ReDeploy EAR file.
You should have one JAR and one EAR to deploy at the end of the day. Infact this is how it should go into UAT and Production. They should not have multiple JARs, multiple WARs to deploy. The reason we have separate Workflow SCA Jar file is, bcos workflow will not change frequenly. And also older versions of SOA/BPM (before 11.5 FP) have a huge disadvantage. Everytime we deploy workflow project, it will make all existing process STALE unless you deploy with a diff version. This is not an issue any more for 11.5 + FP (Feature Pack applied). Hence be careful when you redeploy workflow project.
Thanks
Ravi Jegga

Complex application

Hi
I am trying to build an application. The idea is:
1. connect a mobile to a router
2. Let the mobile search and display the devices that attached to the router (getting the device name and IP address)
3. The user clicks on one of the devices to make a connection with it
I tried to use HttpConnection to access the router and get the DHCP Table but it returns error because the got username and password. My router is Netgear DG834G.
I know this sounds a bit complex but any help will be appreciated.
regards

In my test workspace I created Project1 and Project1 projects.
In Project1, I created a package of BC4J components named package1
I chose "Deploy Business Components..." on Project1 and selected a Simple Archive. Upon deployment, this created me two JAR files Project1CSCommon.jar and Project1CSMT.jar.
I created a library for Project1CSMT.jar and added this library to my Project2 project's library list.
In Project2, I created another BC4J package of components named package2
I right-mouse "Import Business Components..." and I navigate to the Project1CSMT.jar and use the file dialog to "drill down" into the jar file just like a folder, and pick the package1.xml file. This results in importing package1 into Project2.
I chose "Deploy Business Components..." in Project2 and select an AppModule Session Bean (BMT).
I expand the Project2EJB.bcdeploy node in the navigator and right-mouse on the Project2EJB.deploy node to pick "Settings...".
I click on the "Profile Dependencies" tab, and check [x] Project1MiddleTier.deploy and [x] Project1Common.deploy to make those two deployment profiles a dependency of this one.
Then I right-mouse "Deploy" on the Project2EJB.bcdeploy node and it deploys my AppModule in Project2 as a remoteable application module. I go out to the file system and do:
% jar tvf Project2EJB.ear
329 Thu Jun 20 18:31:10 CEST 2002 META-INF/application.xml
1665 Thu Jun 20 18:31:10 CEST 2002 META-INF/data-sources.xml
286 Thu Jun 20 18:31:10 CEST 2002 META-INF/orion-application.xml
4361 Thu Jun 20 18:31:10 CEST 2002 Project1CSCommon.jar
13344 Thu Jun 20 18:31:10 CEST 2002 Project1CSMT.jar
25345 Thu Jun 20 18:31:10 CEST 2002 Project2EJB.jar
9185 Thu Jun 20 18:31:10 CEST 2002 Project2EjbCommon.jarto verify that it includes my imported JAR files into the EAR file.

Need Examples of Large Scale APEX

Hi,
I have a new project deal I'm working on and my potential customer (and I) need more information on the salability of APEX; up to 10,000 simultaneous users. Is this being done anywhere? What's the best architecture? Any Example web sites?
Thanks,
Steve
Yahoo= smanningc3

I know about a non-public APEX application the German T-Punkt shops are using.
Dietmar Aust (http://daust.blogspot.com/) was talking about it at the last DOAG APEX SIG workshop. They have 5.000 concurrent users using it every day. It's running on a Oracle RAC.
Patrick
My APEX Blog: http://inside-apex.blogspot.com
The ApexLib Framework: http://apexlib.sourceforge.net
The APEX Builder Plugin: http://sourceforge.net/projects/apexplugin/

Examples of large, algorithmically complex applications

Similar Messages

Maybe you are looking for