Best Practice Directory Structure

I'm new to programming and I'm creating a large application
that will have a lot of pages and components. As I continue to add
to the project, I find it harder and harder to easily locate the
pages and components that I need to work with. To help me with this
problem, I've decided to add a label to every component with the
location of where it can be found. Now when I view the application
and I see a component I want to work with, I know where it is
located.
This is problably not the best way to handle the problem, it
will be a pain to go back and remove or comment out these when I'm
fininshed with the project, but it wil help now.
Short of having a better memory :>), is there any best
practice guidelines I should look into and follow when I create my
directory structure of files and components that will intuitively
help me quickly locate a component? What do you use to help you?
Thanks for your ideas and suggestions

There are many ways to answer this. If you have a big project
you might consider using a Flex Library Project for your components
- think of it as writing a set of components you (or your company)
might one day sell to others - even if you never have that intent.
The idea is to make the components as reusable as possible in case
you need them for the next project. Having them in a separate Flex
Library project (which creates a .swc file) would make that easier.
The industry best-practice seems to be to create packages
that begin with your company's domain name, but in reverse order.
For example, components I write for Adobe go into a package
beginning with com.adobe which works out to be the file structure
com/adobe and then is further divided by the application and then
its parts (eg, com.adobe.scrapbook.editor). Using your company's
domain to distinguish your components enables you and others to
combine components from different places without naming conflicts.
If you divide your application into logical parts you can
figure out a good package naming convention that is easier to
remember. For example, I might put all of the skins for my
Scrapbook Editor into the com.adobe.scrapbook.editor.skins package.
I can easily find those files and add to them as necessary.
Other people follow a similar pattern and there are books on
the subject, too.

Similar Messages

Best practices code structure for large projects?

Hi, I come from the Java world where organizing your code is handled conveniently through packages. Is there an equivalent in XCode/Objective C? I'd rather not lump all my observers, entities, controllers, etc in one place under "Classes"...or maybe it doesn't matter...
If anyone could point me to a document outlining recommended guidelines I'd appreciate it.
Thanks! Jon

If you have a small project, you can setup Groups in Xcode to logically organize your files. Those Groups do not necessarily have to correspond to any directory structure. I have all my source files in one directory but organize them into Groups in Xcode.
If you have a larger project, you can do the same thing, but with code organized into actual directories. Groups can be defined to be relative to a particular directory.
If really do have a large project, you should organize things the same was as in Java. Your "packages" would just be libraries - either static or dynamic.
As far as official guidelines go, there really aren't any. It would be best to stick to the Cocoa Model-View-Controller architecture if that is the type of application you are working on. For other software, you can do it however you want, including following something like Sun's guidelines if you want.

Best practices in structuring a large Flash project?

I'm building an educational site in Flash. A student works through a series of activities, either watching an animated video or answering a question. You can get a good idea of the basic functionality with this mockup: http://imgur.com/Mi4JyHN.
After a student logs in, the server responds with their current activity and progress. The activity displays, and all student interaction is sent to the server to record (time spent on questions, buttons clicked, correct and incorrect answers, etc). When the student is done with an activity, the server is notified and responds with the next activity to present.
I'm fairly new to flash and would love to hear how people with experience would structure this project. I can think of the following possibilities.
One enormous SWF. All audio files and movie clips are embedded in the SWF and swapped out as necessary. This is not a reasonable option because the size of the resulting SWF would be huge.
Exactly one SWF for each activity. The control buttons and progress bars are obviously shared between each activity, and it seems like a lot of duplication to have them in each compiled SWF. Also, if a button is changed, this requires re-compiling everything, right?
One main SWF that loads others. The main SWF contains the buttons and progress bars and fetches external SWFs from the server to replace the stage area. I don’t know enough Flash to predict how this will go.
Part JavaScript, part Flash. The buttons and progress bars are done in HTML + JavaScript. The page fetches external SWFs from the server to replace the stage area. This is the current system, and the problem is the ugliness/difficulty of managing the communication between JS and AS3.
HTML5. While I would love for this to be a possibility, I don’t feel like it is. Animation is still way easier in Flash and we are still targeting some fairly old browsers. The best part about Flash is the consistency in experience.
Extra questions:
Which options leave us open to publishing to mobile using Adobe AIR?
Which options are best for automated testing / accessibility / version control / general code layout?
Thanks so much for any advice!

For a Flash-based design I would go with option 3. The general controls and objects common for use with each activity would be in the main file. Whether or not the main file would be responsible for sending activity data to the server could depend on there being determinable similarities between data collections... otherwise it might fall on individual activity swfs to interact with the database. There are too many unknowns in this regard for me to offer much.
If this was going to be a loaded application, such as an AIR app, then option 1 might be more reasonable since you only have to install once for everything to be available.
When it comes to mobile you are likely to hit a snag if you rely on using AIR/Flash to try to deal with a main and activity swfs approach... mainly in the Apple realm... unless I forgot a lot of what I was involved with some time ago, a loaded swf cannot contain any code when it comes to iStuff. So you end up having to make the main file contain all of the coding to deal with each activity's processing. Every interface/interactive element can only exist by name and the main file has to target them and assign listeners and processing and etc....a mess in my view. That's why having the one huge AIR file is possibly a tad more reasonable.
I have nothing to offer in the HTML5 end of things. I have not yet journeyed down that path. Since HTML5 is basically wingless without javascript and CSS, it might come to pass that the current system (option 4) is the way to go.

Best practices for package directory structure?

I've only compiled a few java programs & while I find the language itself very easy to adapt to from C++ and Javascript, I get confused about the directory structure and classpaths & such.
My main method of operation is to build small tools and test projects, and I would like to follow good practices for organizing my directory structure, so it will (a) work well with java, and (b) work well with my source control software.
Let's say I am developing packages named "com.example.test.test1", "com.example.test.test2", "com.example.tools.flapper", and "com.example.tools.spinner".
I know I can use this structure:
/java/test/test1/com/example/test/test1/* (files for com.example.test.test1 including test1.java)
/java/test/test2/com/example/test/test2/* (files for com.example.test.test2 including test2.java)
/java/tools/flapper/com/example/tools/flapper/* (files for com.example.tools.flapper including flapper.java)
/java/tools/spinner/com/example/tools/spinner/* (files for com.example.tools.spinner including spinner.java)
But the directories seem unnecessarily deep & lead to a hassle in source control that makes it difficult to browse.
Can I use this? and if so, how do I run "javac" and "jar" properly to compile/jar-ify each package?
/java/com/example/test/test1/* (files for com.example.test.test1)
/java/com/example/test/test2/* (files for com.example.test.test2)
/java/com/example/tools/flapper/* (files for com.example.tools.flapper)
/java/com/example/tools/spinner/* (files for com.example.tools.spinner)
also is it recommended to leave the .class files in the same dirs as the .java files, or better to have a separate tree for compiled .class files?

hmm, there's a lot to absorb here...
I found these articles online that seem good:
http://www.onjava.com/pub/a/onjava/2003/12/17/ant_bestpractices.html
http://www.kevinboone.com/classpath.html
I'd like to use ant if possible & understand it; I've had troubles with IDEs in the past and would like to just stick with basic build mechanisms.
I guess as a practical example let's say I have two different applications that use the serial port, that I want to release separately, each in separate .jar files. One is called com.example.applications.serial1 and the other is com.example.applications.serial2. They both use rxtx.org's gnu.io.* libraries. Serial1 uses some packages that I have developed: com.example.tools.spinner and com.example.tools.flapper, whereas Serial2 uses com.example.tools.spinner and com.example.tools.flopper
I think I want to organize things this way: (everything below under /java/src/)
com
example
applications
serial1
serial1.java
serial2
serial2.java
tools
spinner
spinner.java
flapper
flapper.java
flopper
flopper.java
and I think I want to compile to /java/classes/:
com
example
applications
serial1
serial1.class
serial2
serial2.class
tools
spinner
spinner.class
flapper
flapper.class
flopper
flopper.class
but I'm sort of lost on where to put the ant buildfiles, which directories to be in to run javac, and how to get the jar files right. From what I've searched for online, the external libraries (RXTX's jar files & .dlls for windows) are "interesting", and the easiest thing is to tell users of my .jar files to put the rxtx libs in the same library as the .jar files so I can set the Class-Path: in the .jar manifest to point to them.

Best practices for folder structure in SOA Project

Actually, In my project, I have more than 10 BPEL process and several DBAdapters and some HumanTasks. For each BPEL process, JDeveloper creates .bpel, .componentType, wsdl files. For DBAdapters JDeveloper creates lot of files, and same for HumanTasks as well.
By default Jdeveloper puts all these files in root directory of the Project. It looks messy having 100's of files in root directory.
How can I organize all these files?
What are the best practices for folder structure in SOA Project ?
Thanks

Yes Yatan. I did noticed that Polling service WSDL disabled in EM . But I want to try this approach beacause I want to move all database related stuff to one composite. I have another composite with 3 BPEL Processes. One BPEL queries external webservice and save the response in DB. One BPEL just fires the first Process(Query) and respond to caller with an unique id. I have Another BPEL which keeps polling on database for 5 sec(I'm using PICK Activity), if there is any new record in database it will retrun to caller. If no new records in database, returns empty to the caller after 5 sec.
If I put all these in one composite, having lot of files in root directory, the composite really looks very messy.
As you said, I want to have 2 composites. First Compoiste will have 3 BPEL processes, and 2nd Composite will have DB Polling and DB Save. I will have another project for MDS to store XSDs.
So because I can't have Polling service as exposed service, I have to keep Polling service in the same Composite where the BPEL is. Now again it goes back to Big and messy project.
Is there any way I can separate all DB related including polling in a separate Composite ?
Thanks
--Sreeny

Best Practice for Software Update Structure?

Is there a best practice guide for Software Update Structure? Thanks. I would like to keep this neat and organized. I would also like to have a test folder for updates with test group. Thanks.

Hi,
Meanwhile, please refer to the following blog get more inspire.
Managing Software Updates in Configuration Manager 2012
http://blogs.technet.com/b/server-cloud/archive/2012/02/20/managing-software-updates-in-configuration-manager-2012.aspx
We
are trying to better understand customer views on social support experience, so your participation in this
interview project would be greatly appreciated if you have time.
Thanks for helping make community forums a great place.

Best practice for data migration install v1.40 - Error 2732 Directory manag

Hi
I'm attempting to install SAP Best Practice for Data migration 1.40 on Win Server 2008 R2 (64 bit).
Prerequisite error
Installation program stops with missing file error
The following file was not found
... \migration\InstallationWizard\BusinessObjects Data Services\setup.exe
The file is necessary for successful installation. Please connect to internet or refer to Quick Guide (available on SAP note 1527151) for information regarding the above file.
Windows installer log displays
Error 2732 Directory Manager not initialized
SAP note 1527151 does not exist or is internal.
Any help appreciated on what is the root cause of the error as the file does not exist in that folder in the installation zip file.
Other prerequisite of .NET 3.5.1 met already.
Patch is released since 20.11.2011 so I presume that it is a good installation set.
Thanks,
Alan

Hi Alan,
There are details on data migration v1.4 installations on SAP website and market place. The below link should guide to the right place. It has a power point presentation and other useful links as well.
http://help.sap.com/saap/sap_bp/DMS_V140/DMS_US/html/index.htm
Arun

Best practice for Active Directory User Templates regarding Distribution Lists

Hello All
I am looking to implement Active Directory User templates for each department in the company to make the process of creating user accounts for new employees easier. Currently when a user is created a current user's Active directory account is copied, but
this has led to problems with new employees being added to groups which they should not be a part of.
I have attempted to implement this in the past but ran into an issue regarding Distribution Lists. I would like to set up template users with all group memberships that are needed for the department, including distribution lists. Previously I set this up
but received complaints from users who would send e-mail to distribution lists the template accounts were members of.
When sending an e-mail to the distribution list with a member template user, users received an error because the template account does not have an e-mail address.
What is the best practice regarding template user accounts as it pertains to distribution lists? It seems like I will have to create a mailbox for each template user but I can't help but feel there is a better way to avoid this problem. If a mailbox is created
for each template user, it will prevent the error messages users were receiving, but messages will simply build up in these mailboxes. I could set a rule for each one that deletes messages, but again I feel like there is a better way which I haven't thought
of.
Has anyone come up with a better method of doing this?
Thank you

You can just add arbitrary email (not a mailbox) to all your templates and it should solve the problem with errors when sending emails to distribution lists.
If you want to further simplify your user creation process you can have a look at Adaxes (consider it's a third-party app). If you want to use templates, it gives you a slightly better way to do that (http://www.adaxes.com/tutorials_WebInterfaceCustomization_AllowUsingTemplatesForUserCreation.htm)
and it also can automatically perform tasks such as mailbox creation for newly created users (http://www.adaxes.com/tutorials_AutomatingDailyTasks_AutomateExchangeMailboxesCreationForNewUsers.htm).
Alternatively you can abandon templates at all and use customizable condition-based rules to automatically perform all the needed tasks on user creation such as OU allocation, group membership assignment, mailbox creation, home folder creation, etc. based on
the factors you predefine for them.

"best practice to maintain the SAP OM Org Structure"

Hi SAP Experts,
My client want to have an best practice or an safe process to update, better and maintain their existing SAP HCM Organizational Structure. In one way you can say that i am doing an process oriented job.
Our client system is not up-to-date due to the lack of user awareness and complete knowledge on the system. Due to this they are unsure on the accuracy of the reports that comes out of the system.
As a HCM functional consultant i can look from the technical perspective but not on this process oriented role. I need your guidance in this regard, please sugguest me how can i move ahead and make some really valuable recommendations ? I am confused on where to start and how to start. Please help me in this regard.
Thanks in advance,
Amar

The only thing u need to keep in mind the Relatioships between the objects in OM
check the Tcode OOVK for relationships and assigning those objects PP01 , PP02
Re: Organization Structure
this thread may help u
let us know if there is anything else
Edited by: Sikindar on Dec 4, 2008 9:36 AM

What is the Best Practice for publishing Offline Root CA Cert and CRL to Active Directory?

Hi,
I've read and seen in a few labs different approaches to what is published in Active Directory for a Offline Root CA. I've seen just the Root Cert published to AD as well as the Root Cert and the Root CRL published to AD.
I can understand why the Root Cert is published to AD, but why would the Root CRL need to be published to AD, especially if my Offline Root CA just issues the Cert for my Subordinate Issuing CA? So looking for Best Practices here.
Thanks for your help! SdeDot

On Sun, 22 Feb 2015 18:44:25 +0000, Andrzej Kazmierczak wrote:
Best practice is to publish CRL to 2 alternative paths - LDAP for your internal users to access them on the first place and HTTP as an alternative option to LDAP and as the only option for your external users.
No, the current recommended best practice is to publish to a highly
available HTTP location first (and possibly the only CDP) that is available
both internally and externally. This covers Windows and non-Windows
devices, domain joined and non-domain joined devices and internal and
external devices as well as multi-forest scenarios with no trust between
forests.
Paul Adare - FIM CM MVP

Best practices for Refreshing Unv Structure for SAP based universes

This may sound basic and I apologize for asking this, but the refresh structure in SAP based universes is somewhat different from noraml universes.
Here is what i'd like to do:
- Hide all the L00 objects.
- Rename all the L01 objects and move them to a new Class.
- Change some of the detail (attribute) objects to dimension objects.
- Hide the format and Unit for key figures.
- Hide all of the classes/subclasses that get automatically generated when a SAP based universe is created.
I have noticed that when I do the above and refresh the universe, it assumes that all these objects have gone missing from the original classes and adds them back to the universe.
I also want to make sure that if the select of an object gets updated and the object is re-named, then it should automatically pick up the change.
Lastly, I have some reports which were built prior to this renaming. I want to make sure that the reports do not break.
Thanks,
Kashif

Hi,
This thread is really old.
Yes it was a common problem back in the earlier XI 3.x days . a lot of bugs in this area were eliminated by the time of XI 3.1 SP03 FP3.x - you don't quote your version.
Actually, you need to be aware that a refresh structure is often not needed, and can corrupt the olap universe . Pls check out Note 1278216 - What are the best practices for OLAP Universe Change Management when using SAP Integration Kit?
in essence :
Only use 'Refresh Structure' functionality If:
- A new Object (Dimension/Characteristic) has been added to the BEx query (Rows/Columns/Free Characteristics)
- A new Variable Restriction has been added to the Bex query Filters
Do not use 'Refresh Structure' functionality after:
- Having modified a STRUCTURE in BEx. i.e. 'Detail view of Formula' or 'Details of Selection', or changing the General Description of structure members.
- Doing manual actions on objects/classes in the OLAP Universe like: Move ; Cut/Paste ; Drag/Drop ; Hide ; Delete.
(because these workflow can lead to corruption)
regards,
H

Best practices for using the knowledge directory

Anyone know when it is best to store docs in the Knowledge Directory versus Collab? They are both searchable, but I guess you can publish from the Publisher to the KD. Anyone have any best practices for using the KD or setting up taxonomies in the KD?

Hi Richard,
If you need to configure dynamic pricing that may vary by tenant and/or if you want to set up cost drivers that are service item attributes, you should configure Billing Tables in the Demand Management module in 10.0.
The cost detail functionality in 9.4 will likely be changed to merged with the new pricing feature in 10.0. The current plan is not to bring cost detail into the Service Catalog module.

Advice on Best practice for inter-countries Active Directory

We want to merge three active directories with on as parent in Dubai, then child in Dubai, Bahrain and Kuwait. The time zones are different and sites are connected using VPN/leased line. With my studies i have explored two options. One way is to have parent
domain/forest in Dubai and Child domain in respective countries/offices; second way is to have parent and all child domains in Dubai Data center as it is bigger, while respective countries have DCs connected to their respective child domains in Dubai. (Personally
i find it safer in second option)
Kindly advise which approach comes under best practice.
Thanks in advance.

Hi Richard
Mueller,
You perfectly got my point. We have three difference forests/domain in three different countries. I asked this question becuase I am worried for problems in replications.
And yes there are political reasons due to which we want to have multiple domains under one single forest. I have these following points:
1. With multiple domains you introduce complications with trusts
(Yes we will face complications that is why I will have a VM where there will be three child domains for 3 countries in HQ sitting right next to my main AD server which have forest/domain - which i hope will help in fixing replication problems)
2. and
accessing resources in remote domains. (To address this issue i will implement two additional DCs in respective countries to make the resources available, these RODCs will be pointed toward their respective main domains in HQ)
As an example:-
HQ data center=============
Company.com (forest/domain)
3 child domain to company.com
example uae.company.com
=======================
UAE regional office=====================
2 RODCs pointed towards uae.company.com in HQ
==================================
Please tell me if i make sense here.

Best way to import a directory structure

I have been using iTunes on my new Mac to download stuff from the iTunes Store. I'd like to start using it to play music from my PC's archives.
iTunes seems to organize its display based on metadata in its own files, gleaned from the ID3 tags. It seems to ignore the directory and filename.
The organization method I've followed to this point on the PC is the exact opposite - all music is organized by directory and filename. ID3 tags are used in most cases, but they are not correct across the board. A slim majority are totally correct and ready for importation.
The library is about 10k songs large, so manual retagging is out of the question. Importing the directories as-is leads only to extreme frustration.
Can anyone recommend a utility for retagging files automatically based on directory and filename info?
Thanks in advance for any help you can provide.

Chris CA wrote:
The organization method I've followed to this point on the PC is the exact opposite - all music is organized by directory and filename.
Actually, that's exactly how iTunes does it.
Artist, Album, Song name are Artist directoy, Album directory, song name is filename.
True, that is how iTunes organizes music on the filesystem. My point is that iTunes keys its library database on ID3 tags, not pathname. In other words, I cannot organize my library by directory and filename. (With the intent of updating their tags, honest)
Unless my impression is wrong, I must tag all files correctly before adding them to iTunes, to prevent incorrectly named tracks in the library.
What are your directory and file names?
See this script -> Tags Mirror Artist-Album-Filename
"This script copies the names of the Artist folder, Album folder and filename of the selected tracks to each track's artist, album and song name tags, respectively. Your files and folders must be arranged in this format:
Artist folder:Album folder:filename.xxx"
Or this one...
Filenames to Song Names
This script will rename the selected tracks in iTunes with their filenames (minus the extension)."
This is great stuff. Thank you for pointing me at this site. I'm sure I will need to roll my own script as my directory structure is different than what these scripts appear to address. Ah well, it's a good opportunity to learn AppleScript.
Fwiw, generally my convention has been:
[MusicRoot]/Genre/Artist - TrackName.ext (for one-off tracks)
[MusicRoot]/Genre/Artist - Album/Artist - Album - TrackNum - TrackName.ext
[MusicRoot]/Compilations/Album/Album - TrackNum - Artist - TrackName.ext
Thank you!
-p

Package and Directory Structure

Hello --
I work in a group that supports 3 web sites. (b2b, b2c, b2e)
We're just beginning to develop Java in-house and
currently using Solaris and JDK 1.2.x.
I need to propose a package and directory structure strategy.
The "reverse the domain name" guideline makes sense to me.
My first thoughts are: ("classes" dir could be created anywhere)
classes/com/ppco/b2X/ <--- for .java and .class files (development)
lib/ <--- for b2X JAR files (ready for test or production)
util/ <--- for our utility classes like DBAccessor
lib/ <--- for our JAR files
sun/ <--- for classes like com.sun.mail pkg
lib/ <--- for JAR files like mail.jar
org/ <--- for classes in org.omg.CORBA pkg
lib/ <--- for JAR files
We need to handle 3rd party classes.
Development would be done in the b2X tree and JAR files would
be installed the lib/ dir for testing and release to production.
Does anyone have recommendations or experiences to share ?
Are there some things to avoid ?
Thanks !
Al

Hello Al,
you are on the right track. A typical convention I follow is:
<project>
bin - for startup scripts, etc. to run your application
build - for build scripts (not necessary if you build using your IDE. See below.)
classes - for my compiled classes
lib - for my 3rd party libraries
src - for my source code
test - for my test code (see http://junit.org/ )
That's the project hierarchy. The src (i.e. the package heirarchy) structure is another story.
As you say, you start with the reverse domain name. This is to give your packages a unique namespace. After that, your best guide is practice. Packages can be larger or smaller, depending on your coding practices. Usually you would have these (exact names may differ), plus others:
com/ppco/client
com/ppco/server
com/ppco/common
com/ppco/db
I think your break down of sun, org, etc. is a bit too much. If you would like to do so, however, I recommend you do the separation under /lib. This way, the top level project directory is not polluted by the different types of libraries in use.
Regards,
Manuel Amago.
From build above: I would suggest you always build your release distribution directly with the JDK, not using any IDE compiler. This is because Sun's JDK is the reference implementation, thus, any compatibility issues are not yours.
An easy way to achieve this is by using ANT (see http://jakarta.apache.ort/ant/ ).

Best Practice Directory Structure

Similar Messages

Maybe you are looking for