How to externalize application properties outside mule 4 - mule

I'm currently looking for a solution to externalizing the application properties for the different environments in Mule 4. The idea is to make it flexible but also to make access per role available. I have seen several solution like Custom Configuration Properties Provider. I'm looking for a solution that might be easy to implement.
Does anyone have any experience with this?
thanks in advance

Related

Customize an Archer Application from an external Application

I would like to make a desktop application (C#) that can create a custom Archer application and modify one of my current Archer applications (and subforms). I checked the SOAP and REST API but they have no functions for this purpose.
I was wondering if this is possible, and if it is, then can you help me get in the right direction to solve this?
is there a way to automate the creation of a custom archer application?
i've been thinking about creating this custom apps adding directly their info in the archer database. Nevertheless, i don't know if this could be a good way, since the archer database is complex and i haven't found any documentation about it.
Thank you for your time!
you are going to struggle to get any support from RSA for doing this. It is possible to reverse engineer the database to make changes but you would be working without any documentation and would be vulnerable to any system updates changing the way the data is stored as RSA release new updates.
You could look at reverse engineering the packaging and have your desktop application output packages that could be installed. This limits your impact in making direct database changes but is still not advisable.
Sorry I can't offer more help to a fellow Archer hacker, good luck!

Adding direct reference to a WCF project

Wondering if somebody done it before. We are using WCF in our current architecture and when internally not going over the boundaries I would like the ability to reference the project directly.
I don't want to use tcp or named pipes, I would like to know if is technically possible to alternate between a project reference or normal communication with WCF eg tcp-http etc..
Is there any example or approaches to achieve this?
Looked on google and stackoverflow but I could not find an answer if the title was the same
Many thanks

jqgrid with turbogears 2

I have been recently trying out crudRESTController in TG2.1.
Overall, I found that--
1] The community is helpful.
2] But, it is hard to find a help topic or docs, if I get stuck-up with a particular issue.
e.g. The name of instance of crudRESTController must be a plural of underlying model name.
Otherwise, it won't work correctly. It is nowhere given in their docs.
Hence I am thinking to use jqgrid for crud functionality.
Can anybody please point out whether it would be better to use jqgrid or stick to crudRESTController!
(keeping in mind the control I can have over the code, rapid application development, deployment, speed, etc.
Thanks in advance,
Vineet.
Give a look at EasyCrudRestController from tgext.crud, it provides an easy way to create working Crud Controllers on the fly. For more deep customization you can tune it as you would with CrudRestController.

OpenSwing Framework

Is OpenSwing a good framework for developing professional desktop application?
I was recently using the OpenSwing Framework. I can say only the best for the functionalities which are provided with the framework. It is a multitier concept with excelent data binding possibilities. My App uses a small Derby DB in background and I’m managing it with hibernate.
I’m sure, you will be able to advance very fast and provide a working prototype very quick. I would advice you to read the available doc first and to run the provided examples (http://oswing.sourceforge.net/).
However, it has another side which you should be aware of and you will probably notice by yourself if you run the examples. The GridFrame, GridFrameControler, DetailFrame, DetailFrameControler etc classes are not really generic. There are a lot of dependencies bult in and you will have to customize them again and again for every single implementation (can be seen in the demos).
I had another approach, I invested some time in building my own classes which are generic and using the unchanged OpenSwing classes in the background first. Now I’m only setting the properties file where all details are pre-defined. The rest is generic and I don’t have to re-code again and again for every single frame.
I hope this will help.
Regards
I used the openswing in team for more than two years.
It's a pretty nice swing framework for the enterprise development used in the Internal.
It provide great component based by MVP pattern ,such as grid , document ...
If you try it , It's a good article for you about Model-View-Presenter
And try the demo in the source,It's quite good.
The JAllInOne is also a good demo for the framework also made by the mcarniel
and It's a personal project only developed by mcarniel. Thanks mcarniel's great work.

Staying open with DI/IoC containers

I am involved with several open source projects which taken together provide an application development framework. The question I have is what mechanism(s) should I provide for integrating them with each other?
On the conceptual level the answer is clear - DI/IoC. The "only" problem is to decide which one. In several installations we used StructureMap, but then a user came along who wanted only one of the components and wanted NInject.
So, to qualify the question, how should I go about building my components so that they can be integrated with each other (and 3rd Party) using a variety of DI/IoC containers.
The best I could come up with was to separate out all integration code into separate projects and then have a project per supported IoC container, but this sounds suspiciously like IoC squared.
Any bright ideas? or I am just thinking too hard?
P.S. for the curious: NDjango; Bistro; Workflow Server
As long as you develop reusable components, you can implement them in a DI-friendly way without ever referencing any particular DI Container.
It' only when you need to compose an actual, running application that you need the DI Container, but as I understand, you are developing a framework, and it's best to keep it DI-neutral.
See this very related question (almost a duplicate).
For inspiration about integrating several projects while keeping them independent, see the Castle Project.