with the Eclipse plugin Sirius you can develop a user interface for creating models given the rules of your EMF meta-model much easier than getting your hands dirty directly with GMF. This is really cool!
However, how can you run the models in the user interface? Can you please give me some pointers?
For the purpose of the explanation, let's say that you create in Sirius the user interface for creating statecharts, but how can you run them? Throw Event 1, running State 1 is stopped and State2 starts... (you know what I mean).
Thanks for any pointers!
There are several approaches to do that, one of the most promising one is using a "lab" component https://www.eclipse.org/sirius/lab.html and is called "Sirius Animator".
It is currently compatible with Sirius 2.0 and would probably need to be upgrade for the version you are targeting.
Related
We are three testers and going to prepare automation project with selenium and java code so what are the steps for environment setup , scripts integration and running the testcases and getting the results for the whole project suits
So there are a few things we have to use in order to allow multiple engineers to work on the same framework.
Step 1) Creating the framework, assuming you know how to do this already, you have working tests you can skip this stage, however if not please follow the tutorial i link below.
http://toolsqa.com/selenium-webdriver/
Step 2) Creating a REPO, my preference is GitHub, you can use any git repo however i will post the guide to set one up with GitHub, its a similar process for all. This will allow you to merge code properly without causing conflicts.
https://help.github.com/articles/create-a-repo/
Step 3) Source Control program - to push, pull and fetch from your GitHub Repo, you can do this from Command Prompt however i find cloning the repo into a program like 'SourceTree' is really easy, so i've posted that below.
https://confluence.atlassian.com/get-started-with-sourcetree
If you follow these 3 guides, you will be able to have your automation test scripts on GitHub by the end of the day.
If you have any more questions please do not hesitate to ask.
All the best, Jack
The easiest and most logical way to do this would be to create one branch in your CVS (git or SVN, etc) and have each person setup the dev environment in the same way. Work exactly like developers and pull code before you check-in/commit (this will ensure that one small error does not break your framework) and swear to resolve conflicts during merge (to ensure you don't step on each others' toes).
Also, before you kick off, agree on a standard of coding (including package naming, design pattern usage, filename and methodname usage) and if this is in sync with the dev coding standards in your company, even better.
There will be a few hiccups along the way, but experience is the best way to create a process for your development and check-in practices.
Good luck with your new project and happy coding!
You have asked two questions, in my opinion the answer of your questions is.
how multiple automation testers work in same selenium project - You can use any version control system, Git Hub is the best option which gives you a lot of facilities. You all three can work on same project at same time or you can go for any centralized version control system like tortoise svn which is not much likely used now a days. I will suggest Git Hub for that.
what are the steps for environment setup , scripts integration and running the test cases and getting the results for the whole project suits - It depends on various factors like application and the kind of framework you want to use, there are many frameworks which are widely used for automation testing like Modular Framework, Data Driven, Keyword Driven, BDD, Cucumber, TestNg etc or if you have bandwidth and time you can design your custom framework as per the needs.
I hope I put some glimpse on your queries.
Thanks
I recently upgraded to Flash Builder 4.5 for PHP since I’m working on a big PHP/Flex project and thought that it would be easier to integrate the two.
What do you know? The feature that generate the php class to feed a flex/PHP service trough AMF is appears to be missing!
Time to downgrade or am I overlooking something?
I believe it's still there, but that feature should only be used for building a test application. It breaks a lot of best practices such as creating the connection in the class (it should be injected or retrieved from an outside class that manages the DB) and it puts the database definition info in the class (username, password, etc.). I believe that when you use that feature that a popup tells you that it is not intended for production use. When building a service class it is best to define your classes based off of the functionality you need to provide.
I am thinking there must be some libraries out there that people have developed which can be used as "plugins" or whatever people call them to do simple and common UI types of things.
I am using the message board idea as just an example, but I am looking for a general solution. For example, is there a place where I can browse "gems" for RoR that just take care of some UI component?
How do people usually integrate such pieces as a message board present at the bottom of every page, or some other ui tool without writing their own, or using a CMS?
Thanks,
Alex
Two good places to browse gems are http://ruby-toolbox.com/ and of course http://rubygems.org/
I'm looking for a maven plugin that would generate ActionScript3 classes from Java classes in order to access them by object remoting.
I've seen FlexMojo but it uses the GraniteDS generator which create some problems when it comes to map Enum objects (which can be fix through a workaround that is describe here : http://dev.c-ware.de/confluence/display/PUBLIC/Flexmojos+generated+AS3+model+with+Enum+support+using+BlazeDS?focusedCommentId=7634946&#comment-7634946 if you've googled your way here this might be useful) when working with BlazeDS.
Everything that I found so far are people who explain how to generate VO classes on flex side using Flash Builder 4, but this solution can not be used in an industrial development environment.
Take a look also on http://flex-annotations.aixcept.net/examples/actionscript.html
I also found this one, and while it is not a maven plug-in it could possibly be turned into one:
https://sourceforge.net/projects/cleartoolkit/
It was created by the guys who authored the book "Enterprise Development with Flex". Look for the utility DTO2Fx.
I recently ran across this open-source project: http://www.codeplex.com/Sculpture
Sculpture is a code-generator which allows you to design your domain model and then use persistence 'molds' such as NHibernate/EF and probably more to generate repositories. It takes care of all the mapping and Data access generation. It looks like it does a heck of a lot more, but i'm pretty much sold at this point. However, i would like to know if anyone here used this for real-world solutions. If so, how well does it scale? How did the generated DAL work for you?
We have taken this out for a spin on a new project we had. Always wanted to try out nHibernate but the XML hell of the config files always put me off. Tried a number of these type "helpers" this is the easiest to use so we decided to use it.
I like the facts that you can just point at databse and it will create object based on what it sees and will resync on demand. It was also good that you could add custom methods to these objects and has a very well written syntax for the generated code all using LINQ-To-NHibernate.
It also generated a respository project over the top to allow you to use this design pattern which we did (again another personal goal).
It all worked out very well, created a social (facebook like) site for people with a medical condition, got everything fully working in 3 weeks with onlt 3 people.
On the whole very happy with this - in fact bought a licence for Sculpture 2.1 Nhibernate mold. (note open source but not for nhibernate mold :-) )
There is room for improvement
Windows 7 64 bit had to run visual studio in admin mode (my pc only) - not sure why
TeamCity continuous integration had some issues with paths to dll files - the two generated project (ORm and Repository) use different paths to base dll's - so needed to work out position of source in filestore to automate the build but worked fine.
Would like to heard any views on this or other product which people think are better ..
Looking to do a project where we will create all the objects in sculture and get it to produce/update the database for us - will be odd as we are db type developers but will be good to see how this works..
Thanks
Simon
I would say that most ORMs are generated in some way at this point. I'm using Linq-to-SQL for a current project, you setup your model and generate code (with attributes from the model).
In short, it's nothing new and there's no reason to rebuff it just because it's generated.
Most ORMs can or do work this way, and they do so well.