Take the 2-minute tour ×
Software Quality Assurance & Testing Stack Exchange is a question and answer site for software quality control experts, automation engineers, and software testers. It's 100% free, no registration required.

I've started to evaluate RFT by getting their trial version, I'd like to hear on the strong points that makes you folks to stick with RFT. I'm using QTP and Silktest now and trying to find how useful it would be to replace it with RFT.

Keeping the cost discussion aside, I would like to hear the technical points.

NOTE: I had another post which lacked specifics and could have been resurrected but would be too much hence started a new thread.

share|improve this question
2  
What is your target application? –  Bruce McLeod Oct 16 '12 at 1:57
    
It's a web-based application with .NET, jquery...etc it opens up 2-3 browsers at a time and throws excel, word and pdf files at times for verification. –  user1424385 Oct 16 '12 at 13:24

2 Answers 2

I have never used QTP nor Silktest, so I cannot make a comparison.

In my opinion RFT's strong points are (in no particular order):

  • Interacts well with the required browsers (IE6, IE8, IE9)
  • Object maps with extensive Regular Expression support for handling object recognition and a dynamic find by properties method.
  • Java based (I added a lot of external libraries for extending test functionality, interacting with the Operating System and other tools)
  • Manual Verification Points (I fetch the base data from a database, xml file, RFT own datasets, etc)
  • HTML logs for non-tech people with screenshots, and it's quite structured, I manage all of them with a tool, also good for searching, reporting and so on.
  • Scripts can run from command line or managed from other IBM Rational products, like Rational Quality Manager (we have our own tracking software, I wrote some custom tools to handle it)

I found 2 very weak points. The first is the crappy "Simplified Scripting" that constantly crash/fool up things/delete code without telling you. Disable it.
The second is that sometimes it hangs in some deadlock while reading or writing IE6 memory, but I bet it's more a IE fault ;-)

share|improve this answer
    
Thanks for the feedback, it definitely helps. –  user1424385 Oct 16 '12 at 19:07
    
I tried with RFT and not very impressive at all. Right from day 1 it has shown all kinds of failure, object map takes about 5-8 mins to add an object from "Test Object Browser" or "Drag and Drop method". –  user1424385 Oct 23 '12 at 20:36
    
Wich OS/Java version are you using? Did you disable the Simplified Scripting? RFT isn't fast, but 5 minutes sounds too much to me. –  Alessandro Da Rugna Oct 24 '12 at 6:36
    
I replied to the other thread, but as an update RFT-Java was not helping, it hanged the browser IE/FF. so started with RFT-.Net, works fine except that it takes so much time to add an object. I've kept things as default. do you know the settings in .net to remove simplified scripting? I'm googling for it. –  user1424385 Oct 24 '12 at 16:23
    
I do not know if Simplified Scripting is enabled for .NET, but this is the guide for disabling it: www-01.ibm.com/support/docview.wss?uid=swg21405134 –  Alessandro Da Rugna Oct 25 '12 at 11:45

We've moving away from RFT (with Java scripting) in our department, for the following reasons:

  • Large and slow. RFT takes minutes sometimes to find an object, and generally adds a lot of overhead to our testing
  • Object maps cannot be easily maintained or merged. We found that with large enough maps, sometimes it'd just lose an object, and we'd have to re-map it. They're binary, so you can't merge them in source control, meaning conflicts were a major pain.
  • As we move further away from object maps and "record-and-playback" testing, we found more and more strange bugs (like the JVM segfaulting), library limitations ("RFT just can't interact with this object at all, sorry!"), and performance issues (2-3 minutes to find an object). There's almost no information out there on how to solve these problems, and IBM has been less than responsive. The documentation is a joke -- most of the javadocs are just generic boilerplate.
  • Unregistered Object Exceptions. RFT has some kind of internal garbage collection which we had 0 control over; it kept removing references, thus crashing the test the next time we tried to use the object we just found. We never did find a good solution, though we had a handful of tricks we could use to occasionally mitigate the issue.
  • RFT is incompatible with JUnit, meaning when we now have extensive framework and logic supporting our tests, we can't unit test that logic.
share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.