What needs improvement with Micro Focus UFT Developer?

The solution could improve by working better with desktop applications and websites. It is also suggested that the design and some functionality could be better.

Stability depends on the company's infrastructure and end-to-end infrastructure. When I used the tool in my project, we had a big problem with many users using it simultaneously.

There's room for improvement, especially when I compare OpenText to newer tools like NeoLoad. NeoLoad is a strong competitor to LoadRunner and it's very fast. It saves a lot of time when creating scripts, generating load, and creating reports. OpenText UFT Developer takes a bit more time compared to other tools. JMeter, for example, also does performance testing, but it takes much longer to create scripts, record, correlate, and prepare. LoadRunner is better than JMeter, but NeoLoad saves a lot of time for the consultant, and time is money.

Object definition and recognition need improvement, especially with calendar controls. I faced challenges with schedulers and calendars.

The tool could be a little bit easier.

In my opinion, the price is too high. It could be improved. The support for .NET Framework and Visual Studio in Micro Focus UFT Developer is currently limited. At present, only Visual Studio 2019 is supported, despite the release of a newer version (2022). Similarly, the tool only supports up to .NET Framework version 4.3.8, while there have been six newer versions released. This is an area that could be improved upon, particularly in the Windows environment.

They need to reduce the licensing cost. There's pushback from customers because of the cost. The tool also takes a lot of memory. It's really heavy on the CPU. If I need to run the virtual machine, I cannot go beyond 8GB RAM.

The price of the solution could improve.

The issue with all the integration is that it can become very costly and expensive and we'd like to be able to recommend one single tool that will do it all.

UFT Developer is good, but it requires high-level development skills. Scripting is something that everybody should know to be able to work with this product. Currently, it is very development intensive, and you need to know various scripting languages. It would be good if the development effort could be cut short, and it can be scriptless like Tosca. It will help in more adoption because not every team has people with a software engineering background. If it is scriptless, the analysts who wear multiple hats and come from different backgrounds can also use it in a friendly manner. It is also quite expensive.

You need a more modern language to write test cases in because Visual Basic is not powerful enough. I have to keep the remote machine open while the tests are running, otherwise, it leads to instability.

The pricing could be improved. It's a bit high right now. They could make it more reasonable. Currently, the functionalities do not exist on Ranorex or UFT, however, as the industry continues to evolve and put more applications into a mobile device, the implementation of testing in a mobile device, using either Ranorex or UFT or XYZ, et cetera is just a natural progression of their functionality. This world is not made of only PCs. The solution needs to offer mobile application testing. That said, that is a very difficult requirement as one would expect that they would need to test across various platforms - including iOS, Android, Chrome, et cetera. Every mobile device seems to have its own standard. The mobile industry would need to also standardize a bit. However, that's not likely. There also hasn't been a unification of testing on mobile devices regardless of the browser. It requires a good discipline maturity and if the discipline is not there, I spend my time holding people's hands and showing them how to install everything, how to connect, et cetera. Without users with knowledge, users that understand coding, it's very hard to adopt the application.

With Smart Bear products generally, you can have only one instance of the tool running on a machine. Whereas with Selenium, you can have five instances of the test on a single machine. The lack of multiple instance capability is not as useful.

As I have only been using this solution for a few weeks, I am really not in a position to say what needs improvement. I need to use it more where I can explore all the available features. If I am not able to perform any operation, then I will be in a position to answer this area of improvement better. In the next release, I would like to see integration with different cloud-based tools such as Azure. Microsoft already supports the internet.

UFT is more code-based, and we have to have knowledge of VB scripting to prepare the automation test cases. This is an area that is lagging behind with UFT. One of the biggest challenges we face is not being able to easily interact with ALMs, other than HP ALM. This is an area that needs improvement. In the next release, I would like to see the connectivity improved to be less complex and more stable. Also, they can improve the coding interfaces to be easier and closer to English or any other international language, rather than a programming language.

This is a script-based tool and the usability needs to be improved. Easier connectivity and integration with SAP would be helpful. In the future, I would like to see module-based tests instead of scripting.

A basic level of programming knowledge is definitely needed to use this solution. It would be improved by adding a drag-and-drop interface to help alleviate the coding. We are investigating solutions where a layperson, with an interest in automation, can begin to work with the tool.

UFT is sometimes difficult to run. For example, the customer application is represented by an embedded browser control, waiting for input. If I want to recognize the browser then I need to first start the UFT Pro environment. This can be done from Visual Studio or the management console. The problem is that UFT is not able to identify the object that is inside the browser. In one of my test cases where I have to select the card, I need to right-click on a picture and then select an item from a drop-down menu. I had opened a ticket in version 14.02 and I spent two weeks speaking with people from Nigeria, trying to convince them that there is a bug in the software. I was finally redirected to the engineers who solved the bug, but they sent me a DLL patch as opposed to an official update. The support from Micro Focus needs a lot of improvement. My simulator is able to create a receipt as if it were printed from the ATM. However, in the current version of UFT, I am not able to perform an OCR on it correctly. The accuracy is about 20%. When I told support that our code was written in C#, they showed us some Java code and were convinced that it would work simply by using Java instead of C#. I would like the Object Finder Application Center to be improved. It is a plugin that is used to recognize the object on the screen, but it runs very slowly and crashes often.

There is quite a bit of room for improvement. As time has gone on the product has failed to improve. Basically, Micro Focus' UFT (Unified Functional Testing) was a good product 15 years ago when it was first introduced. They have not really made substantial changes to it since then — which they should have done to make the product more useful and competitive. The gap between it and the competition has shown in the product's lack of development. To improve the product they could better integrate the API and the GUI testing. At the moment, when you run the GUI testing, you run it in Visual Basic Script — which is a very old Microsoft product that Microsoft no longer supports. For the API testing, you have to write your tests in C# or C++. If you write a functional library for one test process, you can not use the same library with another test. A further problem is that even if you have a functional library written in VBScript, you can not use it for multiple projects. You have to make a copy of the library for each project that you use it with. Then, of course, every time you make a change, you have to replicate the change manually through the different projects and that is a real pain. A new feature that I would like to see is better integration between the API and the GUI testing so that you could use the same libraries and the same scripting languages and so forth. That is a major missing piece because of their lack of effort in development over time.

UFT is like a flagship of testing tools, but it's too expensive and people are not using it so much. They should work on their pricing to make themselves more competitive. The performance can be improved. There are much faster tools now. This solution is a bit older and works with older systems, but it's a bit slower because of this. They should modernize the product a little bit. The UI looks okay, but it also looks like something that is ten to twenty years old.

It's now too heavy and they should be making it faster. We do an attempt at automatic regression testing. We schedule a test to start at a certain time. It takes a lot of time to download the resources and start UFT. Competitors in this area have tools that start faster and run the test faster. For example, if the test at our side will take 10 minutes, another tool will do that in one minute. I would like to see them add a feature that tells you if you can run parallel sessions in it. If it were a lot faster than the Chrome version that would be a major win.

The parallel execution of the tests needs improvement. When we are running tests in LeanFT, there are some limitations in terms of running the same tests simultaneously across different browsers. If I'm running a test, let's say to log in, I should be able to execute it through IE, Microsoft Edge, Chrome, Mozilla, etc. This capability doesn't exist in LeanFT. Parallel execution of the test cases across different browsers needs to be added. The default activation of the services should be reduced to a bare minimum. When you install it out of the box, it enables everything and slows down the system. This needs to be adjusted to improve performance. The solution should have better integration with the test management tool.

As far as things that can be improved, it is a good solution so I think I can only do a comparison. We also use QC/ALM (Application Lifecycle Management [Quality Center]). It's a global solution that is managed with information from UFC from all over the environment. It has to be integrated with UFT. Really UFT could have this functionality built-in. We have 40% advantages and 60% disadvantages in our setup of UFT. This is because with UFT, we also have the problem that we have to use Windows Server and I would like to use Linux. For Selenium, we can use Linux so we have good performance. But we can't use UFT with Linux. It is impossible because in UFT we have to develop for UFT with VBScript and VBScript is only for windows and not for Linux. Another problem currently with the UFT — I think it is resolved in the new generation of UFT — is that we can't run tasks in parallel. In the new version, we can improve our workflow if we can choose to allow multiple tasks at runtime. So there is a problem with that currently. In Selenium, our development is done with Java technology — J2EE. So if we have an online community and we have a Selenium grid, we can run multiple tasks in realtime. We can't do that in UFT now because of its requirements, so it's a problem for us. When they come out with a solution for this issue, the product can be more flexible like Selenium and it will be a great benefit to us. To make UFT better, Micro Focus has to make UFT work in a stable environment. Right now, UFT is a problem all the time. It would help to have a community and a special forum for UFT, and even that is missing. We have good forums in Java and for Selenium, so it is possible to get solutions easily for those products. I think it would not be hard to do for UFT, and it would be better for UFT users if we had a good website. Users could help themselves and share knowledge and address problems and make up for the lack of support. We also don't have training for UFT. It is like they just made a product and don't care to support it. It is a good product, but not so perfect that it doesn't need support. I have to go to France to get certified. We don't have that ability here in Morocco. We cannot send everyone there, so it is a problem.