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.

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

The price of the solution could improve.

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.

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.

Number one thing is we are an Oracle shop, so we do Oracle ERP testing, and that add-in from UFT, that technology is not in LeanFT right now. So all of those automations continue to be run in and maintained in UFT. In order to transition those we would need that support, so that is probably the number thing.

Additionally, just more of the UFT functionality that they have now brought in. We're BPT users as well, so they brought in some BPT functionality, but limited. So we can't have a UFT and a LeanFT component existing in the same BPT for example. That is a limitation to us. It's a step in the right direction that we can do some of it now, but we're not there.

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.

In future releases, I'd like to see disruptors such as IOT and test drones. For some of the new stuff that’s coming through, we need to have a clear, well-defined road map of when we’re going to receive new capabilities and new features. This allows us to plan and better work with our customers so that we know exactly when to tell our customers to expect these new capabilities. These industry disruptors actually shape how companies are able to deal with these new technologies. So there’s always room for improvement.

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.

The tool is not the problem. The problem is that we can't get the tool working, because there are other issues. We had a meeting with all the banks and several customers had the same problems. There should be a smarter and quicker way to upgrade UFT Pro.

It must be on the roadmap now, because they really lost with Mobile Center when they released a new version and it was not backwards compatible. It's not very easy for big companies to provision new versions. We have 38,000 PCs in organization, and we're not allowed to do anything on them. Everything has to be taken step-by-step. It does have a learning curve.

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.

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.

LeanFT could support additional technologies because we use it for a lot more than just web and Java and some Windows apps. Further support for other technologies would be nice. I can't rattle any off the top of my head but ones that we use internally.

The tool could be a little bit easier. 

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.

Improvement is still needed for stability and performance. Still, it is at the beginning of a technological shift for HP test automation. The tool would be best if HP made the OR functionality and performance similar to the HP UFT one.

It needs to be able to be used on Chrome, Firefox, and other browsers on Macs and not just Safari. That's a very key requirement for my organization.

I'd like to not have to use Selenium. I'd like to be able to do headless scripting and not just always be UI serving.

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.

I think the one thing we're basically asking for should be JavaScript support, but I think they will start adding JavaScript support in the future. We haven't moved into that; we’re still into LeanFT. We're still finding out what's in there. Once we know what we can really do in LeanFT, then I think we can start providing feedback regarding enhancements we want to see in the solution.

In the next release, I'd like to see ties into C# and Java, enabling us to work more closely with development. However, from what I’ve seen, LeanFT object recognition looks great. The identifiers for an object can be the same for multiple objects, and test automation can detect that. But LeanFT object identification center allows you to see that the identifiers are not necessarily unique, which is very beneficial.

At the moment, we are happy as it is. We don't have any kind of specific technology requirements for improvements, at least not at the moment.

But, support for open source solutions, such as the Robot framework, which is actively used, might be really helpful.

There are still some stability issues. Also, the integration with ALM is not perfect. There issues with parameter parsing.

We don't have ALM Octane now and I've never used it, so I don't know whether it would beneficial. But I think we're too big to easily switch to another test management solution.

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.

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.

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.

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.

In the next release I'd like to see HP enable LeanFT to work with Sauce Labs. I know HP has their own solution, but it would be better if it were more integrate-able in the spirit of being more open-source, more friendly.

As the releases come, support for all technologies in the systems under test that UFT currently support will be the obvious place for improvement.

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.

I think the biggest issues that I've seen, and this is a personal view of mine, is that most of the HPE products have a common look and feel to them. I'd really like to see it be a little more customizable to a use and user standpoint. For example. I happen to be colors blind, so I'd like to see more vivid colors on the UI, and things like that. It would make it a little more flexible from the customer standpoint.

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. 

We push one button and the tests are completely executed at once, so just have to analyze and say it's okay. It would be nice if this could be entirely automated. For the moment, it's people who have to perform this task.