Please share with the community what you think needs improvement with IBM Rational DOORS.
What are its weaknesses? What would you like to see changed in a future version?
One of the things that many people complain about is it's hard to manage attributes. For example, tables or figures. This is something that can be improved. The most important improvement for me right that is needed is based on textual structure type, which has been good, but there are new trends and more model-based are required. For that, it's outdated, it does not work well. It's outdated when it comes to model-based requirements
The software is very dated and old. It's hard to get people to use it because the GUI looks like something from one of those giant Mac computers. It's not very user-friendly and can become slow very quickly, especially if you're not on site. It's been detrimental in this recent work from home era. If you have a lot of employees working from home, DOORS will operate more slowly than if they were on site. The DXL will be very slow if you write an inefficient tool and then your client will suffer. Not everything is necessarily written by IBM software engineers who know the system well and it will slow down as you put more tools and information into it. There's a tendency for garbage accumulation which is the simplest way I can put it. Processing of images needs to be improved.
One thing that I would like to see is a lower-cost version of it that we could use for smaller projects. Sometimes, we do projects for commercial customers who would benefit from something like DOORS, but it's just so expensive. It's just a monster, so a lower-cost version would be the thing that we'd like to see.
Usability and user-friendliness could definitely be improved as far as ease-of-use. The documentation could also be easier to use, I would say. Maybe, in general, they need to see things more from an end-user's perspective in their future development.
I found the user interface to be unintuitive. It's something they need to work on. I wouldn't say it is bad, per se. It is just like learning to write cursive. I would push for more extensive integration with other tools since, for example, I needed it to integrate with MagicDraw. Building in that type of integration and other such integrations would be helpful for our purposes. There could be a better structure around onboarding to get people started. It was unintuitive as to how to get started. It needs to be clear as to what the first things a user has to do in order to get going. There needs to be quicker access to tech support. When I have a two-minute question that takes two minutes to answer, it shouldn't take me 45 minutes and/or a few days of callbacks to get to the right technical support person. It's unnecessary and frustrating for the user.
The strict requirements for synchronization of the data could be relaxed. It requires a permanent connection with good bandwidth. This means that in an environment with remote networking that you need to go through a VPN or use some kind of virtual machine in the middle. We had some issues with the disconnection of desktop software and so on. The strict requirements of time synchronization between the DOS server and the client that request you to have a permanent good connection are difficult now that we are working more remotely due to COVID. The solution has some scalability issues. The performance could be improved. It doesn't run as smoothly as it could. The usability when you're doing writing tends to be similar to Windows. It's a rational style. It needs to be able to do drafting with drag and drop, copy and paste, etc. There needs to be more usability in order to help people move data, create drawings, etc. The solution should be able to support different formats and texts.
Some of the search queries could be improved. The interface needs an area to be able to type your query and actually be able to find them. It could be more stable. In the next release, they could scale it down a little bit and make it more stable.
The user interface for the Change Proposal System could be improved. When creating a proposal it is great and I have no problem with it. On the other hand, during a review phase, when many people are trying to look at the change and decide whether to accept it or not, the user interface is not really helpful because it just shows you the specific change. What we have done over the years to accommodate this is to create a specification module where we pull the proposed change features into that view. I can then look at the changes in the context of everything around it, and we can decide whether it is the change that we really want to make. If there were a way in the change proposal window to view the specific change in the context of the other things around it, including potentially other changes, then that would be helpful. The workaround that we have created allows us to view all of the potential changes in concert with everything that is not changing, which is ideally what the change proposal GUI should do. One of the people that I work with has expressed interest in a process where you have to propose changes to links, rather than just create them. In this way, you can maintain traceability under some form of configuration management for them as well. Personally, in 20 years, I have never had a program where we tried to control links to that degree. We would monitor them, but never had any formal change process for links. We are not allowed to use DOORS as our configuration management tool and instead have to use Agile PDM. This requires us to export data from DOORS and import it there. However, if DOORS were tailored a little bit better then we could use it as our CM tool and avoid using the other one altogether. More and more companies are getting involved with model-based systems engineering (MBSE). I know that DOORS has direct interfaces with many of these tools, although I have never used any of them so I don't know how simple they are to use. That said, anything that can be done to streamline and simplify the tool-to-tool interface between DOORS and other products is a good thing. For example, it should be easy to exchange data between DOORS and MagicDraw, CORE, Genesis, and others.
Complexity, performance, openness are the three areas that can be improved. The IBM architecture and specifically Jazz looks more complex. There are a lot of servers. It's quite complicated. The search capabilities lack in IBM Rational DOORS Classic for customers who have a database with a requirement of more than 25,000 records. For example, you can search easily for a module, but it's really difficult to look for keywords through the whole database because all the modules are separated into small components, which makes the search quite complex. This is something that's really annoying because when we want to make an impact analysis, we would like to analyze the product globally. It's quite difficult to manage. The fact that you can interact externally with data makes it complex. The approach is complex and doesn't work as expected. For example, when I tried to experiment with exporting some records, the tool crashed, but I couldn't find out the root cause, that is, whether it happened because of Rational Windows or lack of memory. It was just crashing. Logs weren't very clear. IBM can try to use more recent technology for different aspects and make it easy. They can also provide free integration from DOORS Classic to DOORS. Currently, all the customization in Excel is lost, which makes it very complex. It would be a feature to make new versions compatible with features in the past versions.
The user interface is old and clunky and in need of some improvement. In order for DOORS to be as good as it needs to be, you have to be able to get normal people using it. The problem is that because the GUI is so bad, you either have to spend a lot of money customizing the interface yourself, or a lot of money on training. Because the users avoid it like the plague, you don't get the power out of it. You end up having specialists in the requirements world doing the work instead of using it the same way you would an easier application, such as Word.
Implementing processes should be easier. The interface is not very user-friendly and has not evolved in a long time. People like to have a nice tool that is up to date with the latest technology, and this is not the case with DOORS.
What could make sense for this product is to improve is to develop a more efficient way to import and export documents from Office 365 like Excel, or Word and the other applications in this suite. Maybe, if possible, add a PDF document export or something like that. There are quite a few single steps that you have to take separately at the moment to make this happen. The parts are already implemented, but there could be a much more unified and efficient way to get that done. Again this is a repeated request from users and nothing has been done to implement it. What I would like to see is in the product is that eventually, IBM will implement additional software solutions for integration. There is one that I know of that used to have the name RPE. It stands for Rational Publishing Engine. I think it is a tool that should be implemented in DOORS because it belongs to IBM after they purchased Rational Software. With this integration, we would have a much better way to actually import and export between Word or other Office documents. On another side, it would be good to also see them integrate the GC (Garbage Collector) trace tool. It is a logical requirements engineering tool that would enhance some capabilities. It could be a specialty add-on but the integration needs to be there. This product is owned by IBM because formerly it belonged to Telelogic as well. It would be good for IBM and the users of these tools to have these solutions implemented in DOORS.
There are problems with the communication between DOORS and Microsoft Office. It doesn't matter which version of MS Office is used. When trying to communicate between Excel and DOORS or Word and DOORS and vice versa, problems arise. I would like to expect a more professional application here. A tool to gradually import scripts into DOORS Next Generation (DXL to Java) would be very helpful. Merging classic DOORS with next-generation DOORS using a web client would be a great idea. I would like to see a much more professional way to generate documents. A tool to import scripts, step-by-step into DOORS Next Generation would be very helpful. The merging of classic DOORS with Next Generation DOORs using a web client would be a great idea.
The HMI is difficult to use and the user interface should be better.
It's difficult to set the code on the solution. The testing part needs to be improved, but they have too much legacy to move to the latest IBM versions. I do not believe there are any more plans for this product. It's now a legacy solution. They should just ensure that they keep the technical support they have in place to assist existing users.
The GUI needs improvement in the following ways: * The OLE embedding is not very user-friendly. * The whole concept of having to lock and unlock, in order to switch the edit mode, is not user-friendly. In the next release of this solution, I would like to see integration with other tools. For example, for change management, and with tools like IBM Rational ClearCase. I know that IBM has now linked ClearCase, ClearQuest, and DOORS, but we have an older version so I do not know how good the integration is.
They have recently upgraded the product. And, it is definitely more user-friendly. It was really clunky to use previously. It might not be very fancy, but it always gets the job done. I think it would be nice if there was better integration with other tools, particularly diagrammatic tools so that you can do things like enterprise architecture.
If you were talking to someone whose organization is considering IBM Rational DOORS, what would you say?
How would you rate it and why? Any other tips or advice?