What needs improvement with Sparx Systems Enterprise Architect?

The product could be improved in terms of its ease of use and documentation. While it offers a lot of functionality, it can be difficult to grasp how to utilize these features effectively.

In general, we now have a problem with Sparx Enterprise Architect because many analysts think it's too complicated to use in a giant environment. So, two years ago, it was used very, very often, and it was a basic tool in the organization. Now, there's a discussion. Is it worth using Enterprise Architect in the company? We need a better, lighter tool to create, for instance, user storage. So, our company is at a specific stage in our product development or software development, where there are some people who want to use it in connection with something like Papyrus, for instance. For engineering, it's not a good description because it's complicated and for analysts in a dry environment. In the agile environment, the analyst focuses on use cases or creating user stories for support. It's not user-friendly for them. It is not not user-friendly. But in general, it's complicated for very simple activities. For example, when we know and use Enterprise Architect well, and we can use all these features for such analysts, it's no problem to use it. But new analysts, modern analysts who work in a giant environment, don't want to learn how to use Enterprise Architect in the environment because they think it's too complicated and has too many functions. There should be ways to find what you need easily and use only the functions relevant to our project. For example, if we're just analyzing a small project, we don't necessarily need tracing or tons of diagrams. We've also had issues installing the Proof of Concept (POC) for Sparx Prolaborate. Many people found it overly complicated and needed improvement. Overall, user-friendliness is a major pain point for Sparx in my company. So, Sparx could look at similar products with a more user-friendly interface in the industry. There are people who use Enterprise Architect religiously, and there are others who prefer simpler tools like point UML for just drawing diagrams. It depends on the needs of the project. Another point for improvement would be enhancing the versioning of models. Although there's a way to establish timelines, it's not as powerful as it could be.

The dashboard and connectivity could be improved.

The Portfolio Management features can be added in the next release. As it helps you to manage more portfolio of projects and architectures of cost projects on a portfolio level. This would be an important feature in the next release.

The modeling tool is targeted toward a sophisticated user.

One room for improvement in Sparx Systems Enterprise Architect is that it's not very friendly. Another room for improvement in the tool is that it doesn't enable you to import the metadata from a database very easily, so reverse engineering of a database was very difficult. Its database modeling and entity-relationship modeling functions need improvement. An additional feature I'd like to see in the next release of Sparx Systems Enterprise Architect is the re-engineering of databases into logical models, if that feature's not yet there, because I haven't checked.

When the model is large, it is a bit slow to render. Also sometimes it is difficult (selecting and holding it) to move a single attribute from one entity in the diagram to another. At times check-in and check-out procedures are slow. Navigating from EA Models to Business Process Models becomes a nightmare if the repository is heavily loaded. I have worked on some other tools which work pretty well in this area. SPARX Systems should consider improving on this point in their upcoming releases.

It can be improved in the area of shared documentation. The idea is that the architecture tool can call back to an enterprise asset, pull that information, and link that as a sub-artifact. Shared environments are a little bit tricky. Looking at it from an enterprise perspective, there should be a much better shared environment. I've got multiple people in different business units. Once I model an object, others should be able to reuse that same object multiple times. Currently, it's a lengthy process to set that up from scratch.

The model integrity is not the same as MagicDraw, so MagicDraw is a bit better as a tool. The integration could be improved.

The solution could improve by having more aggressive working sessions with other product vendors. It would be a benefit to have demo sessions where users of the solution can ask questions to product experts and receive answers. For example, people who are struggling with the initial setup. The team setup is very crucial for the success of a tool like this. I don't know if we looked into it properly. However, if some part of it, such as user setup can be automated, then it would be great for adoption. In a future release, they should improve portfolio planning.

It would be beneficial to incorporate features like document management usage of video models or PowerPoint visuals that you can import and easily use, instead of having to buy extended modules. When collaborating with other people, it needs to be more user-friendly. I cannot get businesses to use enterprise architect as it is too complicated for them.

I have found the solution is lacking options. The general usability needs improvement and more compliance to CML definitions.

I check the Gartner report pretty regularly and I see that Sparx keeps on going down in the rating. The Sparx EA needs to catch up. I believe the one reason is the strategic architecture. Also, other tools might work better for those high-level executives for whom the quality of the diagram is very important. Sometimes these little differences set a product apart.

Their technical support is not good in India. I wrote to them because I had a question, but I never got an answer. So, I just left it behind.

The tool is, to some extent, clumsy and in some areas slow (especially on mid or low-performance workstations). Quite a lot of data entry would be very tedious if you could not develop your own automation or data entry tooling (or have it developed for you). The automatic creation of reports based on the model elements could be improved and overall the diagrams could be more beautiful (or more visually appealing content could be added) to the toolbox. The wireframing support could also be improved and the roadmap capability is not ideal.

From a practical point of view, we need speed and reliability for creating a model and doing some really meaningful tasks such as application landscape, refactoring, etc. These are two primary criteria. Sometimes, when you import something, it creates the object duplicates, or it allows you to do something that you're not supposed to do. For example, validation is missing. This could be frustrating because when you work at a high speed, you need to come back and start fixing things that the tool allowed you to go with, which is not quite good. So, there should probably be some internal mechanisms to advise you about what you're doing and what is probably not the best idea. For example, you can do many things with ArchiMate, which is modeling language, but people can interpret many things incorrectly. They start modeling and then realize that it is not a good idea. So, it is not the tool itself. It is probably a combination of the modeling language and the tool that validates it. It would be very good if validation mechanics are embedded in the tool to, at least, advise people that a particular thing is allowed to be done in this way, but doing it would also mean something else that you may not want. The languages themselves are not perfect. In a large company, you have many people doing the modeling. If they interpret things differently and the tool allows them to do that, then you would have to do some rework.

Its documentation is not 100% perfect, but it is good enough. It is powerful, but because it is so powerful, it is sort of arcane. It took me a while to figure out how to use the report generation features effectively. So, it would be really nice if they had a way to make that a little bit more interactive and a little bit more straightforward. Something that is a little bit annoying is that the changes that you make to models are instant and permanent. So, you have to be careful with what you do, and you should do a backup or have some kind of a version control scheme in there. I'm used to Word where you can hit Control+Z and undo what you just did. It is not quite that way in Enterprise Architect. There are cases where the model changes, so you got to be careful with what you do before you change the model.

My customers use Bizagi, although I do not. I don't want to change this. What should be improved, however, are the integration capabilities of the solution with Bizagi.

This solution is quite complex to use. It would be nice if the learning curve wasn't so steep. I can't think of any additional features that I would like to see in the next release. There are already plenty; if there were any more, it would only become harder to use.

It could be more user-friendly. The tools could be more simple to use. It's a very complex solution. Because I am a business analyst, I use these tools to manage requirements, and I make models in UML, BPMN, and ArchiMate, and it's complex. In the next release, I would like to see more integrations.

The UI is a little bit outdated. It should be more fresh and clean. The other thing that I would really love to see improve is the roadmap capabilities. They advertise that you can use Sparx Systems Enterprise Architect to do roadmapping, but I don't think that this is really accurate because the roadmapping capabilities are very basic and you can't really do a lot with them. Also, the Veeva Forms Management capabilities, which are built-in, could be improved. They are okay, but they could be much better as well. I would love to see more emphasis on Agile product development within the tool itself. So, if I am managing an Agile project or a scrum project, I would love to be able to plan my sprints within the tool and manage user stories, use cases, and test cases within the tool itself without the need to use any other tool.

Its usability needs to be improved. For non-technical users, it is a little difficult to understand how Enterprise Architect works. Users who are not engineers find it difficult to understand how this tool works. This is something they need to work on. They can develop a BPM model to simulate processes.

One of the drawbacks is that it is oriented on architecture and not user-presented. Also, one of the disadvantages is that it doesn't provide a better representation level for the readers. It is not integrated with solutions such as Confluence or Jira. This is something that is missing in this solution. Because it is oriented on architecture, it is not convenient to use this information for the presentations to our clients. They also have a cloud-based deployment solution and it has a bit more capability to communicate to clients and to the sales team. I would like to see integration with Confluence or any other TRM, and the capability to integrate with the data storage, such as a repository similar to GitHub.

I would like it to be less of a general tool. Currently, it is not a Swiss army knife that can do everything. It is not specialized for our purposes. We are a civil engineering company. We build things. We work mostly in what is known as Infra world in the Netherlands, which comprises objects such as bridges, locks, and water management. We would like to see more focus on such types of projects. It would be nice if it has more specializations. At the moment, it is very generic, and you have to create everything yourself. Our focus is more on user requirement management, which is currently very basic. I would like to see a lot more functionality in this area. Its basic functions for adding user requirements are perfect, but we need more features. Currently, it has limited possibilities for our requirements. I would also like to see better contract management and have it managed in a certain way.

The presentation graphics need to be improved in future builds. It's primarily an architecture tool. Therefore, it's using certain formulas, and they aren't really very useful in terms of presentation graphics for executives. It's an ongoing issue. You do some kind of diagram, you then have to convert it into a Microsoft PowerPoint in order to get a certain look and feel. Otherwise, the design is just too obscure for executives to understand. The product needs better tools for defining report templates. Sparx will generate automated reports based on whatever you select from the repository. It has this templating tool that's very flexible, however, I can't get the damn thing to work properly. It's just not very user-friendly. It's almost like a programming language. That's the thing that we keep coming back to tech support with to say, "What are we doing wrong?" If they offered a better report generating front end that will let someone quickly and easily configure what they want in their reports, that would be very useful.

The fact that you can do a lot yourself is a plus point, but it also becomes a challenge because you need an understanding of the programming languages to get things to work. It becomes challenging for those who are not very good at programming. You have standard reports, but if you want to make your own reports, you have to program it. Similarly, if you want validations rules, you have to take care of them yourself.

Many items are not actually handy, I would say. Many things which should be working - regarding some techniques - are not working and seem to be buggy. Recently, I was trying to perform a task and I was thinking about getting in touch with Sparx about it. However, I started with confirming my concerns with colleagues, and now I would like to somehow report it to Sparx and ask them if the way the task is being handled a bug or feature. For example, there seems to be inheritance between blocks. I know inheritance from programming, and I checked that this mechanism is still working in MagicDrawer. However, it's not working in Enterprise Architect almost at all. Some things which should be inherited are not inherited, or if they are inherited after, let's say, deleting the inheritance dependency that still remains in the target block. It is not logical, at least for a programmer. The solution needs to offer better support for the mobile-based system. Right now, it's not working.

The Business Process Modeling or BPM feature can be improved to make it more interactive and user friendly because it is a tool for technical people. My current use is only for business process modeling notation and putting in the icons etc. You need to take them in as a class, which makes things very complex. Because of this complexity, it is not an easy-to-handle solution. Enterprise Architect is not very good for mockups. We cannot create user screens and other similar kinds of stuff, which is bad. For these things, we prefer to use Axure RP and other similar solutions. They should either remove this feature from this product or provide some kind of connectivity with Axure RP so that people can do better mockups of screens and import them. They need to augment and strengthen the BPM feature, which is the main feature. They need to put in some elements like artificial intelligence and augmented reality. They should look into such features because these things are coming up.

The interface is not as fancy as it is in some other software solutions. Some of the features may be fine but I prefer a different type of interface. Sometimes, it's difficult to read the text when you are presenting. The text is not very large or easy to read. It needs a zooming option. The user interface could be better, it needs improvement. This solution has some limitations from a business perspective. In the next release, I would like to see multi-language support.

For data modeling, it is not very mature when comparing with other data modeling tools. In terms of the workflows, we were initially thinking of having something automated where you have the options to check-in and check-out your data models. This would mean that you can get your data modeling changes reviewed by some of the team members. The option to check-in and check-out option is not available in this tool. We are doing the steps manually to run the workflow that we defined. Even with the changes that the other team members will make, the owners of these respective areas will have to manually identify those changes and then merge them back to the enterprise models. That is what is lacking with this solution, that we have seen so far. In the next release, I would like to see an automated way to check-in and check-out your data models and with the review process, where multiple people can make changes to a model, and with the workflow, everything will be automated where the data models can request someone to review the modifications. This would be everything that is needed in the next release.

Even if there are web-based tools in the Enterprise Architecture tool ecosystem (like Prolaborate), the main modeling application is still a fat client application. For some organizations, it is still a concern and a significant disqualification criterion for adoption. The capability to model and analyze while maintaining coherent traceability within different variants (variations or versions) of a future architecture has been greatly enhanced in the recent versions of Enterprise Architect. It requires a very mature, systemic, and methodic approach that is not easy to grasp for junior modelers.

It is a good affordable that is actively evolving, I think the modeling of activity diagrams could be optimized - currently, they insist on you specifying whether a connector is a control flow or an object flow for instance. It is a minor point, but since this sort of diagram is popular in that it affords both the chance to effectively constrain the model whilst leaving freedom for the next stage in the dev process - which key in good design then it should a high priority to optimize this rather than waste resources unnecessary 'bells and whistles'? There are several little things they could and should optimize. But the platform is good and could be the base a whole tranch or really useful features. for example: to be able to easily run code set up in unit tests to reverse engineer specific code blocks to yield sequence/activity diagrams, would be really useful when as a contractor you have to 'firefight' the design from the code. Personally I would like to see the database normalized better. It's really just a data dump whose business rules are contained in the front end client code - it is way way way off 3nf. Because its easy to create diagrams one needs to be vigilant on the housekeeping of orphaned fragments - I have written my own scripts to do this, may they are available now. I don't make much use of the traceability Matrix, yet that should be a feature that I should use if I could see it made it easy to ensure the traceability of ALL the design to the code (completeness) However, it works. It’s good to use and it’s affordable for a single contractor. It has REALLY helped me. It is a good product and I am sure it will only continue to improve.

The product is good. When I'm trying to do something specific for a part of project documentation, it's hard to get it figured out if you don't use it all the time. It's such a massive tool, it's hard to figure out how to dig in and get to the documentation where you have to be to get some idea of what to do. There are not a lot of examples that I'm aware of to be able to do that. The user interface is going to be difficult because of all the things we have to do when we're doing design and trying to figure out use cases and stuff like that. Tutorials would really help out where you could just start to pick it up and say, "Okay, for this kind of thing, I'm going to go through this tutorial step by step and get it completed." They don't have that. There's nothing wrong with the tool. It's how to use it.

The documentation could be better. Where I work, we speak French and we don't speak English, so we don't have anything in French. It's perfect in English, but we need something in French.

The solution is difficult to learn. The documentation needs a bit of improvement. What I find is that when I'm trying to do something specific for some part of a project, in terms of documentation, it's kind of hard to get at figuring out if you don't use it all the time. It's hard, because it's such a massive tool, to figure out how to dig in and to get to the place you need to go. The documentation would give you some idea of what to do. There's just not a lot of examples that are fully baked that I'm aware of. The user interface is difficult, however, it was going to be difficult anyway, because of all the things we have to do when we're doing design and trying to figure out use cases and stuff like that. What I'm getting at is more tutorials are needed. You should be able to just to pick it up and say, okay, for this kind of thing, I'm going to go through this tutorial step by step and get it completed. And I don't see that as an option for getting to know the solution.

I think the solution could be improved visually. It's not as visually appealing as other solutions. There is a difference between being focused on a technical audience or a business audience. A business audience requires a greater effort and good visuals. Integrations with SharePoint, for example, could be beneficial. If you are using on-premise installation, the sharing of your artifacts with people that don't have a license, can be somewhat complicated.

This solution should have better ease of use for the uninitiated.

Using EA involves a steep learning curve if you want to understand its capabilities and functionality. Providing more detailed information about how to configure and adapt EA for consumption by users with less technical knowledge or experience would be helpful. Plus, provide online training that covers the basics of as well as more advanced topics. An introduction on how to do the basic configuration for the non-technical users would also be of benefit.

Their business model does not include consulting services but this can be a weakness in some cases. Companies that include or even highlight the consulting arm of their companies can easily create a situation in which the on site staff become dependent on the vendor consultants and are unable to progress without them. Knowing which button to select or which icon to click on is one thing but understanding how a tools like this fits in to the overall methodology takes an in company investment in integrating the tools and software process. I've worked for most of the big name vendors and I know first hand that the only real way to learn is to roll up your sleeves and start using the tool on a regular basis. Training and instruction are very available for Sparx via webinars, videos, pdfs, whitepapers and email support which I have always found to be first rate. One cannot use these tools using the cup o soup paradyme, "just add hot water". If you want that , stick with Visio,

The platform has support for Windows and Mac, but not Linux. It would be nice to have it supported on the Linux operating system. The product is quite heavyweight, but if you want to do many things with it, it's like a Swiss Army Knife. If they move in the direction of a cloud-based version then I think it is a good idea because people do not want to install so much software. I have seen it with several IDEs, where it is easier to use cloud versions instead of keeping local versions up to date.

Sparx Systems Enterprise Architect is okay. I don't see anything that can be improved at the moment. For me it's perfect. There used to be that feature in ArchiMate Modeling of Enterprise Architect called inherited relationships. The tool used to be equipped with this functionality of showing those relationships and now it's gone. It would be nice if it comes back. I would like to see Sparx Systems Enterprise Architect more user-friendly and intuitive. This would be great. The product is okay for the experienced user, but for the entry-level people, it's somehow hard. They cannot start without one hour of introductory training.

I think that collaboration can be better. It isn't bad but it can be improved. It would be nice if you can have interaction with other users, in a community of some sort, where you can have a discussion and frequently asked questions. It would be best if you don't have to send a text and wait for a measurable time before getting a response. Some of the attributes do not exist for models, and you have to customize in order to have them. They are supposed to be standard attributes. The entire interface can be improved.

* Documentation generation needs significant improvement. * More rigorous support of the ArchiMate rules would be a great improvement.

Look and feel.

A better deployment model for the enterprise without relying on HKEY_CURRENT_USER in the registry.

* It should be made Windows compatible. * It should have a more appealing UI with zoom in and zoom out features. * It needs improvement on the communication part to stakeholders.

The reporting needs improvement.

It is weak in regards to data modelling. It does not provide logical or physical data modelling or the ability to migrate from conceptual to physical and back. It is limited in its reverse engineering capabilities. It uses a UML-based modelling technique which does not provide sufficient control of metadata and standards. There was no way to implement data integrity and referential integrity constraints. I do not even think it has the domain concept. It does not provide the package concept, so model elements can be inherited by other model elements.

Improve the following: * Report builder * Diagramming capabilities * Connector routing and layout algorithms * Inconsistent UI elements must be tidied up with the toolbox gaps removed. * Alignment and button sizes need to be made consistent. Recommended features would include: * A refreshed diagramming canvas (even outsourced to a third-party diagramming vendor, if necessary). * Replace built-in report writer with direct integration to Microsoft Word. * Improve dashboard "Getting Started" guidance to direct modelling activities according to "what" I wish to achieve, e.g., "I wish to generate a Platform Application Map," and all the most appropriate/popular types of diagrams are made available. * Improve support to intuitively generate hierarchical navigation diagrams.