EA for the masses

[skiwi]

An EA user group is planning on a discussion on

Enterprise Architect has a footprint across so many organisations and an installed user base of over 250,000 seats, yet we see so many examples of organisations that only use the product in a limited way. What are the barriers to wider adoption? Why are people confused and often scared of adoption? What’s the message we need to give Senior Management and Executives to get commitment to using Enterprise Architect more widely and realising its full value?

I thought it might be useful to get some feedback from the community on this very subject!

[skiwi]

Naturally I have some thoughts on the subject to kick start the discussion.
Perhaps these are a little provocative, if so, tell me.



1) EA is a powerful tool, with a stack of features, but at the same time it has many quirks, inconsistencies, and minor bugs that are outstanding often over long periods of time. I base this assertion on the evidence from the EA user forums. It is hard to recommend EA to a wide range of users exactly because of this. What is EA planning to do, and by when, to address all the minor issues that appear to be neglected in favour of adding significant new functionality. In particular what is known as the EAUI (EA unique user interface) seems to me to be a major drawback to widespread adoption.

2) One of the potentially most useful features of EA is generation of documents from models, yet this, and the quirky EA RTF editor, is an area that newbies have by far the most problems with, and even advanced users cannot generate profession quality documents out of the box - they generally need tweaking, restyling, and polishing afterwards. The EA RTF editor does not adequately support the range of style formatting options people are familiar with and expect from MS Office, and generated documents override user styles supplied in the underlying (normal.rtf) style document, and generate extra white space. Additionally it is was noticeable that the most recent release ( document generation broke many documents. What is EA planning to do to improve document generation usability and quality?

3) EA is a tool generally used by groups of practitioners, yet has poor to non existent support for sharing external resources such as spelling dictionaries, rtf template, EA template, and more, or internal resources (as found in the resource tab (such as document templates)). Should EA perhaps focus on fixing a number of usability issues and hold off on major functional advances?

4) EA is a modelling tool, as noted above, yet a major expectation of it is to be able to produce diagrams. It is often (inappropriately) compared with Visio (a diagramming tool), yet if it is to be adopted by the masses I believe that diagramming has to be significantly improved. In particular the fact that connectors and similar objects are "second class" citizens in diagrams (see the forums for this and many many more issues). I believe it is acceptable to expect EA diagrams to look as good as Visio diagrams.

5) Sparx have striven to provide a stable environment for users (and their scripts) by freezing the underlying data model (which by default install is on an access 97 database). Based on forum feedback what is EA planning to do to release the logjam and update the model which will assist with fixing some of the issues now recognised to be related to that data model. These range from columns that are too narrow, inconsistencies and kludges in table usage, to differences between schemas on different database types.

[beginner]

My simple experience is: people do not speak UML. At least they do not speak it well enough. In fact the average (from my stomach) is less than 10 percent of IT people which understand UML basics good enough. And the number of architect with excellent UML skills is again less than 10 percent of these. Any UML tool is doomed when the basics are not know by its users.

Another aspect: Project managers rely on Excel instead of model data. There are only few that see the potential of a model. I always hear: we produce the EA documentation after the project pressure is gone. I must not comment on this.

Finally: They (management) usually don't know what they want to achieve with a model. Reasons for introducing tools like EA are from gossip about re-use and potential time saving. Ha!

b.

[fwoolz]

Some additional thoughts:

1. EA is attractive to individual practitioners largely because of its price: they can either afford it themselves or their organizations are happy to purchase it for them since it's dirt cheap compared to many other modeling tools. (Plus it is still a good tool AFAIAC, gripes notwithstanding.) If an organization is really looking to go whole-hog into model-driven  development, they are probably more comfortable going with an IBM, say, than a small firm like Sparx.

2. I suspect a lot of work by individuals and small teams is "UML as Sketch" that only draws on a fraction of the power of a UML modeling tool. Expensive tools like IBM's are overkill. (EA is also overkill in such cases, but it's affordable!)

3. Large firms/ large teams will almost certainly do some benchmarking/ evaluation of several tools before making a purchasing decision. EA's quirkiness will invariably surface during such a process; this would, I think, tend to cause firms to shy away from using it as a heavy-duty production tool.

4. Sparx may be perfectly happy with the current situation and their market share; if so, the incentive to making a major overhaul that would result in the requisite reliability improvement just isn't there.

[beginner]

2... (EA is also overkill in such cases, but it's affordable!)

Cheap overkill. Nice thought.

4...

Yes. That's a fact. There's only one true communism and it will last ever since.

b.

[Van Lepthien]

1. The EA model is not consistent with the UML model, nor (my specific headache) is the SysML add-on consistent with the SysML specification. I am having to write my own audit code to check that the diagrams and relationships are consistent with the OMG UML and SysML specs. The mapping between the OMG models and the EA model seems to be kludged rather than well though out.

2. As mentioned by others, the RTF documentation generation leave a lot to be desired. This may be an intrinsic problem of a generalized reporting tool, but the problems are exacerbated because the tool is based on the EA model, while I need reporting that reflects UML and SysML models. The inventive (is that a better word than kludge?) mapping between the models occasionally gives some confusing results in the reports.

3. Lack of visibility of bugs and bug status is especially irksome for those of us responsible for supporting a group of EA users - partially because more users turn up more problems. If there were a bug list available, it would save me a significant amount of the time I take to see if a problem is reproducible (and definable) before I submit a bug report. I suspect that such a list would save a lot of work on the Sparx side, too, by reducing the number of duplicate bug reports submitted. What would also be nice would be a feature that would allow voting on the importance of getting particular bugs fixed.

4. Use of antequated and proprietary technology. The EAP files are in Access 97 format, the Automation Interface  documentation specifies "Microsoft Visual Basic 6.0, Borland Delphi 7.0, Microsoft C# and Java". Thank goodness for Java, but VB6 and Delphi7 have been dead for years. If I need to be writing auditing and reporting programs because of model quirks and RTF reporting inadequacies, I might want to be able to write those in something like Python. If a company bought EA because it was a relatively inexpensive option, it probably doesn't want to shell out the cost of a Visual Studio seat if it doesn't have to.

Van

[fwoolz]

Van L.,

1. The EA model is not consistent with the UML model, nor (my specific headache) is the SysML add-on consistent with the SysML specification.

I wholeheartedly echo your sentiments regarding SysML.

3. Lack of visibility of bugs and bug status is especially irksome

See http://www.sparxsystems.com/cgi-bin/yabb/YaBB.cgi?num=1277671922/1#1.

[Nicole Tedesco]

A year later, looking back, not much has changed.

I am finding great use in tracking enterprise application and infrastructure portfolios.  In the end, I have increased my productivity by using EA as an object query system of sorts.  I import applications from "master application lists", augment with anecdotal information, map applications to business domains, map changes to requirements and feature sets, and so on.  Upon request, I can always use the traceability feature (or plain old Access queries) to find one or more applications in the system that might do what someone needs (especially during acquisition activities).  I can data mine to generate some implied relationships.  I can write external tools to suck in data of all kinds in order to augment my growing database.  (You need to move out of the "relational" concept of queries in your system and provide queries more suited for the directed graphs we create.)

My diagrams are rich with information.  Shape scripting does wonders for automatically generating useful, reusable visual metaphors, though the shape scripting mechanism itself can use improvement.  I do wish however that diagramming abilities were strengthened.  I need to improve readability because, frankly, results from EA are a bit sloppy.  I need better controls of stupid things like arrow heads and labels.  A better rendered would be really nice, better clipboard conversion to Enhanced Metafile format... boy, it would be a killer if EA would allow me to assemble and manage and publish PowerPoint decks directly from the tool.

In the end, I need to communicate complex ideas.  First I have to assemble much information.  Then I have to organize it to make sense.  Then I must design a communication form that others will understand, UML or not.  (Applying images to objecs and shape scripting are very helpful, thank you!)  Remember, Edward Tufte be damned, the world's corporations live, eat and breathe PowerPoint and Visio.  How is EA feeding/competing against those applications?

In some respects EA must become "PowerPoint with benefits" and "Visio with benefits".  Perhaps EA doesn't have to be Visio or be PowerPoint, but perhaps think of those tools like code and provide "round trip" experiences between them.  Perhaps concentrate on round-tripping between the Microsoft Office XML file formats; doing so could prove to be a potential data mining coup for EA—think about it (data mining of existing information would be a real market differentiator).

Oh yes, what ever happened to "frequent release", frequent bug fixes and so on?  I used to tell people I liked the rapid response model of EA, but that seems to have fallen by the way side.  You are losing touch with your user base.

[Geert Bellekens]

Nicole,

The reason we haven't seen a release since november last year is because Sparx is working on the next major release (v 9).
I have a feeling the first beta release won't be long now.
From what I picked up here and there there are a few major improvements coming, including some which relate to the diagramming aspect.

Geert

[Nicole Tedesco]

Is Sparx discussing v9 at all?  Can I read anything of their plans?

[Geert Bellekens]

Not publicly, but we sometimes get hints from the Sparxians when they reply in this forum.

Geert

[eliteisv]

because the program utterly unintuitive the only reason will eventually purchase is for inconix support and code generations other wise my $600 would go to casecomplete which has a logical work flow
I"LL end buying the 650 bundle for multimedia tutorial

[Thelonius]

Is Sparx discussing v9 at all?  Can I read anything of their plans?

Quite honestly - this is becoming a worrisome portent for professional users of Sparx. If not a "fatal flaw".

Other vendors:

(1) Provide limited information about product strategic roadmap / directions / future functionality - with all appropriate caveats attendant (no guarantees, express or implied, etc.)

(2) Do keep the community of bug reporters and feature requesters "in the loop" in some meaningful way. Sparx refuses even to acknowledge when a member of the community submits a defect report. This arrogant rudeness boggles ones' imagination.

(3) Prolonged inability to maintain product documentation in synch with successive upgrades and versions - indicative of immature comprehension of the principles of software configuration management

(4) Unwillingness to promote strong third-party partnerships and assume some degree of accountability and vendor leadership in ensuring that functionality that is not strongly delivered as core product functionality will be shared out to capable third-party developers - prime example being reporting - which has been utterly tragic for as long as I can remember - and would have been an obvious candidate for farming out to a strong software engineering third-party. The TOGAF and ArchiMate "add-ins" are ... how should one describe them ... well, never mind ...

To a loyal supporter and user of Sparx technology for many years, this proprietary, closed, mode of dealing with their customer community is worrisome.

If "Version 9" doesn't enter the world with comprehensive, new, strong functionality, new architecture, and a better, more consistent user interface, I fear for the future commercial success of our friends in Creswick. (Think "Plan 9 From Outer Space".)

Version 8 was not a major release - it was more than a point release, but it did not warrant a new version number. If the same thing happens with "Version 9" ...

The market for UML, TOGAF, ArchiMate tools has speeded up, expanded, and gotten much richer in terms of competition and capability over the last three years.

Time for Sparx to pull a huge, instantly loveable, cuddly rabbit out of a massive Akubra.

:-)

[Nicole Tedesco]

Thelonius, Right on!

At the very rapidly expanding North American bank I am currently working at, I am at this very moment in a position to recommend an enterprise modeling tool.  I have recently turned EA into a decent productivity tool with a few tweaks I developed on top of its relational model. I am on the verge of either recommending wider scale adoption of EA, or something else...

I am not happy about the level of support Sparx has been providing over the last couple of years.  In the beginning, I was impressed not just by the price-function balance of the product but also by the level of support Sparx provided, specifically in terms of community integration.  Now, however, I no longer recommend them on the basis of their support model in fact I tend to make caveats about the opposite,

"EA can work for us, but their current support philosophy is really bad."

If I can migrate my enterprise-level work to MagicDraw, I may recommend that as an enterprise-wide modeling solution.  I am willing to wait to see what EA Version 9 delivers.  However, I must see:

• More than just a point++ release (to Thelonius' point)
• A much better commitment to community feedback, returning to their roots
  
• Increased visibility into the product road map, and not just hearsay from a few friends who talked to some other friends who overheard some other folks chatting in a coffee shop in Canberra...
• A return to regular bug releases

This kind of behavior from a product company can a sign of financial distress.  The decline of WordPerfect comes to mind.  This could be a sign the product may not be long for this world.  (I really, really hope I am reading the tea leaves wrong.  I am including this titillating suggestion mostly to see if anyone at Sparx is listening, anyone at all... Bueller?  Bueller?)

[eliteisv]

I think EA could benefit from a well organized ribbon I realize that possible since it's programed in Delphi