in
Forums
Blogs
Files
Devexpress.Com
ClientCenter
Support Center
DevExpress Channel

Gary's Blog

The eXpressApp Application Framework, Project Mojave and Show-stoppers

The provenance of the word show-stopper dates back to the music hall days when a performer’s act was so good that the show was quite literally brought to a stop by the audience’s applause. Of course, time moves on and the word has taken on an altogether more sinister meaning, with it now conjuring up the idea of a fault so catastrophic as to bring the use of something, or the sale of something, to a premature end. A good example of this is the Millennium Bridge, the most recent bridge to be built across the Thames in London.

The Millennium Bridge is a foot bridge crossing the Thames at St. Paul’s. It opened on June 10 2000. Now every structure has a resonance frequency, a frequency at which it will oscillate at maximum amplitude. Unfortunately for the Millennium Bridge, it’s resonance frequency was close to the natural footfall of the average person, not good on a foot bridge! The result was that when people walked on it, it swayed, as it swayed people naturally fell instep with the swaying motion, which amplified the swaying motion and so and and so forth in a feedback loop. Before long the bridge was swaying wildly and it was finally closed to the public on June 12 2000 to enable remedial engineering work to be carried out. A true show-stopper in the modern sense of the word.

Of course we are more familiar with the term show-stopper to mean a fault, or a series of faults, that render a software product unusable. The most stark example of this at the moment is probably Microsoft’s Vista operating system. When Vista shipped it received some bad press and in this age of the blogsphere echo chamber that negative press was passed on and amplified, to the point now where users who have never seen, let alone installed, Vista think it is a terrible product and will quite happily tell anyone who will listen all about it’s failings and short comings. This negative impression of Vista could become, not so much a show-stopper, but something which could seriously retard the sales of Vista.

It’s hard to get out from underneath the snowball of such negativity, fueled as it is by the echoing nature of the blogsphere, but Microsoft have attempted to do just that with Project Mojave. In this experiment a number of people, who had a pre-existing negative view of Vista, were shown a “new” operating system called Mojave. They all thought the new operating system was amazing and wanted to know when they could upgrade to it; only then were they told that they had, in fact, just been shown Vista. Whether initiatives like this are enough to combat the negativity surrounding Vista remain to be seen, but it is certainly a good start.

To a certain extent we suffer a similar problem with XAF, not around the quality of the product, everyone I show it to loves it, but there is a perception that XAF has a steep learning curve, and that learning curve is putting people off taking up the framework. Well guess what, those people are not wrong, they are just not completely right. Now XAF has a learning curve, there is no doubt about that, however it is worth noting that every framework, which increases developer productivity, is going to have a learning curve and the more a framework does for you, the more productive you are, the more there is going to be to learn.

So, now that we have established that there is no getting away from the learning curves associate with an application framework, what are DevExpress doing to help? Well we are tackling the problem on a number of fronts. Firstly, there is the documentation, then there are the videos and lastly, but I hope not least, there are evangelists like me. If you have an issue regarding not being able to work out how the framework works in a particular area, then please feel free to get in touch with me at garysATdevexpressDOTcom. Trust me, the view from the top is worth the effort. :-)

Technorati tags: ,
Digg This
Published Aug 06 2008, 01:01 PM by Gary Short (Developer Express)
Filed under: ,
Technorati tags: XAF, Videos

Comments

 

Jascha said:

Hi Gary,

I am a committed xaffer (as I believe we are now called) and have battled through most of the learning curve - one which I would agree with you is worth the effort. I think there are a few things DX could do to improve the situation (all intended to be constructive):

There is the perennial cry of "improve the documentation". Now I accept that documenting such a beast is a difficult task and not one that DX has done before so you have a learning curve with respect to how to document frameworks as opposed to component sets. IIRC I have seen DX reply to this cry with "tell us which bits need improving". With all due respect I think that is a bit of a cop out and you cannot expect a comprehensive answer from an individual user. Perhaps providing us with a well-publicised place that we can register what we consider to be areas in need of improvement would allow us to slowly build the picture as we encounter them. Here is a example of an area that could benefit from some TLC - controllers are at the heart of XAF and a large portion of the learning curve. Sub-classing controllers is an important aspect of using them so why does the documentation not contain references to their protected members?

You probably also have to accept that there is only so much you can achieve with documentation alone. To my mind, the next biggest potential area from improvement is the reference application(s). IMO MainDemo just does not do XAF justice. Bear in mind that, given that XAF has a learning curve, you need to convince newcomers to invest the time it takes to properly evaluate it. Apart from de facto usage that will take time to build, the next most compelling evidence has to be a comprehensive and cohesive reference application in a problem domain well suited to XAF. While XAF can produce such a thing, MainDemo does not embody it - ask yourself how many of your target customers are thinking of writing a partially functional Outlook lookalike? IMO, the impression MainDemo gives is along the lines of "XAF can create simple applications but if XAF was capable of more that that then surely DX would have invested the time to show me - therefore it can't". Additionally, if you were to complement a decent reference application with a set of how-we-did-it videos then it would be that much more accessible and, in itself, would represent a useful learning aid.

Finally, I think DX would also benefit from someone unfamiliar with XAF (it was once upon a time going to be Julian) actually building a production quality and scope application with XAF - IMO an awful lot would be gained from the exercise including what the reality of the learning curve is and where the documentation is lacking for someone that is new to the party.

HTH,

Jascha

August 6, 2008 10:58 AM
 

Mormat said:

Per Jascha, good point

August 6, 2008 11:07 AM
 

Gary Short (Developer Express) said:

Hello Jascha, and thanks for your comment. So, we are called Xaffers now eh? Cool, I never knew. :-)

To follow up on your comment, I don't believe us saying "tell us where the documentation is weak and we will fix it" is a cope out. I mean, no one, let alone us, sets out to produce documentation that is not fit for purpose, so if that is the result, then clearly, we need to know where we have failed you, simply saying "improve the documentation" doesn't help us; it is tantamount to an athletics coach telling an athlete that to win, he must simply run faster. However, I agree, it is also not possible for one single user to provide us with this information. Perhaps we need a collaborative evironment in which everyone, who feels the doucmentation is lacking, can point it out to us. Perhaps a wiki site where users who have already solved the particular problem can add their advice, a sort of living documentation for XAF?

With regard to a reference application, do you feel one big monolithic application, demonstrating all that XAF is capable of, is the right way to go, or do you think that several, smaller applications, surfacing a particular element of XAF, would be better?

Regards,

Gary

August 6, 2008 11:38 AM
 

Robert Teague said:

I don't intend to purchase a Universal subscription in order to get it. It is too expensive for a lot of individual developers, like me.

August 6, 2008 1:13 PM
 

Bill Hazelwood said:

Although still new to XAF (does that make me newbie Xaffer or perhaps Xaffer Jr???), I love it. I can't imagine trying to develop my app without it.

Having said that, it seems to me that that we are left to wolves when it comes to the learning curve. The online documentation is good and the help forums are great, but only (I believe) scratch the surface. There is a feeling of being disconnected. I crave interaction. Nothing beats 'in person' training. I have mentioned it before, but will say it again. Xaffer University!!!!!

Gear it towards people that have already read the online docs and done the demos so that there is some level of competence before beginning. Keep to the major sections (XAF,XPO,UI, other) and have different levels. Expand later to more detailed areas. Webinars and classrooms are great. Microsoft does a great job with their interactive (and archived) Webcasts. In person classroom enviroment is expensive, but then again, we could all use for a week or two in Vegas!!!! ;)

Other than that, it would be great to see (graphically, I like the 'visual') class maps and hierarchies, more detailed descriptions or links to the online docs while in the XAF editors (Model.DesignedDiffs.xafml).

I am very impressed with what I have seen (can't believe that I got all this functionality for that price!!!!) and look forward to a long and prosperous future with DevExpress.

XAF integrated Mobile Device projects would also be great (xxxx.Win, xxxx.Web and xxxx.Mobile).

August 6, 2008 1:31 PM
 

Keith Lawrence said:

I think Jascha nailed it.  

I installed XAF just 4 days ago infact, and the first thing I did was to look at the MainDemo.  I was less than amazed at what it showed.  Fortunalty I had watched some of the videos on XAF already or it would be on my uninstall when I get time list.

I think that MainDemo or the Better Ref App that DX will produce :) should not be a part of the main Install but a link to a page with a Video Series on how the App was made.  Then warn before downloading the App that the Videos need to be watched first.

August 6, 2008 1:39 PM
 

Norbert Kustra said:

I agree with Jascha - I think we really need a good reference XAF application. Few months ago I tried to use XAF for my new project, but there was lack of information for me to do that. I am not very experienced developer, so I often look at DX demos how they do some things. Unfortunately neither documentation nor XAF MainDemo was not enough useful for me to come through. I have purchased Universal subscription but still not able use XAF :-(

Regards

Norbert

August 6, 2008 1:50 PM
 

Jascha said:

Hello,

Xaffer is not quite in the Oxford English Dictionary yet but who knows...

Ok, apologies for being a bit harsh calling it a cop out. My point was that, from a client's perspective, being asked where the documentation is lacking feels like being asked to write a long essay where the facts are buried in the (in my case) hazy memories of an extended period of development. I should perhaps be more disciplined in noting deficiencies but a place where we can register them easily would make it more likely to happen. My thinking was more of a repository for documnetation areas in need of improvement rather that us actually writing the documentation but I guess it could be a combination of the two although you would have to think about how to keep it current for the latter. I think something more structured than a wiki might make it more usable - perhaps a specific forum area i.e. divide the XAF area into a number of subsections of which one is for documentation. Clearly, it would need to be actively monitored by DX. On reflection, perhaps a xafipedia is not such a bad idea as long as DX actively edit / maintain it.

My feeling is that any reference application(s) would need to feel substantial in its/their own right to fully convey the scope of what can be done with XAF. Given the investment that this will take, my guess is that DX are unlikely to implement more than one in the short term. Specific elements of XAF can be demonstrated within the application but part of the objective would be to make it real-world so that it sells XAF as a serious dev tool and helps keep XAF focused on what is actually required "out there" rather than what seems sexy to implement in the "development lab". Perhaps a ShowMeTheRelevantXAFAspectsOfThisView_Controller is called for ;)

Jascha

August 6, 2008 2:07 PM
 

Robert Johnson said:

Jascha made a good point about the demo app. I got the impression from it as well as samples that XAF was only intended for simple applications, not real world stuff.

I would like to see several simple as well as complex examples. Perhaps you can start a video series that develops a complex application from start to finish.

August 6, 2008 2:21 PM
 

Welliton Toledo said:

A Northwind sample will be terrific!!!

I'm like Norbert Kustra, I purchased the Universal, but can't use XAF yet.

August 6, 2008 2:43 PM
 

Christopher Andrews said:

Just installed the demo version of XAF recently. The first question I asked myself was "how can I customize the cookie cutter generated UI?" If I'm going to spend $2,000 on a subscription I'd better be able to place a button anywhere I like on the form. Is there any documenation that could walk me through something as simple as this? The closest thing I found was under Concepts > Extend Functionality > Controllers  and the reference section.

Chris

August 6, 2008 8:41 PM
 

Marco said:

A reference application would be very nice, although i think it will need a subsection of the xafiki to determine what has to be in it. IMHO there are tooo many users having varying demands to expect DX to create a reference app perfectly suited to everyones needs. By the way, I think the FeatureCenter demo coming with 8.2 is a good place to systematically register various parts  of XAF. Currently, there is not very much in it, it would be nice to create a subsection for View / WindowControllers (including the various actions), one for the various modules coming with XAF, a subsection about Templates and what you can do about it (placing a button anywhere on the form for example) etc.

As for the documentation, I agree with the contributors before me. I think that at least the following parts could be enhanced:

1) Currently there are 3 main sections I use a lot, Tutorial, Concepts and Task based help, but it's quite frustrating that when I want to know how to access the grid control, it's within Tutorial \ Extend Functionality. When i want to create a custom detail view item, I have to go to Concepts \ UI Construction and when I want to access the scheduler control, it's within Task based Help \ Extend Functionality.

I think that elements covered in a tutorial should not be omitted within concepts \ task based help. Put it in other words, a tutorial should be an extra.

2) The Reference section of the Documentation is by far not complete. There are quite a number of undocumented classes used in XAF (also in code-snippets in the support center).

For example: View.Info is a DictionaryNode, but what does DictonaryNode has to offer me?

DictionaryDifferenceStore is another one.

Speaking for myself, I've been a DX Enterprise user for about 6 months before I made a switch to XAF. I realize when people start with XAF immediately, there are a couple of layers used within XAF which themselves already have a learning curve. IMHO, it's would be good recommended to first play with the various components (like XPO, XtraBars, XtraEditors, XtraGrid) to get a feeling about what they are capable of. Then it's also more clear what can be achieved by a powerful tool like XAF. Maybe an idea for the introductionary week at XAF University?

I think one of the biggest advantages of XAF is the combination of abstracting away a lot of annoying things (like saving settings for all imaginable customizations a user can perform) but still giving me the possibility to influence or even completely replace this functionality.  And yes, al this has a cost, being the learning curve. Although quite new to XAF (about two months now), I know for sure it's worth it.

Marco

August 7, 2008 4:08 AM
 

Steve Sharkey said:

Jaschas comment (and others who have followed up along a similar theme) about a "proper" demo app is quite critical I feel - to be honest the "Line of Business Systems " Case-Study is what brought me back to a second trying out of XAF. (And my subscription being renewed at Universal level).

My work so far is progressing reasonably - I made more progress by forgetting XAF for a while and really concentrating on XPO. I am getting close to completing my objects and moving onto the GUTS of the application and I worry that I will meet another set of delays on the developments.... So my postings will (probably) rapidly increase some time soon.

August 7, 2008 4:12 AM
 

Mike_Grace said:

Jaschas is right. I have no idea whether XAF is capable of creating the application I want to write and I don't want to invest time and effort only to find out it can't. The mainDemo is not enough.

August 7, 2008 10:12 AM
 

Gary Short (Developer Express) said:

Okay, so more demo apps and better documentation, right? In the meantime, if anyone has specific questions then don't hesitate to get in touch.

August 7, 2008 10:31 AM
 

Seth Juarez said:

A showstopper for me in the XAF framework is the inability to use other ORM tools for the DAL (aside from the cost which becomes prohibitive at times for a developer like myself). If there were good docs, plugable provider architecture, and maybe a bit of a price reduction, I would certainly consider the switch.

August 7, 2008 3:07 PM
 

Gary Short (Developer Express) said:

@Seth, although we feel that XPO is an excellent ORM tool, there is ongoing work being done to allow this element of XAF to be pluggable, look forward to seeing that feature in future releases.

August 8, 2008 4:49 AM
 

Adam Leffert said:

- DevExpress, the DevExpress .Net controls, XAF and DevExpress support are outstanding.

   Literally, they stand out in the field of .Net development.  No other set of tools that I am aware of enable

you to get so much good work done so quickly.

- For a professional developer, XAF is definitely worth the learning curve.

- DevExpress should pay more attention to the word "steep" in the phrase "steep learning curve".

The problem with the XAF learning curve is not that it's long.  The problem is that there is a very steep section

in the curve.

Here's how the XAF learning curve goes:

A) Learn the very basics, e.g. create a class, inherit from BaseObject, XAF creates UI for add, edit, delete

automatically.

This part of the curve is fun.  Small effort, big return.

B) Learn to customize ListView's and DetailView's.

Also fun.  Small effort, big return.

3) Try to do something that seems like it should be simple (e.g. change a button caption, hide a button, etc.).

This is the first place that a new XAFfer will run into trouble.

The desired task seems so simple that the developer figures there must be a simple way to do it.  At this (steep)

point in the learning curve, the boss whom you're trying to convince to buy XAF for you is getting impatient.

Yes, once the developer knows how to inherit from ViewController and chase through collections to find

ActionItems it's not very difficult.  But this part of the XAF learning curve is like playing hide-and-seek.  

Once you've found the person hiding, it seems very obvious.

The XAF doc is very helpful, but there are some simple (and not so simple) ways to improve it that would make it

much more helpful.

I'm working on a post to offer some suggestions on how to improve the help doc and the three help db's (Issues,

KnowledgeBase, CodeCentral), but I want to take some time to organize my thoughts, and check whether DevExpress

already has solutions to the problems I'm considering.

I'm not at the top yet, but I agree with Gary ("The view from the top is worth the effort").

Back to work on my XAF app!

Best regards,

Adam Leffert

August 8, 2008 10:42 AM
 

Alain said:

I think that might be is necessary the following:

1. An schema of how each piece of XAF application is called from the main thread of execution, in other words, how is the activation flow and execution of controllers, listview, navbar items, etc. Might be this can help to new developers to understand views, controllers and actions functionality.

2. A decission matrix to select when use or not XAF, based on the kind of application to develop.

3. A document about how to solve more complex issues in software development using XAF, for example, how to work in a multi-user enviroment, when two people(user) try to get and modify the same record from DB, or how to mix XAF with others tools to allow multi-tier development (UI, application server, database).

4. A document with advises of how to test a XAF application, when for example, you change the name of some views, and controllers doesnt work (but not error is raised), because are "linking to old string names" of views.

5. A procedure of how config and use some software version control software like subversion to work many developers over a XAF application.

6. A document of how setup a software factory based on XAF, nothing complex, something to control a team of 4 or 5 developers for example. I think that many potential users of XAF will be small software companies.

My five cents.

Regards, Alain

August 8, 2008 6:36 PM
 

Adam Leffert said:

A few points on requirements for the XAF demo:

- It has to be live on the Web.

There's no excuse for not doing this.  DevExpress ASP.Net controls have live demos on the Web.  They instantly silence any competitor who doesn't have live demos, or has ones that show inferior products.

Is there currently a live XAF demo on the Web?

Also, let the monkeys bang on the typewriters.  If the XAF demo was on the Web, it might throw a few exceptions.  The development team would get emails and fix the bugs.

- Make the dogs eat the dog food.

If the XAF demo is contrived, i.e. not an actual application in use by people trying to get actual work done, it will never get the optimal level of attention and investment.

Make the XAF demo something that XAF developers use every day.  Take Jascha's ideas about a more structured system and create a live XAF app that relates concepts, tasks, classes, questions, answers, code samples, XAF versions, etc.

There are plenty of arguments for and against this suggestion.  Rather than trying to predict them, I'll wait for the group's response to the idea.

- As others have said, we also need a non-trivial demo app for non-programmers.

I suggest a pseudo-accounting app.  I don't know a good name for these apps (anyone out there?) but I've been writing them for years (OK, decades).

They are not true double-entry accounting systems, but they cover the things that accounting systems don't do.  These apps have been a large percentage of my consulting practice over the years.

They are more impressive than CRM apps, because they show the type of data found in CRM apps, plus dollar amounts (Euro's, Yen, ...), graphs, etc.

- Enough wiki.  We need more structure.

Yes, it's wonderful how knowledge can emerge from wiki free-for-alls.  But we have enough of this in the DevExpress doc.  We've got forums, KB articles, issues db, etc.  These are all very helpful and necessary.

Now we need more structure.  There's a lot to say on this issue too, but to summarize, DevExpres should be able to click a button and get a list of the XAF features that need to be documented and have not yet.

If there are too many of these to enter into a db, we really need a db! :)

That's why I suggest the two-birds-one-stone solution of an XAF documentation demo app.

Best regards,

Adam Leffert

August 10, 2008 12:21 AM
 

Jascha said:

"Make the dogs eat the dog food" LOL. Although maybe we can come up with one that uses caviar instead ;)

Contrived is a very appropriate word to use here - a contrived demo app. will reflect poorly on the tool used to create it. Bear in mind that granular demos showcasing features are appropriate for component sets since such products are purchased precisely for that reason - to add features to applications. Application frameworks incur a far higher cost and risk of adoption and consequently their demos must address those issues as well as showcasing the features that are inherent in the applications they produce.

Gary - I am aware I have banged on about this quite enough and will quitely step down from my soap box now. If you intend to look into creating a new demo app. perhaps you should start a new post / thread to conduct a straw poll about what types of applications XAF is being evaluated / used to produce.

Jascha

August 11, 2008 5:24 AM
 

Gary Short (Developer Express) said:

Hey Jascha please don't feel you have to get off of your soap box, we need feedback like this to take our products forward in the market place. Reading these comments however, it's clear that there is a number of things that our users want, things that we just don't have time to produce. All is not lost though, that is the beauty of the community, if the community feel there are things lacking from our documentation and demo applications then the community can provide them. There is already at least one blog outside of DevExpress for example (expressapp.blogspot.com). If you feel you can't wait for us to catch up to where you need us to be, then feel free to start your own blog or Google Group or wiki, we will support your efforts as much as we can.

August 11, 2008 6:16 AM
 

Gilles Muys said:

Gary,

I think that adam Leffert hit the nail on the head with his opinion.  The videos are absolutely outstanding.  However, in the "Blog" video Oliver stated that it would be extremely easy to put the Blog Post into the XAF UI shell.  I have done the video on my own and I am struggling bigtime on how to insert the Blog Post into the UI.  It seems like it should be a very simple thing to do, but it isn't.

When doing the videos, I think many of the users understand the basics, but the UI isn't very easy to grasp.

August 12, 2008 10:35 PM
 

Gilles Muys said:

Another thing that I would really like to have a demo for is a custom Module so that I can implement it in any application.  Very similar to the Windows Scheduler.  I know that it has just been available for web apps, but something that that would be very nice.

August 12, 2008 11:12 PM
 

Adam Leffert said:

Quoting Gary:

"All is not lost though, that is the beauty of the community, if the community feel there are things lacking from our documentation and demo applications then the community can provide them."

This is the right answer to a different question.

This is the right answer to a developer who asks DevExpress to create a demo app for his favorite vertical market (gas stations, movie theatres, etc.).

This is the wrong answer to a developer who politely points out that the current XAF demo app is unacceptably simplistic.

The XAF community needs a way to express to DevExpress just how unacceptable some parts of the XAF offering are, without sounding rude or using hyperbole.

I suggest a very simple system.  Let's rate the fundamental parts of the XAF offering on a 0-10 scale.

It's very important to define what these numbers mean.

10 = The best that is actually found in the real world.

For example, if you rate something a 7, you should be able to provide a URL for a product that is much better.

0 = worthless

5 = any worse and it would start to be unacceptable

It's important not to peg 10 to a fantasy, because doing so flattens out the ratings, making them useless.

We all shoot for 10's in our development efforts and business lives.  If we get 7 or 8 on most issues, maybe that's OK.

If a significant number of respected, reasonable XAF developers rate a fundamental part of the XAF offering as 4 or less, DevExpress needs to have a real plan for addressing the issue in the short-term.

A "thanks for the input, we'll get to it as soon as possible" is not good enough.

A couple of weeks after agreeing that a fundamental part of the XAF offering (code, tech support, doc, etc.) is 4 or less, DevExpress should be able to provide concrete examples of progress in that area.

We want to help publicize DevExpress and XAF, and we want DevExpress to like us, but we have to express ourselves clearly on these issues.

PLEASE chime in with your ratings.

The $2,000 we spent on license fees is a drop in the bucket of our commitment to XAF as a platform for delivering real software solutions to paying customers.

As a side note, I would rate the DevExpress ASP.Net controls as 10's in ALL major areas.

Here are my ratings on some fundamental XAF areas:

XAF CODE

- Functionality of Web apps you can produce with XAF once you learn it: 10

Rationale: There are several key areas of the offering that seriously need work, but I know of no better ASP.Net app framework overall.

- Look and feel of XAF apps: 8

- Cross-browser and cross-platform compatibility of XAF apps: 10

Rationale: Works on IE, Firefox, Safari on PC and Mac.

DOCUMENTATION

- Quality: 7

Good enough to be useful.

I don't care about grammar mistakes.

- Completeness: 3

This is the MOST crucial issue.

Proof: It takes literally dozens of hours to figure out how to do things that every developer delivering Web solutions costing more than $5,000 will have to do.

I have spent months learning XAF, and I am still spending many hours per week trying to piece together answers that should be in the doc.

Proposed solution: As I've mentioned before, I suggest that DevExpress create a database of major issues that have and have not been documented.  This is NOT the same thing as the issues db or the KB.  The issues db lists bugs and Q&A.  There is one record per bug or question.

The doc db would list:

- issues that DevExpress admits are fundamental, but have not yet been documented.

If DevExpress employees had one place to go to find these items, they'd know what to work on next.

- doc items that DevExpress knows are out-of-date, are fundamental, and need to be fixed

- doc items or code that exist and are up-to-date on one platform (Windows or Web) but need to be created for the other

Gary: Telling the community to "let us know where the doc is lacking and we'll fix it" is both a generous offer and a cop-out.

It's a generous offer in the sense that DevExpress tech support's responsiveness is a 10, but if DevExpress takes a 100% reactive stance and only fills in the doc where developers complain about it, the doc will NEVER be complete.

Not "will be complete far off in the future".  Never.

Because the coverage of the doc will be uneven: questions from newbies who didn't RTFM, hard-core requests from veteran XAFfers, etc.

OK.  The bell in my head that rings when I write a post that's too long just went off. :)  I'll finish with a few more ratings:

TECH SUPPORT

- Intelligence: 10

- Responsiveness: 10

I.e. length of time it takes to get an answer.

- Willingness to look at our code samples: 10

- Willingness to provide code samples: 10

Apologies for the long post.

XAF developers: PLEASE list which XAF issues you know to be fundamental and how you rate XAF on those issues.

It's obvious that DevExpress cares about the quality of their offering.

If developer after developer rates a fundamental issue less than 5, I believe DevExpress will fix it quickly, the product improves, and we all win.

Best regards,

Adam Leffert

August 13, 2008 11:28 AM
 

Alfred Ortega said:

Gary

As a person new to DevExpress and XAF I can attest to the steepness of the learning curve. IMHO I don't think that one monolithic demo app is going to help though.  So I'd like to offer a small list of specific things that can be done since I agree with you that general statements about making documentation better aren't really helpful.

Forgive me in advance, I don't know the proper DevExpress/XAF Termininology!

1) Changing the UI of a listview

2) Changing the UI of a individual item editor thing

3) Implement Role-Based Security

4) Business Rule Validation

5) Update the defaults provided by the XAF Wizard thingy to use other DevExpress Controls instead.

For example there is a TreeViewGrid demo for a simple task-list.  Well make it so that the tasks have an taskid/parenttaskid hierarchy and build it using XAF.  Two birds - one stone.

I'd go as far as saying that all (or at least most) data driven demos should come in simple binding format and a XPO or XAF format. Anyway I think that a set of small demo apps that each highlight a particular issue is better than a big "real world" app.

There called "demos" for a reason...

Clear commenting in the code of these apps which points to specific places in the documentation to learn more about a task or method or whatever.  Right now as I'm going through some of the documentation I feel like Winnie the pooh looking for a heffalump.  He's sure it exists but couldn't tell you what it is or what it looks like!  I'm sure there are methods or classes for what I want to do - if only I knew what they were! :-)

August 25, 2008 9:43 AM
 

Tim_Pau said:

Gary,

Our company loves the controls in our DevExpress Enterprise subscription. We are on the verge of creating a new application and are considering using XAF. The framework offers many compelling reasons for its use.

But I (like many others) can't figure out if XAF can build the application that we envision. I spent several hours today watching you and Oliver in your umpteen XAF videos. They were helpful...but I'm still left with the nagging question, "What if I spend hours building my new app only to find that XAF won't do what I need it to do?"

A demo application would be great *after* I embark on the development of the app in XAF. But right now, I just want to know what are the capabilities and limitations of an XAF app.

Here's an idea. Can you create a place where others can post screen shots of their XAF apps? Seeing some screen shots would help immensely. You at DevExpress could start by posting some screen shots.

Thanks,

Tim Pau

September 4, 2008 5:31 PM
 

Garth Henderson said:

Please continue to vote yes on this by tracking Jascha's Suggestion S92325.

www.devexpress.com/.../S92325.aspx

We are a growing community and I appreciate all the wonderful perspectives presented here.  I think that everyone is right.  I especially like the idea of a XAFiki.

September 11, 2008 8:10 PM
 

Yann Duran said:

Yes, yes, yes, yes & YES!

I agree with EVERYTHING that has been expressed here. I was nodding my head so much while reading this set of posts that I thought it was going to fall off, lol.

I feel a bit better now too, knowing that others have experienced the same difficulties as I have.

I ran out of evaluation time, & being a self-employed developer spending a couple of thousand dollars HOPING I would be able to use it eventually was not very palitable.

Maybe I'll have to have another go at evaluating XAF & stick with it a bit longer. I love it, but it's been HARD so far.

Yann

October 17, 2008 2:13 AM

Leave a Comment

(required)  
(optional)
(required)  
Verification code: Required
   
Add
Copyright © 1998-2008 Developer Express Inc.
ALL RIGHTS RESERVED