DabbleDB, the not so good …

When launched I really liked the idea of DabbleDB.  The video showing what it can do and how to do it is great too, so I finally decided to open a demo account and create a little test application.  I think I now walk away a little disappointed, but then again, maybe I was expecting too much after watching the demo movie?

Issues:

• Responsiveness; I found the interface to be very slow, adding a column, adding a filter, applying a filer (one of the features is “powerful filtering”), deleting a column, everything seems to be taking too long (manage expectations? it is very fast in the demo video!)
• Field Types; I like most of the field types but it took me a while to figure out the “Link to Entry” field type.  I would have liked to have the option to have the “Location” field type to display the Google Map within DabbleDB.  The “List of Entries” only allows a drop down select box but I would prefer a check box option, similar to the “Choice” field type but then with check boxes.
• Maps; the map feature was something I was really interested in, but it didn’t end up to be the solution I was hoping it to be.  There are only a limited number of maps available and there is no way to display regions or states within a country – other than the states in the US.
• One more little thing; when in a form the “Save and View Next” doesn’t seem to be going to the next row you think it will go to.

Anyhow, overall DabbleDb is a very nice online database and certain aspects are very dynamic and user-friendly but my overall experience is that it is a very slow online application which after just a short time starts to get annoying.

Point Of Sale

I recently started an exciting personal little side-project. For this project I am looking at the business processes and possible IT systems for small restaurants or café’s; i.e. establishments with about 30-40 seats.

In the – small hospitality business – market segment, the majority of successful businesses still seem to operate the old fashioned manual way.

During my analysis of existing systems – and there are plenty of existing systems out there – I noticed that most systems mainly target larger businesses. The majority of these systems claim they can also be used by the small restaurants and cafés. A lot of the systems are POS systems only. The bulk also seem to try to be a generic “retail POS system”. Generic POS systems tend to group businesses like; grocery stores, hardware stores, souvenir shops, hairdressers and restaurants & cafes; even though they have completely different business processes.

I have come across a number of systems that do have some good features; from the commercial systems I would like to mention http://www.redcat.com.au and http://ordermate.com.au and from the open source world I can see plenty of potential in http://www.openbravo.com. Even though it’s a retail POS system I also want to refer to the award winning http://checkoutapp.com.

Why has this niche market not jumped on the new technology?

• No tangible added value
• Relative high initial investment
• Too complex or no good fit in the existing operational model

What is missing from the retail POS to meet the needs of a typical small restaurant?

• Reservation and seating management
• Portable electronic order taking
• Focus on the different stations; i.e. restaurant, bar, kitchen
• Useful reporting

The essential features for a successful system;

• Seating management; manage reservation, quick and easy overview of tables. Note: some systems have implemented a graphical table layout but I don’t think this would add any benefit for small restaurants.
• Order management; allow the waiters to take orders electronically in a similar way as they do now using the paper order pad. Since the Palm and iPaq’s are on the way out, I am thinking iPhone/iPod Touch or Nokia smart phones. Either display or print the ordered items to be prepared at the appropriate station. Note: there are PDA systems to take orders but I find them not very user friendly and they tend be look ugly.
• Billing; create bill and process payments. It would be nice if the waiter could carry their own hip-pocket printer and credit card processor.
• Menu management; allow the manager and chef to alter the menu, change the prices, create set menu’s for large parties and print the menus.
• Stock management; assist in stocktaking and projecting required stock. Why not use the same mobile device as used to take orders to assess actual stock levels.
• Reporting; provide useful reporting assisting the manager with staff rostering, stock ordering and supplying the accountant with the required data.
• Staff management; manage staff shifts, have their shifts even sent to their calendar, phone and/or email. Register staff start and end times and calculate wage.

I am trying not to focus too much on the implementation but I think it’s clear that the portable electronic order pad will be the main supporting device rather than the touch screen POS.

Still a work in progress …

Adobe Flex

I recently bought Peter Armstrong’s Flexible Rails (by Manning) and so far I have to say like the book and I am impressed with the possibilities of Flex. I am interested in Adobe Flex as the User Interface for Rich Internet Applications. Coming from a Client/Server – good old Sybase/PowerSoft PowerBuilder – background I have always been frustrated how long it takes – using java, .Net, PHP or whatever – to put together a decent, good-looking web application and then I am not even talking about how long it takes to make changes afterwards. Compared to the 4GL Client/Server technology the web application frameworks are way too complex for one developer to master. A developer will have to be an expert in back-end component development (back-end frameworks, business logic components, object-relational mapping …) as well as a master in front-end development (GUI frameworks, CSS and more recently AJAX ) to be able to do the same thing one 4GL C/S developer could do.

I am therefore following the much hyped Ruby on Rails, Django and Seaside framework developments and I also keep an special eye out for WaveMaker (which actually has some RAD PowerBuilder roots) and they all seem to head in the right direction. WaveMaker looks very promising – I need to give their latest Mac version a go – the only thing that makes me a bit reluctant is that it is based on java, and I have been on so many java projects which spiral out of control due to its complexity.

The GUI design aspect is exactly where the frameworks were still lacking – unless you are a CSS and AJAX expert – and that is why I am very interested in the combination of Rails and Flex. Flex needs the Flash 9 player (which is not – yet – supported on the iPhone) and in my view lacks a decent WYSIWYG RTE Editor with similar features like FCKeditor or TinyMCE, other than that it seems pretty promising.

Enterprise Architecture

Having worked at a large number of private companies as well as government organizations I still haven’t come across one that has a decent enterprise architecture methodology.

An enterprise architecture – in my view – is important because each organization should have a good understanding of their business processes, all the data entities, all the IT systems and their overall IT architecture. Instead of having a complete mapping of all the business processes and data stores, most organizations seem to focus on IT systems only. The only way this will ever change is when management recognizes that the only way to reduce IT costs and complexity as well as improve business alignment is to have – and maintain – the business processes and data maps as well.

The only time someone usually will look at business processes is during a system development project. The main problem with this approach is that during a system development project – with the time and budget being allocated to the project – there will be not enough scope to look at all business processes and nobody will make sure that this analysis work will be maintained. Another problem is that depending on the business analyst there will be a different level of detail.

There are plenty of frameworks and methodologies that can be used to overcome some of these problems though. The Zachman framework for Enterprise Architectures is a good mechanism to place artifacts in a nice and organized structure. It’s basically a grid to place the information about; What, How, Where, Who, When and Why with different levels of details depending on th target audience (the most important once in my view being the; Planner, Owner, Designer and the Builder), but it doesn’t really tell you how to do the work.

How to maintain all these artifacts? Why not look at how IT projects maintain different environments and versions – Development, System Test, User Acceptance Test and Production environments with each a specific version – because the details of each of those cells may all go through changes just like IT systems do.

Which tools and notation to use is another issue. Traditionally these artifacts are documented in Word and/or a CASE tool, but I find both to have big shortcomings when it comes down to environments and versions.