× The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.



G4

Thanks for your response to my post.

Geeky400 wrote:
Hi Rob,

Thanks for your input and your interesting thoughts on the subject. I spent a bit of time on your website and will go back to it very soon to finish what is an indeed thought provoking read on a rather unusual approach for the IBM midrange world. I will also contact you privately when I have finished going through the info on your website, in order to request access to test drive Erros.
The screen dumps there are somewhat out of date as, for instance, the function key prompts have been removed from the screen and the buttons now have a legend showing their purpose together with the alternative short cut function key. In addtion, where the screen will fit into the browser window, I try to make all screens have the same number of lines to that everything is in the same place on the screen as far as possible. This will often mean writing blank lines to the browser. Constantly changing screens are confusing to the end user.
The scope of my original post was much more humble and did not necessarily extend to reviewing accepted standards like the use of a relational database for example.
When I was a child, I was taught to challenge everything, not necessarily out loud, but to myself at least and particularly where the world at large accepts the status quo just because it is there. I was not taught that change for the sake of change was a good idea. These lessons from my father were very subtle and I was also taught not to be too aggressive if I felt the need to express publicly the need for change. Perhaps my father was less successful in that lesson or may be I changed when I worked for one of the world's leading investment banks (Warburg's) and then for IBM (UK). At both, you were encouraged to be a free thinker and express your thoughts.

As far as RDBMSs are concerned, I found that I could not build the solution that I wanted using them.
I however totally understand and agree with your definition of architecture in that context. My definition of architecture in the previous post was in fact merely describing the design of components (and the layer they belong to) and their interaction. As for my aim with this post, I think I was hopping to get feedback on the pros and cons as well as the best ways and practices when extending modularity to an n-tier model.
I have been using an n-tier model for over 20 years with no exceptions. It does make change simpler but in itself does not bring the dramatic changes in productivity that are required if we are to keep up with our end user needs.

And that is the test. If you asked people in charge of development what level of productivity change would they describe as stunning, many people will say 25%. Where major new systems are being developed, as, for instance, in government and these may take 3 - 4 years and then often fail, in the UK at least, 25% off 4 years reduces it to 3 years which is still totally unacceptable. I believe that a 25 times improvement is what is required. It is possible if we challenge the status quo.
Most shops today use a certain level of modularity, but not necessarily to the extend of totally separating the presentation from the business logic and the data access. I believe that an n-tier architecture model is a worthwhile step forward, but its attractiveness seems to hide a fair number of questions which I hope would be discussed here by people who have attempted such a move or who are currently contemplating an n-tier model in their future developments.
I don't see this as difficult and I am not aware of the disadvantages. Maybe I have missed something.

My original idea was to separate data, logic and presentation in a three tier architecture. This simply seemed common sense. As I was not trained as a programmer, it was a way that I might have a better chance of solving my problems. But I then put the major part of the logic into the database rather than into code, and this simplified the job much further as it dramatically reduced the amount of code that I needed to write, giving me my 25 times productivity gain and more


<Rob>
Can we design an architecture that will make us more efficient?
If enough people are interested, it might be a useful discussion.
</Rob>

I really hope this will lead to useful discussion where ideas and concepts as well as technical options and issues can be part of the debate. Isn't Efficiency a nice ideal?
I certainly think so. I know plenty of IT people who believe that IT is an end in itself and who argue very strongly against me if I say that we are merely a service industry. But without our end users, we wouldn't have a job. If we as IT people can react more quickly and with reduced cost to our users' requests, our companies will have a competitive advangage and they are more likely to retain our services.
Maybe we should carry on on two fronts in order to get the most from both threads:
- On thread purely related to n-tier model in an RPG environment,
I am not sure that I see the point of this as I still say that language is part of the implementation. You can have an n-tier model, whatever your language.
- Another thread on Application Architecture in general where broader concepts could be discussed.
It's not up to me, but I would prefer to start the discussions at a high level and not get bogged down in detail at this stage. If we can agree an overall structure, then we can discuss how to implement it.

There are many other points that I would like to raise about how we might move forward, but I didn't want to put too many ideas in one post. If there is interest, I will post them. First we need to see just how many people want to join in the discussion. I hope that there will be many but I fear there may not be.

Best wishes

Rob Dixon
www.erros.co.uk
www.boarstall.co.uk


As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.