Bob,
As an application developer, provider and supporter for over 30 years, I sometimes am motivated to comment on development tools coming from 3rd parties in general terms.
In theory, the idea is simply wonderful.
Having been bitten by SB+ and the product that Raining Data/Rocket etc. launched/adopted/dropped many years ago ( so long I cannot even remember what it was called ), we decided to live with the tools provided with the database. Granted, we do use AccuTerm and Wintegrate to “put lipstick on the bulldog” in terms of the presentation layer.
Here are the real life factors to be considered from our point of view:
Of course, being in the brick and mortar POS market space reduces the pressure to be web based for some of the same reasons stated above as they relate to ISP’s and failure points, 3rd parties etc.
We are experiencing some minor demand for web page integration as a result of the pandemic, but since we have been doing wholesale distributor data exchange with our retail stores since day one, many of the tools for such things are already in place.
I certainly wish you guys continued success and demand for your product.
Keith Grill
President
DBMS Inc.
318-635-0757 – office
318-572-3196 – cell
--
You received this message because you are subscribed to
the "Pick and MultiValue Databases" group.
To post, email to: mvd...@googlegroups.com
To unsubscribe, email to: mvdbms+un...@googlegroups.com
For more options, visit http://groups.google.com/group/mvdbms
---
You received this message because you are subscribed to the Google Groups "Pick and MultiValue Databases" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mvdbms+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mvdbms/56866935-853c-4b78-b46a-e1733509f99bn%40googlegroups.com.
Hi Keith (and Will) – thank you for responding. I have tried to “educate” our comrades on the direction of business apps for a while.
For those that do not know me, a quick review. I have been in the Pick marketplace since day 2. I worked at SMI in Chicago and we received Reality serial #6, oddly I was employee #6. SMI grew close to 150 people with most being Systems Analysts (remember that job) and Programmers. I have worked for the ultimate (forgive the pun) providers of Pick systems.
14 years after joining SMI we became a Pick licensee for the development of Pick on IBM mainframes. The product was sold to The Ultimate Corp. and I went with it. When Ultimate began having financial problems, I was lucky enough to join Sequoia. 10 years later when Sequoia was having financial issues, it was sold to GA, I went. GA began to have financial difficulties and I joined jBASE. jBASE morphed into BlueFinity! At BlueFinity we discovered a product we were developing was a low-code/no-code platform. It is called Evoke and it creates new apps as well as modernizing legacy applications for MultiValue and Oracle and SQL Server and... I have been in product marketing, training and sales.
I know how good MultiValue applications are and I know that unless the they transform into customer/employee/partner/market focused native and/or web apps they are running on borrowed time. MultiValue applications can become competitive again if they join the digital marketplace. I see what is happening in the mainstream software marketplace and the requirement for apps running on smartphones, tablets, watches and diverse operating environments; Android, iOS, Linux and Windows.
We are all aware that business processes have significantly changed in the last few years. Technology has exploded. MultiValue developers can easily take advantage of technology advances AND changes in customer/employee/partner/market behaviors. They cannot do it by putting lipstick on a bulldog!
I had a customer when I was with Sequoia in the subscription/service business. One of their customers used their order-entry service; the customer published a catalog. The Sequoia user’s offices were in a warehouse. When I first walked into those offices, the warehouse was filled with over a hundred order-entry clerks, they were on phones taking orders. Within two years most of the phones and clerks were gone. The warehouse was empty with the exception of a few customer-service representatives on phones resolving customer issues.
The company had outsourced a provider of web catalog order-entry applications (could not be done in Pick). Customers now entered their own orders on the web. The data was batch “EDIed” to the Pick system overnight and that was about 30 years ago. It was kinda like an early Amazon.
Today Amazon like apps can be built in WEEKS using Pick business logic from the legacy system. It will take much longer to refactor the business logic from the Pick Basic/Python/Proc/Other code and to load the pictures and descriptions. It can be done incorporating modern Microsoft technologies supporting the MultiValue database. Of course, when I say use Evoke, the only low-code/no-code platform to support MultiValue I get…
Now in response to your real life factors:
My apologies for the Tony G. length of the posting and to Tony!