ph3n1ks
July 22nd, 2006, 08:39 PM
First off, let me say hello to everyone. I am a software/web developer in addition to many other things I do. Mostly I do the other things to make a living. But occasionally, I do get to engage in my chosen profession in the small business sector. Anyway, on with the thrust of things.
I have been considering starting an open source project. I am soliciting general interest and specific developer interest in this project that is in the concept phase right now. Read below to find out more.
For several years now, I have noticed inadequacies in SME (Small to Medium Enterprise) targeted Enterprise Resource Planning (ERP) softwares. Now lately on sourceforge.net, I have seen an abundance of them. However, they are all web based. While I am sure that these solutions address the issues that I have with small accounting packages like QuickBooks and larger accouting packages like the MAS series, many small business owners couldn't tell you the difference between the "web" and their local network, if they even have one. I have literally run into business owners who have told me that they would stick with QuickBooks even if the product is hard to work with, next to impossible to get data out of, and incredibly expensive with its endless upgrade treadmill and market of third party tools that become necessary to get anything usefull out of the application. Many of them do however see the advantage to role-based, multi-user access and all of the data being together and accessible in one place/application.
So here is the concept:
A large enterprise class ERP package targeted to the SME sector that uses smart client technology to connect to a server application. The entire thing needs to be extensible so that vertical market modifications are possible. I prefer to build it on open source platforms (MySQL/PostgresSQL, Mono/.NET, Apache (yes, I did say something against web based... but you will see later)).
The Server End:
I believe that there are two ways to implement this and follow my concept. First, using a VMware Virtual Appliance (created by the project), a web service with database that can be deployed with VMware Player. The second idea, one that I don't really like per se, is to create a server application bundle that gets installed on the end-user system just like another program.
Personally, I go with the web service concept on a Virtual Appliance. This presents many advantages. In addition to allowing access to the accounting/ERP data across the local network, it makes it possible to expose certain parts of it to an ecommerce or EDI web service as well. This results in elimination of double or even triple data entry. Centralized data management also makes it easier to back-up the data. And further, exposing a web service makes it up to the end-user to decide whether to use the available client application or maybe get one custom made for his/her need.
The Client End:
The Client End actually gets split into two parts, the application that through some mechanism has the ability to access the entire web service and the client that connects to the ecommerce web service.
The client for general access is where the smart client application comes into play. I believe that this should be modular to support not installing certain modules if an end-user doesn't want them, with the ability to install them later if needed. I prefer an application based on what features of the .NET Framework 2.0 that the Mono Project (http://www.mono-project.com)supports. In this way, the client could be ported to any .NET capable platform.
The client that connects to the ecommerce web service will essentially be nothing more than an ecommerce website that either can connect directly to the Server end or can be synchronized manually from the general access smart client application.
Conclusion:
At this point, the project is in the concept phase. I am looking for general interest in such an application and specific interest from developers in the open source community. So, if you read this and meet either group, please leave a reply and let me know where you would fit.
I have been considering starting an open source project. I am soliciting general interest and specific developer interest in this project that is in the concept phase right now. Read below to find out more.
For several years now, I have noticed inadequacies in SME (Small to Medium Enterprise) targeted Enterprise Resource Planning (ERP) softwares. Now lately on sourceforge.net, I have seen an abundance of them. However, they are all web based. While I am sure that these solutions address the issues that I have with small accounting packages like QuickBooks and larger accouting packages like the MAS series, many small business owners couldn't tell you the difference between the "web" and their local network, if they even have one. I have literally run into business owners who have told me that they would stick with QuickBooks even if the product is hard to work with, next to impossible to get data out of, and incredibly expensive with its endless upgrade treadmill and market of third party tools that become necessary to get anything usefull out of the application. Many of them do however see the advantage to role-based, multi-user access and all of the data being together and accessible in one place/application.
So here is the concept:
A large enterprise class ERP package targeted to the SME sector that uses smart client technology to connect to a server application. The entire thing needs to be extensible so that vertical market modifications are possible. I prefer to build it on open source platforms (MySQL/PostgresSQL, Mono/.NET, Apache (yes, I did say something against web based... but you will see later)).
The Server End:
I believe that there are two ways to implement this and follow my concept. First, using a VMware Virtual Appliance (created by the project), a web service with database that can be deployed with VMware Player. The second idea, one that I don't really like per se, is to create a server application bundle that gets installed on the end-user system just like another program.
Personally, I go with the web service concept on a Virtual Appliance. This presents many advantages. In addition to allowing access to the accounting/ERP data across the local network, it makes it possible to expose certain parts of it to an ecommerce or EDI web service as well. This results in elimination of double or even triple data entry. Centralized data management also makes it easier to back-up the data. And further, exposing a web service makes it up to the end-user to decide whether to use the available client application or maybe get one custom made for his/her need.
The Client End:
The Client End actually gets split into two parts, the application that through some mechanism has the ability to access the entire web service and the client that connects to the ecommerce web service.
The client for general access is where the smart client application comes into play. I believe that this should be modular to support not installing certain modules if an end-user doesn't want them, with the ability to install them later if needed. I prefer an application based on what features of the .NET Framework 2.0 that the Mono Project (http://www.mono-project.com)supports. In this way, the client could be ported to any .NET capable platform.
The client that connects to the ecommerce web service will essentially be nothing more than an ecommerce website that either can connect directly to the Server end or can be synchronized manually from the general access smart client application.
Conclusion:
At this point, the project is in the concept phase. I am looking for general interest in such an application and specific interest from developers in the open source community. So, if you read this and meet either group, please leave a reply and let me know where you would fit.