Today I’ve met a senior programmer. He used visual foxpro for his latest ERP system. Honestly I’m amazed with what kind of features that foxpro can provide and how good they are. Foxpro, really? That old not-popular programming language? Maybe that’s what you have in mind. But honestly, you’ll be amazed by how good Foxpro interact with database objects. Moreover, how fast the database is! Though it haven’t being tested over 50 users though.
I’m amazed at the features of foxpro!
So? Do we all move to foxpro all together? No, foxpro is good, but not in every aspect. One of the reason is that application structure is very tightly coupled with the data object, which is good for administrative tools but not for the other. I’m not using foxpro, so I can’t give more reason, and that topic is not what I want to discuss now.
It’s a desktop app!
I just want to shared that the visual foxpro that has been developed is desktop app. I must admit that no matter how good you are at web programming, you can’t defeat the UI functionality and keyboard-friendly functionality of desktop apps. There isn’t even safe cross browser hotkey that can be used in web. While in app, you can easily navigate using arrows, tabs, function keys (I’m pointing at you, F1-F12!), page up page down, control + keys, etc. Not to mention the responsiveness, and the functionality of one-page application (nah, ajax can’t even compete with it).
Web apps won’t win against desktop in UI and keyboard functionality
Man, I missed the desktop-based development time of my university time.
But that’s it. Desktop integrated system is losing popularity over year, as better web environment is emerging such as web socket, html5, css3, etc. But I must say that web apps will never beat desktop apps for transaction-intensive activities. No POS (point of sales) in hypermarkets using it.
Hypermarket using web-based POS? How slow will it be?
Do you imagine integrating barcode scanner with web apps? Do you imagine using mouse to choose which payment method? What will happen if they need to wait for respond after submitting, waiting for the payment change? That won’t do. We must back to our very first factor of application development: requirement. If they require it, then it must be done.
The very first factor of application development: requirement
So, we back to desktop app?
Yes for that particular activities. But how many of those transaction-intensive activities are there? There won’t be many of that kind transaction in today’s business. More often we need flexibility and portability more than keyboard shortcut. Let’s say reports, approvals, notes, attachments, all of them are better cross-browsers and portable. Employee self service (reimbursement, leave) is better online. Marketing will be more awesome online, since it can directly communicate to media via links and connections.
Of course we must ignore web-specialized features such as online trading, blog and social media though.
reports, approvals, notes, attachments, reimbursement, leave is better online
We go back to the very first rule of application development: requirement. If it is required to have intensive-transaction application or embedded devices (printer, barcode scanner), then use desktop. If not, then it’s better web since it’s easier to access and easier to maintain and publish. So, compare and plan before deciding the approach!