Desktop GUI Move To Web/Browser Based - When?


ecologix_akelly's picture

I realize this topic may engender passions on either side but I thought I would at least broach the subject as in my mind it is definitely the elephant in the room for future xTuple development. I am thinking we are at a stage where the pure desktop application and more specifically as it applies to xTuple is possibly an outdated model. I understand the considerable ramifications in such a move going forward possibly in regards to license management but you can use php encryption frameworks for commercial editions. I see the pro's heavily outweighing the cons, including:
- The largest Pro would be true multi platform and form factor compatibility (think tablets which to my knowledge are shutout from running xTuple currently, except in a drastically limited way)
- No need for a third party web app add-on or specific iPhone/Android app.
- Faster development and deployment of updates/fixes.
- Much wider exposure to developers for contributing functionality as growth in the current environment is certainly in web based application not pure desktop apps. Myself for example all major extensions I have made to our implementation have been web based (including full import of products, entry of a single product, p/o importing, prospect/opportunity creation). My reason for choosing this method is that I figured a full blown web front end to xTuple is a way off so I would attack it piece by piece myself as our needs arrive.
- Nothing at the database level would need to be changed and in many ways the attention to detail in pushing as much logic to the database makes me believe that xTuple is already planning for such a transition.
- Much greater flexibility with the interface and inclusion of external applications and modern technologies.
- Immediate regaining of countless hours of development time currently used to add scripting hooks into the C++ core, this becomes immediately obsolete and unnecessary.

The biggest con I see is that us as customers would need to endure a period of little to no new development while the desktop app was transitioned to a web based app but I really think this is a small price to pay now as I really see no valid argument against the reality that at some point the xtuple team are going to be forced to go this route at some point anyhow. Other issue would be need for a web server on customer install but this is minor and the best web server in the world today is open source and supported on multiple platforms.

Thanks for the platform to share my 2 cents

jrogelstad's picture
Joined: 12/10/2008
We're working on it. Stay

We're working on it. Stay tuned for an announcement in May.

ecologix_akelly's picture
Joined: 10/19/2010
Figured you'd already be a

Figured you'd already be a few steps ahead! Thanks Andrew

ma23693's picture
Joined: 03/19/2009

I can't think of a single web based application that I like.

Does anyone have a example of a web based application that is really user friendly?

Let's assume plenty of bandwidth, memory, and processor power on both ends.

That information might help me understand what I can expect.


johnf's picture
Joined: 02/03/2010
I have to agree in general.

I have to agree in general. However, they can be made to work. In general web apps tend to be prettier, slower, and lack that overall feel that users today expect. But it is possible to negate those three. Websites have been known to be very ugly, using an intranet server increases the speed, and with speed comes a better overall feel.

So NO I can't really provide a website that really works as well as a desktop app. But I can say the Google Doc's and a few others work well! So there is a trade-off.

I do think it is funny that several of my customers pay for the web apps but tend to use their desktop app to use the web app. They create a document in word (or something else) and cut and paste into google docs. Same with email. They don't always use the google web interface - just the services.

When customers ask for web apps I deliver. So far I have delivered three major website (not much to judge by). But I would call them complete failures. Each one is dying due to the lack of use.
Customers paid up to $20,000 for their website (pro graphic design, programming, high speed access, etc. it adds up quickly) and I think it was a waste. In one case last year they had 24,000 opportunities to use the web and it was use 391 times. That's as cost of ~$51.00 each. That was not a good ROI.

In another case everyone wanted the customer to have access to their data. Great and it sounded good to me. Two years later customers accessing their data is limited at best. In fact my logs show only one customer takes advantage of the web on a regular schedule. It seems all the electronic things the program was doing (email, electronic invoices, shipping notices, driver updates, etc.) filled the need to know what was going on with their products.

I guess what I'm saying is you have to be very sure of the need - make sure the need is not filled by other means.