As james said it in his first reply, you should become clear what you need and what you want. How critical and how important is it for the company?
If it is important or even critical for the core business of the company (or say if it doesn't work, it costs 100.000 per hour) I would engage you to write it by yourself. Otherwise it can happen to you, that features you considered important are droped down from the framework or your technological or organizational direction is considered obsolete. There is always a risk factor working with open source projects (there are chances too), but the biggest problem is, that you just have noone to kick in the ass if something is broken. By the way, the abovementioned "direction cut" seem just to happen to us (with struts) so we are thinking about abandoning it, but thats a different story. If it's not THAT important, you should still choose something which isn't just in the first release. Talking about frontend I wouldn't even think about JSF or SHALE, since they are extremely new. Struts 1.2.7 is believed to be a high quality, stable product, so it could be a very reasonable option, but you should test it yourself first. Moving along the application, for the mid-tier spring or pico-container seem to be the best available choice. On the other hand, what's wrong with your POJOs? Do you really need a container? Spring is a lightweight container, but it's still a container. My first choice would be POJOs, as you already seem to have, and in case you need distribution CORBA or RMI (last if performance isn't an issue). For the persistence layer, there are many frameworks available, iBatis and Hibernate among them. I don't know enough about your application to give you reasonable advices (which I normally bill by 100$ per hour :-)), but I think you should make a requirement analyses and compare the frameworks. Since you already work with XML on the persistence side (my interpretation of a "custom built XML based data access layer") you maybe should look into XML-DBs. After all, you should really invest some time into proper requirements analysis to make the right decision, based on your requirements and not someone else's preferences. Regards Leon > -----Ursprüngliche Nachricht----- > Von: Balasubramaniam, Sezhiyan [mailto:[EMAIL PROTECTED] > Gesendet: Donnerstag, 1. September 2005 21:07 > An: user@struts.apache.org > Betreff: Tech-Stack Choices > > I am kind of new to this forum... Need some advice from the gurus. > > > > In eBay, for our all intranet applications, we use "JSP > presentation, STRUTS 1.0.1 as controller, java based > application objects for business logics and custom built XML > based data access layer". > > > > Now we are in the process of upgrading all our apps/apis/servers. > > > > I thought of utilizing this opportunity and try to upgrade > our STRUTS applications, to come back to the industry-wide tech-stack. > > > > But by following the forum, I am getting confused which one > to choose and what are their purposes: > > > > STRUTS 1.2.x? > > JSF? > > SHALE? > > SPRING? > > HIBERNATE? > > > > Can someone give a high level some idea and some pointers? > > > > Our main focus is to align with the "viable industry > standards", the same time not to jump into the usage of beta > API's which are not tested out thoroughly. > > > > Appreciate the help. > > > > Bala > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
