|
This depends on your architecture, Don. A JSP is really a servlet - in simplest terms, it runs on the host, formats HTML and outputs it to the browser. You can output JavaScript in the HTML stream, which can in turn do some pretty fun stuff, but IMHO that's not a good design for any serious business logic. I'm not entirely clear on your requirements. Client-side logic can entail one of two scenarios: 1. Creating a "local" data image and updating it purely locally, and then showing views of the local data to the end user (in the early 90's we coined the term "detached processing" for this architecture). This can even involve a batched update back up to the host. 2. Interactive update of host data through logic on the client using an ODBC bridge. Personally, I won't do option two. From what I know of IIS and RDS, if you open an ADO to an OLE DB on the AS/400, you're directly updating data on the host. And even for temporary files you've got a whole lot of traffic going on, and a bunch of overhead on your host. It's just hidden from you as a programmer, so it's not only slow, but you also have less control over it. Other than that, how was the play, Mrs. Lincoln? If this is your design goal, then I probably can't help much. So let's assume you're doing option 1. This is really a thick client running in detached mode. While there are applications where this is a good thing, there are disadvantages as well. Unless you are actually storing data on the local hard drive, a failure on the client means you lose all work up until that point, and workstations are notoriously less stable than AS/400's. I'd be interested in knowing more of your requirements here. Anyway, if you're designing a true thick client interface, I'd be much more likely to use a server on the host. The server can be called directly from the workstation, using sockets or IBM's Java Toolbox. One call to the server downloads the data, one call uploads it back to the host. Editing can be done at upload time or upon user request. You have all the capabilities of a true thick client application, but without being locked into any particular transport mechanism. You can even design this as an applet, I suppose, if your clients refuse to install thick client software on their PC. In any event, you're now moving into an area where you're defining "how" you want to do things, rather than what or even why, and so it becomes more difficult to analyze. Why do you want to have a "local" recordset that you update? Why not simply send update requests to the host? As I pointed out above, if you're using IIS RDS and ADO talking to an OLE DB source on the AS/400, you're doing that anyway, except you're adding all the extra overhead of ODBC, and you're tied to Microsoft's architecture. Anyway, give me a little more idea of what your application needs to do, how the users want to interact with it, and so on, and maybe I can make a better recommendation. Joe > -----Original Message----- > From: Schenck, Don > > Joe -- > > Our operating condition: Our clients will ONLY use Internet > Explorer 4.0 or > higher. Nothing else. > > IF we use IIS and RDS, we can get a recordset from the server and do a LOT > of database manipulation (adding, updating and deleting records) AT THE > CLIENT without going back to the server. > > That's really nice. > > Can I do the same thing using XML and JSP with the 400? > > -- Don Schenck
As an Amazon Associate we earn from qualifying purchases.
This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].
Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.