i have to admitt i was wondering as well, but after the below explanation i like the idea. for areas that you have multipule people working on single progects it would allow greater interaction. it might also,if i understand correctly, allow the use of clusters. On Friday 09 May 2003 13:20, you wrote: > Since both of these posts seemed so much alike, i decided to combine > them and contiue the thread under Rolf's much more accurate subject. > > On Fri, May 09, 2003 at 05:43:12PM +0200, Thomas Schmidt wrote: > > You talked about client/server architecture. Could give me some more > > information about that client/server idea? What should be the task of > > the > > server? Just data storage? Computing solutions for algorithms > > (modeller)?= > > =20 > > And of course, what would be the benefits for a user? > > and, > > On Fri, May 09, 2003 at 06:05:48PM +0200, rolfinternet@xxxxxx wrote: > > What is the meaning of "CAD with client/server architecture"? > > For me, this term is clear from applications of database.(i.e. ERP) > > What I mean by client/server for a cad application is mainly the > seperation of the CAD engine from the user's interface. This seperation > could mean that the engine runs on a server ( one instance for all the > clients ) and the gui runs on a workstation, or they could both be > running on the same machine. > > The server/engine would do the bulk of the work that wasn't directly > related to the user interface. Such work would include calculations, > database manipulations, etc... > > The client/user-interface would handle the interaction between the user > and the engine. Its main task would be object rendering. It would also > send manipulation requests (draw a line from point A to point B) to the > server for processing. > > The benefits that I see in this approach are twofold: > > First, the user isn't tied into UI/toolkit that they don't want and/or > aren't interested in. This is my problem with QCAD, I don't have any other > programs that use QT, and I don't want to install QT just for one > program. If the engine is seperated, then it is just a matter of writing > an UI for the toolkit that you prefer. And if the functions used by the > engine are well documented, then writing a UI should be fairly simple > for someone familiar with that toolkit. I consider this the primary > benefit of engine/interface seperation. > > Second, the client/server approach may allow UIs that don't require X. > While I haven't fully thought out this possibility, it at least seems > interesting. > > There are probably also benifits with regarding the engine as > middleware, allowing one engine and one database to serve multiple > clients. > > More important, in my opinion, is that the engine and UI be extensible > to allow modules to be added for functions that apply to specific > engineering disciplines. This however, may actually be made more > difficult by a client/server architecture, but I'm not certain. > > I'm interested in hearing any and all comments (short of 'you are a > dumb-a**) regarding these ideas. > > I'm also _extremely_ interested in laying or helping in laying the > groundwork for Generic All-purpose CAD for Linux GACL (tm) :-) > rather than seeing anymore YASCLs littering the abondonware road. > > Regards, > Lee -- Oh i've slipped the surly bonds of DOS and danced the skies on Linux silvered wings. http://pfrostie.freeservers.com/cad-tastrafy/ //www.freelists.org/webpage/cad-linux