hevesh_c@apollo.HP.COM (Cathy Hevesh) (05/08/90)
Hi, I'd like to conduct a little informal survey to find out how popular and viable a package PHIGS is for all you application programmers out there who write graphics applications. It has always been my belief that solution suppliers demand standards and prefer to base their own software on standard platforms. It has recently been suggested to me that this is currently not necessarily the case and that there is a lot of resistance to using and/or porting to PHIGS for graphics applications. Is this true and if so, why? Thanks for your input. -Cathy Hevesh
dud@Alliant.COM (W.A."Dudley" Gaman) (05/09/90)
In article <4a45a410.20b6d@apollo.HP.COM> hevesh_c@apollo.HP.COM (Cathy Hevesh) writes: > >there is a lot of resistance to using and/or >porting to PHIGS for graphics applications. >Is this true and if so, why? > There is usually resistance to change, however PHIGS has been selected as the 3-D extention to X (PEX) and as the parallel graphics library in the Intel/Alliant i860-based Parallel Architecture eXtended (PAX) standard. While it is hard to create a graphics library which will suit everyone's needs, and while standards tend to fall somewhat behind the technology curve, PHIGS (and PHIGS+) can handle the requirements of a large number of users. Add this to the benefits of preservation of software investment (ie. it can be ported to other PHIGS platforms without changing source code) and high-performance available in systems which use PHIGS as their native instruction set and, IMHO, you have a package which is here to stay for a while. In my experience (sales support) I am seeing more situations in which PHIGS is a requirement. dud .
kauel@mentor.com (Kendall Auel) (05/10/90)
Re: PHIGS Any graphics library must have two important features in order to be accepted by the user community: o Functionality o Performance An important thing to keep in mind is that a well-designed application is 90% independent of any particular graphics library or standard. In other words, the low-level graphics routines are only called by a graphics interface layer within the application. Thus, standardization of the underlying graphics library is not a benfit to the majority of the application. The virtue of being a "standard" does not help PHIGS if functionality and performance are not superior to other options. (PHIGS is not truly standardized, anyway, since there is no guarantee of even source-level compatibility.) PEX does a good job of addressing a lot of problems with PHIGS. First, it is a window-based interface. Second, it allows for immediate-mode rendering. These help alleviate concerns with functionality. But what about performance? X-Windows is not necessarily known for being lightning-fast, and PEX suffers from the same reputation. I have seen some demos of PEX on high-end (read:expensive) workstations that are highly interactive. This should improve over time. For applications which have already decided to use the X-Windows system, PEX looks very attractive (especially considering the alternatives -- are there any?) Right now though, I don't think many 3D applications will decide to port to X just so that they can use PEX. The best hope for PEX is that X-Windows and Motif become the industry standards. These opinions are not necessarily those of MGC -- Kendall Auel | Mentor Graphics Corporation | Ich bin ein Berliner kauel@mentor.com | Silicon Design Division | "I am a pastry"