korp@atlantis.ees.anl.gov (Peter Korp) (11/28/90)
In article <1990Nov27.012631.15725@odin.corp.sgi.com> erik@westworld.esd.sgi.com (Erik Fortune) writes: [lots of X - NeWS history deleted] >X had been through 9 releases (one number was reserved but never used) and >the X community had learned a lot from those first 10 releases. Lots of >the lessons learned from those early releases of X were applied to X11. >That experience also pointed to some likely problems with the design of NeWS. >Bob wrote a paper describing some of his concerns about NeWS in late 1986. >I'll see if I can dig up a copy... > I'd be interested in having a copy of that. What I find absolutly amazing is how every new release of X looks more and more like NeWS. Shared memory extensions, circular windows, PostScript imaging, synthentic input extensions, and I don't know what else. Don't you think one of the reasons Sun was able to implement the X11/NeWS server as cleanly as they did was because of the great similarity of X11 and NeWS? In computer terms 1986 is a *long* time ago. I think it speaks well of Goslings "research effort" that his design has remained relatively unchanged since then. >In 1986 it was widely believed that what UNIX really, really needed yesterday >was standard graphics. The only things that were not proprietary and were >remotely close to being understood well enough to be standardized were X10 and wm, >neither of which was acceptable. Hence X11. > You spoke of X11 *having* to be developed quickly due to X10s acceptance in the UNIX world, I think the battle scars of this rapid development still show. > >-- Erik Peter
jg@crl.dec.com (Jim Gettys) (11/28/90)
In article <1990Nov27.170031.23812@mcs.anl.gov>, korp@atlantis.ees.anl.gov (Peter Korp) writes: > > I'd be interested in having a copy of that. What I find absolutly amazing > is how every new release of X looks more and more like NeWS. Shared memory > extensions, circular windows, PostScript imaging, synthentic input extensions, > and I don't know what else. Don't you think one of the reasons Sun was able > to implement the X11/NeWS server as cleanly as they did was because of the > great similarity of X11 and NeWS? The same thing can be said about NeWS... Selections, properties, etc. Also remember that NeWS was looking over our shoulders as well, during its development, an advantage we did not have, as Sun was close mouthed until NeWS was announced (prematurely in my book; the first version was buggy, and kissed off the imaging market entirely, due to lack of color map capability, for example; I may not like color maps, but they enable a class of applications I used to make my living writing. From my perspective, I think it very unfortunate that Gosling et. al. were not allowed more time. Once things hit the street, it is very hard to revisit any major design choices.). > In computer terms 1986 is a *long* time ago. I think it speaks well of Goslings > "research effort" that his design has remained relatively unchanged since > then. As has X11. It has been about as stable in the same period of time. > You spoke of X11 *having* to be developed quickly due to X10s acceptance in > the UNIX world, I think the battle scars of this rapid development still show. As does NeWS. I was quite disappointed when it was announced that it had not faced up to certain issues I reguard as essential (and are present in X, and have been for a long time). Retrofitting them into NeWS afterwards has been a painful process. Both systems explore different points in a very rich design space; I happen to believe some fundamental choices in NeWS are wrong, and James and Dave think likewise about X. Such is life. If you prefer one over the other, so be it. - Jim