kaleb@thyme.jpl.nasa.gov (Kaleb Keithley ) (10/24/90)
A couple of weeks ago, someone (sorry, due to lack of disk space, we expire all articles in seven days or less, so I can't recall who) posted an article acknowledging a "consortium internal" fix to TMstate.c in Xt, that dramatically improved the performance of Xt in general and Motif 1.1 in particular. (How's that for a run-on sentance?) 1st question: Would that person, please send me the patch? Much obliged, I'll buy you a beer/wine/whatever when you're in the area. Take you on a personal guided tour of the Lab even. Most, if not all, projects here at the lab run under the intense scrutiny of configuration management, probably not unlike most other places. Trying to get CM to apply the public patches is hard enough; trying to get them to apply non-public patches is even harder. If this fix does indeed exist it would greatly simplify a lot of things. CM for one, and it would eliminate the need to write a replacement for the Motif Scrolled List Widget, etc. for another. 2nd question: MIT, please release this as a public fix? If R5 is truely eight plus months away, that's a long time to wait for this much needed fix. Thanks. -- Kaleb Keithley Jet Propulsion Labs kaleb@thyme.jpl.nasa.gov causing trouble again.
gabe@hpcvlx.cv.hp.com (Gabe Begeddov) (10/27/90)
/ hpcvlx:comp.windows.x.motif / kaleb@thyme.jpl.nasa.gov (Kaleb Keithley ) / 9:12 am Oct 24, 1990 / A couple of weeks ago, someone (sorry, due to lack of disk space, we expire all articles in seven days or less, so I can't recall who) posted an article acknowledging a "consortium internal" fix to TMstate.c in Xt, that dramatically improved the performance of Xt in general and Motif 1.1 in particular. (How's that for a run-on sentance?) We provided a rather extensive rewrite of the entire translation manager to MIT and OSF. This was our snapshot at the time. We have since made quite a few bug fixes which we will also be providing to MIT and OSF. One of the many areas of performance related changes that we made had a significant impact on the interactive performance of the text widget. Due to pressure from various sources we decoupled that one change and made it available as a diff to the current MIT public source pool. The diffs for this specific change were to TranslateI.h and TMstate.c 1st question: Would that person, please send me the patch? Much obliged, I'll buy you a beer/wine/whatever when you're in the area. Take you on a personal guided tour of the Lab even. Management's preference is that we don't send it directly but allow MIT/OSF to be the intermediaries. 2nd question: MIT, please release this as a public fix? If R5 is truely eight plus months away, that's a long time to wait for this much needed fix. My impression is that MIT is planning to sit on the changes until they feel comfortable with them and have modified them to their preferences. While we made the specific change for interactive text widget performance separately available it is our preference that the entire suite of translation manager changes be used. This makes sense since that is what our software is based on. I talked with Vania Joloboff at OSF today and he said that he had made that specific interactive processing change available to MIT so that's where things stand at the moment. Thanks. -- Kaleb Keithley Jet Propulsion Labs kaleb@thyme.jpl.nasa.gov causing trouble again. ---------- Gabe Beged-Dov Interface Technology Operation Hewlett Packard