User contributions
Jump to navigation
Jump to search
- 21:44, 16 October 2007 diff hist 0 m Code Racer →Sample Results: revise to match latest code
- 21:37, 16 October 2007 diff hist 0 m Code Racer →Code: ouch ouch ouch count wins0 as a win for 0, wins1 as a win for 1, not the other way around of course
- 21:32, 16 October 2007 diff hist 0 Code Racer →Code: ouch another tupo -- report the more wins as the winner, not the lesser average
- 21:29, 16 October 2007 diff hist +133 m Code Racer →Sample Results: 1000 runs now producing reasonable results for both averages and wins
- 21:18, 16 October 2007 diff hist +193 m Talk:Code Racer catch up the bug report re wins not summing to near scale to the changes in the main article's code
- 21:13, 16 October 2007 diff hist −6 Code Racer →Code: fix the stuttered line of code that was making wins for 0 and 1 nearly match always
- 21:06, 16 October 2007 diff hist +183 m User:Ppaatt Lynagh →Thanks: to Strife for "clear & conventional" and LSL Script Efficiency, to Xavier for Efficiency Tester, to Thraxis for startswith and endswith in llSubStringIndex
- 20:47, 16 October 2007 diff hist +126 m User:Ppaatt Lynagh →Functions: add Code Racer - Race two versions of code forever, to see which runs faster.
- 18:08, 16 October 2007 diff hist +402 N Talk:Code Racer bug I guess -- often the sums of wins exceed the scale
- 17:29, 16 October 2007 diff hist +21 m Hex →Implementations: consistently always include a summary of how different in the section headings of the different code
- 17:28, 16 October 2007 diff hist +102 Hex →Implementations: label our too many examples as clear, conventional, fast, small, and/or different -- never "correct at a glance"
- 17:22, 16 October 2007 diff hist +918 m Talk:Hex quote the Wiki.secondlife.com edit page and Wikipedia on the neutral point of view we're building here
- 17:15, 16 October 2007 diff hist +25 LlSubStringIndex restructure startswith to work like endswith, rather than burning run time to search all the haystack in the often common case of needle not found
- 17:14, 16 October 2007 diff hist +74 LlSubStringIndex useful snippets -- trust llGetSubString to understand negative indices, rather than adding llStringLength(haystack) on ourselves
- 17:04, 16 October 2007 diff hist +140 m LlSubStringIndex collect all our clear and conventional examples to look for one string inside another all together
- 16:58, 16 October 2007 diff hist +29 LlSubStringIndex confirm our Efficiency Tester claims of equal run time with the much different Code Racer benchmark
- 16:57, 16 October 2007 diff hist +6,757 N Code Racer Race two versions of code forever, to see which runs faster
- 16:31, 16 October 2007 diff hist −27 m Category:LSL Library anonymise my contributions - as I learn the netiquette here, I begin to think that's allowed
- 16:29, 16 October 2007 diff hist +95 Category:LSL Library add Code Racer - race two versions of code forever, to see which runs faster
- 16:24, 16 October 2007 diff hist +69 m Talk:Hex →Correct At A Glance: ouch ouch ouch forgot to sign again sorry
- 16:23, 16 October 2007 diff hist +410 Talk:Hex →Correct At A Glance: how about clear/ small/ fast rather than correct-at-glance/ small/ fast
- 16:21, 16 October 2007 diff hist +378 m Talk:LlSubStringIndex and next would be startswith a la endswith, then clarity & space & time cost comparisons here, with the winner in the article
- 16:19, 16 October 2007 diff hist +214 m User talk:Thraxis Epsilon I see you improved llSubStringIndex by contributing startswith/ endswith. Thank you!
- 06:38, 16 October 2007 diff hist +29 m Talk:Hex →Due Diligence: now I understand the difference - we can say found at Efficiency Tester and then explored in more depth at LSL_Script_Efficiency
- 06:36, 16 October 2007 diff hist +349 Talk:LlSubStringIndex propose snippets startswith(whole, part) and endswith(whole, part)
- 06:32, 16 October 2007 diff hist +428 LlSubStringIndex also present the drearily-conventional correct-at-a-glance snippet that means contains, not just the smaller snippet -- and note as yet we have no faster snippet
- 06:14, 16 October 2007 diff hist +188 Efficiency Tester try the empty test of deleting the { counter += 1; } source line to see the astonishing inaccuracy of this instrument
- 06:09, 16 October 2007 diff hist +44 m LSL Script Efficiency →How Fast Does That Code Run: add llOwnerSay llGetFreeMemory to resynch to Efficiency Tester
- 06:09, 16 October 2007 diff hist +44 Efficiency Tester add { llOwnerSay((string) llGetFreeMemory()); } up front: running that there measures code cost before heap grows and announces start of run time
- 06:06, 16 October 2007 diff hist +116 m LSL Script Efficiency →How Fast Does That Code Run: guess which code cost we meant by "optimize" and complete the thanks line
- 06:03, 16 October 2007 diff hist +127 LSL Script Efficiency →How Fast Does That Code Run: Please note this code and its comments should exactly match the code and comments presented by the Efficiency Tester page.
- 06:02, 16 October 2007 diff hist +94 m LSL Script Efficiency resynch the llGetTimestamp code here with the master copy at the brief Efficiency_Tester article
- 06:01, 16 October 2007 diff hist +118 LSL Script Efficiency →How Fast Does That Code Run: reword the rule against reducing accuracy to avoid contradicting the overriding invitation for all the world to mercilessly edit and redistribute our writing at will
- 05:57, 16 October 2007 diff hist −3 Efficiency Tester launched by Xaviar Czervik, not just copy-edited by Xavier, says the history at LSL Script Efficiency
- 05:54, 16 October 2007 diff hist −100 LSL Script Efficiency strike the proposed rules for posting - to advance the science, we must invite people to present reproducible results, not forbid the posting of results
- 05:51, 16 October 2007 diff hist +10 m Efficiency Tester reword the link with LSL Script Efficiency
- 05:49, 16 October 2007 diff hist +16 m Efficiency Tester sort the paragraphs - give the reader what matters most to the reader first
- 05:47, 16 October 2007 diff hist +61 m Efficiency Tester For a less brief discussion, see: LSL Script Efficiency
- 05:44, 16 October 2007 diff hist +24 m Efficiency Tester clarify the English - words more short and real - the quick point first, the depth second
- 05:42, 16 October 2007 diff hist +94 m Efficiency Tester clarify - add comments and white space and less distractingly creative variable names and never single letter variable names -- also strike the old never-reached un-explained llInsertString code
- 05:24, 16 October 2007 diff hist +618 Efficiency Tester guess why this page thru yesterday mystically presented exactly the same code with less comments than the LSL Script Efficiency page
- 04:54, 16 October 2007 diff hist +374 LlSetObjectName gather the caveats together, add the Chat History specific, add the Edit General Name caveat
- 04:42, 16 October 2007 diff hist +152 LlSetObjectName grow example to "$yyyy-$mm-$dd $Object" from "Object X"
- 21:49, 15 October 2007 diff hist +32 m Talk:Hex →Due Diligence: link with the apparently pointlessly redundant LSL_Script_Efficiency article also, hope someday to distinguish that clearly from Efficiency Tester
- 21:47, 15 October 2007 diff hist +147 m LlGetTimestamp link with the Efficiency Tester and LSL_Script_Efficiency that calls this llGetTimeStamp function, also link with http://en.wikipedia.org/wiki/ISO_8601
- 21:41, 15 October 2007 diff hist +252 m Talk:Hex →Due Diligence: link with Efficiency Tester example of measuring speed by watching llGetTimestamp progress
- 21:32, 15 October 2007 diff hist +370 Talk:Hex →Due Diligence: yes consistently now, all five llGetFreeMemory LSO sizes reported in the article I now reproduce
- 21:24, 15 October 2007 diff hist +987 Talk:Hex →Due Diligence: sometimes 347, sometimes 351, is code space measure of first instance
- 21:13, 15 October 2007 diff hist +145 Talk:Hex I don't yet understand how the first copy occupies 351 bytes and subsequent copies occupy 347 bytes. Do you?
- 20:08, 15 October 2007 diff hist +1,508 m Talk:Hex →Due Diligence: ouch post the beginning of the delete experiment, not the end