Jump to content

Talk:Rugg/Feldman benchmarks

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Benchmark results?

[edit]

I have another printed source from 1983 with benchmark results for BM1 - BM8. Most of the values match the examples in the article, with an exception for BM8 on the Apple II with Applesoft BASIC which in this article is listed at 55.5 seconds but my source says 107 seconds. I know it would be fairly easy to take new measurements, either on real hardware or an accurate emulator but perhaps that counts as own research? I could also add more examples to the list, if that is Wikipedia worthy to have. --Carlsson~svwiki (talk) 12:17, 12 February 2020 (UTC)[reply]

OR is a tricky one. The issue is not really the source, but whether that source can be independently verified. So if you say you have an antigravity machine and there's no other demonstration, OR applies. But in this case, anyone can verify the numbers by running the same test on their machine/emulator. So I say go for it. Maury Markowitz (talk) 16:09, 14 April 2020 (UTC)[reply]
Thank you for your support. I did my best to amend the article with additional benchmark results, both published and community generated. Carlsson~svwiki (talk) 22:18, 29 April 2020 (UTC)[reply]

My oompliments to those who wrote, edited, and updated this article. I think you did an excellent, thorough job. I am the primary author of the article discussed. My co-author, Phil Feldman, agrees. We are both still around, although it's been many years since we wrote a magazine article. --Tom Rugg. — Preceding unsigned comment added by 2600:8802:5519:3C00:43D:B6E7:8BDA:A700 (talk) 11:03, 21 January 2022 (UTC)[reply]

Errors in the table

[edit]

Looking at the article in the reference list ( Coll, John (February 1978). "Direct Addressing: Where to get your personal computer". Personal Computer World. pp. 55–58. (the link to it works fine)), it looks like there have been some mistakes in tranferring the numbers to the Wikipedia article.

The numbers in the Wikipedia article for Benchmark 8 seem to be from the last column in the original article, which actually holds the sum of benchmark 7 and benchmark 8. Also, this sum is incorrect in the original article for at least one platform.

Someone has added benchmark times for the Commodore 64, including 119 seconds for benchmark 8. When I try to reproduce it, benchmark 8 takes 11.4 seconds. If I increase the loop to 1000 iterations, it takes 116.2 seconds. 2001:2002:D540:AA27:250D:FDA0:F3A9:28AB (talk) 17:35, 11 January 2023 (UTC)[reply]