4 ms·

>It means the claims are unverifiable. The test set especially should be available. Eh..Why should anyone in the world care that a compressor can compress test

by ksk 10y ago

>It means the claims are unverifiable. The test set especially should be available. Eh..Why should anyone in the world care that a compressor can compress test data well? Except vaguely, it might serve as a motivator to investigate the use of said compressor for their own purpose. But then again, this is simply marketing, for developers. To me, its like saying "the new compiler version is faster" on this collection of source files. Personally, I couldn't care less what the test set is, if the compiler is not faster for my own use case. So yeah, if anyone is interested, they could probably get an evaluation copy, try it out on their own data and decide for themselves.

sp332 10y ago
Of course it would be better, but I thought it was implausible that the code would be public given that the corpus is private. I guess it's possible though.
michaelmrose 10y ago
Then the blog post is a waste of everyone's time and ought to be deleted from hacker news. Either it provides enough information to replicate the benchmarks or its spam. If it provided more freely available information other parties could test immediately and post and this would provide additional feedback to help parties decide if it is actually worthwhile.
- ksk 10y ago
  >Then the blog post is a waste of everyone's time and ought to be deleted from hacker news. You're entitled to your own opinion. > Either it provides enough information to replicate the benchmarks or its spam. If it provided more freely available information other parties could test immediately and post and this would provide additional feedback to help parties decide if it is actually worthwhile. Using that logic, any posts here advertising a new untested product should be banned. That would include all product announcements from $Company. Infact, that would probably include reviews because they contain subjective impressions about "performance" and other criteria that nobody can objectively verify and "provide additional feedback to help parties decide if its actually worthwhile".
  - michaelmrose 10y ago
    Numerical analysis is different that a description. People expect you to back up numbers

by ksk 10y ago

sp332 10y ago

Of course it would be better, but I thought it was implausible that the code would be public given that the corpus is private. I guess it's possible though.

michaelmrose 10y ago

Then the blog post is a waste of everyone's time and ought to be deleted from hacker news. Either it provides enough information to replicate the benchmarks or its spam. If it provided more freely available information other parties could test immediately and post and this would provide additional feedback to help parties decide if it is actually worthwhile.

ksk 10y ago
>Then the blog post is a waste of everyone's time and ought to be deleted from hacker news. You're entitled to your own opinion. > Either it provides enough information to replicate the benchmarks or its spam. If it provided more freely available information other parties could test immediately and post and this would provide additional feedback to help parties decide if it is actually worthwhile. Using that logic, any posts here advertising a new untested product should be banned. That would include all product announcements from $Company. Infact, that would probably include reviews because they contain subjective impressions about "performance" and other criteria that nobody can objectively verify and "provide additional feedback to help parties decide if its actually worthwhile".
- michaelmrose 10y ago
  Numerical analysis is different that a description. People expect you to back up numbers