Oracle now crossing on Rubin's prior testimony, which is all terrible
-
-
Remember when Douglas Schmidt conducted build tests to see if Android would run without the API packages?
-
Testimony responding to those build tests.
-
Google: What do you mean, failed to build? Astrochan is describing what it means for code to build (source code to object code)
-
It is a very clear explanation.
-
Not sure I heard correctly earlier but I think Astrachan tested his own builds?
-
Ah yes, he did. So first, when he removed all the Android packages, the platform failed to build.
-
Then he removed one of the 37. He removed one Android packages, and again the platform failed to build.
-
(These are all mirrors of the Schmidt tests)
-
(Wait for the mic drop I think I know what's coming)
-
He removed declaring code, and then that failed to build.
-
Google: Were you surprised? Astrachan: No. Because when you change the source code, it's common for it to fail to build.
-
Astrachan: I don't think this shows that the 37 packages have any more importance or criticality than any other packages.
-
Not sure I understood correctly — Astrachan did this with non-Java packages?
-
Now we're looking at the visualizations of the API packages provided by Schmidt.
-
Astrachan said that these visualizations are nothing like what Sun and Oracle provided to developers to explain Java.
-
He says there are books and posters showing classes that are distributed by Sun-Oracle for devs.
-
oh man
-
Astrachan is responding to Reinhold it looks like. Oracle objects, says it's not in the scope of what Reinhold said.
-
Reinhold wasn't disclosed as an expert witness, he was brought in as an employee. Oracle withdraws its objection.
-
We're looking at the examples of declaring code, explaining that he can understand what a method does just from the name.
-
He said that during trial, he "suspected" he could guess what the method did, then after trial he went back and looked it up.
-
When he looked it up, his conclusion just based on the name was correct.
-
(So this is about the functionality of the code)
-
Astrachan: ... each part is very descriptive and functional about what the method does.
-
Now addressing Reinhold testimony that declaring code is sometimes more important than implementing code for devs.
-
Astrachan: You can't do that [come to that conclusion].
-
Astrachan: Developers would absolutely care about implementing code.
-
Astrachan: If the declaring code were good, easy to understand, but the implementing code were buggy ... it would be very difficult.
-
Astrachan: Typically, bugs for developers are in their code, not in the library code. .... it would not be so good.
- 42 réponses de plus
Nouvelle conversation -
Le chargement semble prendre du temps.
Twitter est peut-être en surcapacité ou rencontre momentanément un incident. Réessayez ou rendez-vous sur la page Twitter Status pour plus d'informations.