New candidate JEP: 495: Simple Source Files and Instance Main Methods (Fourth Preview)

[Brian Goetz]

As you know, naming things is one of the hardest problems in computer 
science.  During the course of this project, we tried on several names: 
unnamed classes, implicit classes, and others.  And there were 
*extensive* discussions about the pros and cons of each.  Rest assured 
that the point you made about implicit classes was well understood and 
represented in the discussion.  In the end, especially given the 
audience for this feature, the simplicity of "simple" won over the 
explicitness of "implicit".

One of the many reasons naming is hard is that there is never an 
obviously-right answer; there's only answers that are more or less 
compelling to more or fewer people.  So even if 90% of the people think 
it is 90% better, there will still be that 10% who are disappointed.  
It's the nature of the game.

(And yes, the 23 ship has sailed.)


On 11/14/2024 4:47 PM, Eirik Bjørsnøs wrote:
> On Thu, Oct 24, 2024 at 4:50 PM Gavin Bierman 
> <gavin.bierman at oracle.com> wrote:
>
>     What this JEP is proposing is a new sort of compilation unit. In
>     the draft spec we call this a “simple" compilation unit, to
>     contrast with “ordinary”. As no human says the phrase “compilation
>     unit”, but rather everyone informally speaks of “source files”, we
>     ended up calling these new sorts of source files “simple”.
>
>
>  I guess this ship has sailed, but..
>
> I just watched Gavin's Language Futures talk from Devoxx, and realized 
> I would have preferred the adjective "implicit" used in JEP-477 over 
> "simple". "Implicit" implies a class without an explicit class 
> declaration, whereas "simple" implies everything and nothing at once.
>
> Cheers,
> Eirik,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/amber-dev/attachments/20241114/b5af0238/attachment.htm>