Re: Draft JEP for upcoming work on snippets

2021-01-26 Thread Pavel Rappo
Roger, I'd add to what Jon has already said about indentation. Early on in our prototype we recognized the need for clear rules and convenient control for indentation. We knew that having those would allow "Snippets" to improve on the solution provided by the "pre-code" compound: {@code

Re: Draft JEP for upcoming work on snippets

2021-01-26 Thread Jonathan Gibbons
On 1/26/21 7:47 AM, Roger Riggs wrote: * In the @snippet tag, can the "class=" indicate a module name in addition to the class name? A related question is whether and when it makes sense to specify a module name. The current thinking is that external snippets are searched for in 3 lo

Re: Draft JEP for upcoming work on snippets

2021-01-26 Thread Jonathan Gibbons
On 1/26/21 7:47 AM, Roger Riggs wrote: Hi, A very useful proposal. A few comments, I expect these will be addressed as the development proceeds. * In the @snippet tag, can the "class=" indicate a module name in addition to the class name? We'll need to address that. Ideas are to eith

Re: Draft JEP for upcoming work on snippets

2021-01-26 Thread Roger Riggs
Hi, A very useful proposal. A few comments, I expect these will be addressed as the development proceeds. * In the @snippet tag, can the "class=" indicate a module name in addition to the class name? * Does the "lang=" tag need to be a closed set; if it were open, it would give more f

Re: RFR: JDK-8255059: Regressions >5% in all Javadoc benchmarks in 16-b19 [v2]

2021-01-26 Thread Hannes Wallnöfer
On Mon, 25 Jan 2021 14:25:59 GMT, Hannes Wallnöfer wrote: >> I have narrowed down the performance regression to the modularity check >> introduced in JDK-8240169. Since this check is not necessary if we know that >> the element list's modularity matches that of the library we can omit the >> m