Expected distribution of lambda sizes (Re: Syntax poll, take 2)
sam at sampullara.com
Fri Jun 17 09:17:26 PDT 2011
There are some people that program with this style even though the syntax isn't there, like myself (relying on the IDE to generate most of the boilerplate). In my current project I have 158 .java files that generate 1113 .class files. 10+ lambda type usages of anonymous classes (implementing interfaces like Function and Predicate) are fairly commonplace in this codebase. Additionally, I make heavy use of the Google Guava APIs which are designed to be used using lambda type expressions. It isn't clear to me how many examples there are in the wild, but I could give some numbers on what it would look like at a minimum.
On Jun 17, 2011, at 12:53 AM, Neal Gafter wrote:
> Alex's comment makes sense to me: we cannot reasonably expect to glean the
> expected size of lambda bodies from an analysis of existing code bases that
> both lack lambda expressions and lack APIs that were designed in the
> presence of a language containing lambda expressions.
> On Thu, Jun 16, 2011 at 11:49 PM, Reinier Zwitserloot <
> reinier at zwitserloot.com> wrote:
>> On Thu, Jun 16, 2011 at 9:36 AM, Alex Blewitt <alex.blewitt at gmail.com
>>> On 16 Jun 2011, at 08:03, Reinier Zwitserloot <reinier at zwitserloot.com>
>>>> Using the apparent aggregate preferences of the greater java community,
>>>> gleaned by looking at a lot of open source code, I conclude that the
>>>> java community will likely end up putting a lot more linebreaks in
>>>> bodies than we are doing with these examples.
>>> Using this logic, we can conclude that no-one in the greater Java
>>> will be interested in lambdas, since looking at open source code shows no
>>> sign of lambda use at the moment.
>> This flippant commentary makes no sense and does absolutely nothing to
>> further the discussion.
More information about the lambda-dev