Re: [collections] Revert a performance related fix in 4.1

2015-01-26 Thread James Carman
On Mon, Jan 26, 2015 at 7:03 AM, Benedikt Ritter wrote: > Hello Adrian > > 2015-01-24 19:43 GMT+01:00 Adrian Crum : > >> Slightly off-topic but somewhat related... >> >> I saw a recent commit where a "performance improvement" went something >> like this: >> >> StringBuilder sb = new StringBuilder(

Re: [collections] Revert a performance related fix in 4.1

2015-01-26 Thread Benedikt Ritter
2015-01-26 14:05 GMT+01:00 Gary Gregory : > On Mon, Jan 26, 2015 at 7:03 AM, Benedikt Ritter > wrote: > > > Hello Adrian > > > > 2015-01-24 19:43 GMT+01:00 Adrian Crum < > adrian.c...@sandglass-software.com > > >: > > > > > Slightly off-topic but somewhat related... > > > > > > I saw a recent com

Re: [collections] Revert a performance related fix in 4.1

2015-01-26 Thread Gary Gregory
On Mon, Jan 26, 2015 at 7:03 AM, Benedikt Ritter wrote: > Hello Adrian > > 2015-01-24 19:43 GMT+01:00 Adrian Crum >: > > > Slightly off-topic but somewhat related... > > > > I saw a recent commit where a "performance improvement" went something > > like this: > > > > StringBuilder sb = new Strin

Re: [collections] Revert a performance related fix in 4.1

2015-01-26 Thread Benedikt Ritter
Hello Adrian 2015-01-24 19:43 GMT+01:00 Adrian Crum : > Slightly off-topic but somewhat related... > > I saw a recent commit where a "performance improvement" went something > like this: > > StringBuilder sb = new StringBuilder(); > sb.append("foo"); > > was replaced with > > StringBuilder sb = n

Re: [collections] Revert a performance related fix in 4.1

2015-01-24 Thread Phil Steitz
On 1/24/15 1:16 PM, Oliver Heger wrote: > Hi Thomas, > > On 24.01.2015 19:21, Thomas Neidhart wrote: >> Hi, >> >> from time to time some researchers trying to find performance >> bugs in >> open-source software create issues for collections. >> >> One of the easy targets is the Collection#retainAll

Re: [collections] Revert a performance related fix in 4.1

2015-01-24 Thread Oliver Heger
Hi Thomas, On 24.01.2015 19:21, Thomas Neidhart wrote: Hi, from time to time some researchers trying to find performance bugs in open-source software create issues for collections. One of the easy targets is the Collection#retainAll(Collection) method as the default implementation in AbstractC

Re: [collections] Revert a performance related fix in 4.1

2015-01-24 Thread Adrian Crum
Slightly off-topic but somewhat related... I saw a recent commit where a "performance improvement" went something like this: StringBuilder sb = new StringBuilder(); sb.append("foo"); was replaced with StringBuilder sb = new StringBuilder("foo"); The change reduced the code by one line, but

[collections] Revert a performance related fix in 4.1

2015-01-24 Thread Thomas Neidhart
Hi, from time to time some researchers trying to find performance bugs in open-source software create issues for collections. One of the easy targets is the Collection#retainAll(Collection) method as the default implementation in AbstractCollection calls contains on the provided collection. Now,