Re: Regression and failure to clone/fetch with new code Re: git-svn performance

From: Hin-Tak Leung <htl10@users.sourceforge.net>
Date: Oct 31, 2014, 02:30 UTC
Message-ID: <1414722617.89476.YahooMailBasic@web172305.mail.ir2.yahoo.com>

(apology for the empty message again - sticky fingers in smart phone...)

------------------------------ On Thu, Oct 30, 2014 08:46 GMT Eric Wong wrote:

Show 24 quoted lines

>Thanks, I'm not able to reproduce the issue, but can you try the
>following?
>
>diff --git a/perl/Git/SVN/Ra.pm b/perl/Git/SVN/Ra.pm
>index 75cdac9..82d6108 100644
>--- a/perl/Git/SVN/Ra.pm
>+++ b/perl/Git/SVN/Ra.pm
>@@ -153,6 +153,7 @@ sub url {
> sub check_path {
>     my ($self, $path, $r) = @_;
>     my $cache = $self->{cache}->{check_path};
>+    $r = int($r);
>     if ($r == $cache->{r} && exists $cache->{data}->{$path}) {
>         return $cache->{data}->{$path};
>     }
>@@ -169,6 +170,7 @@ sub check_path {
> sub get_dir {
>     my ($self, $dir, $r) = @_;
>     my $cache = $self->{cache}->{get_dir};
>+    $r = int($r);
>     if ($r == $cache->{r}) {
>         if (my $x = $cache->{data}->{$dir}) {
>             return wantarray ? @$x : $x->[0];
>---

I added those two lines to my git and there is no improvement. It still won't svn fetch the next revision. I think it may be important/interesting to find out when or how it becomes non-int, so I have tar'gz'ed my wont-fetch virtual box .git and in the middle of uploading here:

http://sourceforge.net/projects/outmodedbonsai/files/R/

I am also uploading my old R clone also - maybe you'd like to see why its .git/svn/.caches is so big compared to a recent one, as well as how and why there were an extra merge and two missing merges compared to a recent clone?

(FYI - I have some local stuff on my old R clone. They are mostly there because the R devs don't agree etc but there is nothing confidential about them).