summaryrefslogtreecommitdiffstats
path: root/src/sp-string.cpp
diff options
context:
space:
mode:
authorLiam P. White <inkscapebrony@gmail.com>2015-02-25 01:48:17 +0000
committerLiam P. White <inkscapebrony@gmail.com>2015-02-25 01:48:17 +0000
commit9b72cebb7ccb500cf447bf75b66fe77875d9945a (patch)
tree5110bf55e68dee4b60030f2008a7127ff347c065 /src/sp-string.cpp
parentTranslations. Hungarian translation update. (diff)
parentRestore libinkscape.a (diff)
downloadinkscape-9b72cebb7ccb500cf447bf75b66fe77875d9945a.tar.gz
inkscape-9b72cebb7ccb500cf447bf75b66fe77875d9945a.zip
Restore libinkscape.a .
I have a feeling this will be a controversial commit, and I feel the same way: I liked the way the SPObject factories were set up. However, they placed undue stress on the linker to extract every single symbol from every object file, and this isn't something that it easy for the linker: it continually thrashes my system with 8 GB of RAM for more and more memory, with my system often times running out. This is not the only solution, but for now, it's quite a good one, and the comment in Makefile.am remains true: libinkscape.a does speed up the build. In the future, I'd hope to see proper code modules and an incremental link, which should really help speed up the build. (bzr r13940)
Diffstat (limited to 'src/sp-string.cpp')
-rw-r--r--src/sp-string.cpp17
1 files changed, 0 insertions, 17 deletions
diff --git a/src/sp-string.cpp b/src/sp-string.cpp
index b561187d0..26bb44006 100644
--- a/src/sp-string.cpp
+++ b/src/sp-string.cpp
@@ -22,30 +22,13 @@
*
*/
-
-
-#ifdef HAVE_CONFIG_H
-# include "config.h"
-#endif
-
-
#include "sp-string.h"
#include "style.h"
#include "xml/repr.h"
-#include "sp-factory.h"
-
#include <iostream>
-namespace {
- SPObject* createString() {
- return new SPString();
- }
-
- bool stringRegistered = SPFactory::instance().registerObject("string", createString);
-}
-
/*#####################################################
# SPSTRING
#####################################################*/