[SERVER-566] Don't require spidermonkey for building libmongoclient Created: 22/Jan/10  Updated: 12/Jul/16  Resolved: 08/Mar/10

Status: Closed
Project: Core Server
Component/s: Internal Client
Affects Version/s: 1.2.0, 1.2.1
Fix Version/s: 1.3.4

Type: Improvement Priority: Minor - P4
Reporter: Ash Berlin Assignee: Mathias Stearn
Resolution: Done Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

Boost 1.41, OSX 10.6


Participants:

 Description   

Attempting to link against the libmongoclient.a provided in the 1.2.0 and 1.2.1 release tarballs yield unresolved externals when using a different version of boost. This is apparently expected, and I was encouraged to compile it myself.

The crux of the matter is that compiling the client lib shouldn't require spidermonkey. A simple hacky change to the SConstruct file:

diff --git a/SConstruct b/SConstruct
index 5d0b835..5029467 100644
--- a/SConstruct
+++ b/SConstruct
@@ -260,8 +260,8 @@ if ( usesm and usejvm ):
     print( "can't say usesm and usejvm at the same time" )
     Exit(1)
 
-if ( not ( usesm or usejvm or usev8 ) ):
-    usesm = True
+#if ( not ( usesm or usejvm or usev8 ) ):
+#    usesm = True
 
 if GetOption( "extrapath" ) is not None:
     for x in GetOption( "extrapath" ).split( "," ):
 
 

compiles fine and links into a program fine. Its obviously not the right fix, but it does show that it isn't actually used in the client code.

In addition, perhaps libmongoclinet.a shouldn't be distributed in the tarballs, or it should at least specify which version of boost it was compiled against.



 Comments   
Comment by auto [ 08/Mar/10 ]

Author:

{'login': '', 'name': 'Mathias Stearn', 'email': 'mathias@10gen.com'}

Message: Don't require spidermonkey for building client lib SERVER-566
http://github.com/mongodb/mongo/commit/434a2e4b72b6dae16d90fc990a047da4ef39a477

Comment by Eliot Horowitz (Inactive) [ 03/Mar/10 ]

not sure this is even really possible without massive work - but take a look.
we need to clean up SConstruct at some point.
I know richard was looking at it - so may want to talk to him

Generated at Thu Feb 08 02:54:32 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.