[Bioc-sig-seq] GenomicFeatures package

[Steve Lianoglou]

Hi Seth,

On Nov 2, 2009, at 12:47 AM, Seth Falcon wrote:

> On 10/31/09 4:14 PM, Michael Lawrence wrote:
>> On Sat, Oct 31, 2009 at 2:15 PM, Martin Morgan<mtmorgan at fhcrc.org>   
>> wrote:
<snip>
>> I'm wondering if the custom compile is really necessary? Is it not  
>> possible
>> for rsqlite to provide that feature by default?
>
> As far as I know, recent versions of RSQLite have the RTree support  
> enabled.  The following, taken from the RTree documentations (http://www.sqlite.org/rtree.html 
> ), seems to work (RSQLite_0.7-3 DBI_0.2-4):
<snipped working example>

Wow .. I didn't even check to see if it compiles like that be default,  
that's great.

I assumed that since it *doesn't* compile by default in the sqlite  
distro, I expected it to not do so in the RSQLite install since "- 
DSQLITE_ENABLE_RTREE=1" wasn't added to the src/Makevars.in  
PKG_CPPFLAGS entry, which is currently:

PKG_CPPFLAGS=@PKG_CPPFLAGS@ -DTHREADSAFE=0

But searching the package source again, I see that it is defined  
somewhere deep in the configure file. I'm not sure what codepath has  
to be "active" to land in the particular `if` block that that sets - 
DSQLITE_ENABLE_RTREE=1 (my shell scripting mojo is quite weak).

Just to be sure, would it be better/safer to assume (by those of us  
who might want to rely on this feature) that it's always compiled this  
way if this flag is also set in the src/Makevars.in/PKG_CPPFLAGS line  
as well?

Lastly, can I humbly appeal to include a more recent version of the  
sqlite src in the RSQLite package for the future?

It looks like the version included is 3.6.4, and we're up to 3.6.19  
with a plethora of bug fixes and performance enhancements happening  
between the two versions.

Thanks,
-steve

--
Steve Lianoglou
Graduate Student: Computational Systems Biology
   |  Memorial Sloan-Kettering Cancer Center
   |  Weill Medical College of Cornell University
Contact Info: http://cbio.mskcc.org/~lianos/contact