Skip to content

[Bioc-devel] rtracklayer: BigWigFIle on Windows

4 messages · Kasper Daniel Hansen, Michael Lawrence

#
In the man page for trackLayer::BigWigFile it says "These functions do not
work on Windows".  Despite this statement, which seems to have been around
for a while, I and several of my students seems to be able to execute
something like
  import(bwFile, which = my.gr, as = "Rle")
on Windows on a BigWig file.  In my case the OS is Windows 7 run under a
virtual machine on a Mac.  In the virtual machine I don't have full
support.  For example, the code above works when bwFile is a BigWigFile
pointing to a BigWig file stored on my local hard drive, but does not work
when I query a url; in that case I get a weird error message.  However, I
have at least one student for which this does not seem to work on Windows.

I was unaware of this when I designed a quiz problem involving importing a
BigWigFile.  I am surprised that it works at all.  What is the current
state of BigWig file support on Windows? Clearly some of it works on some
systems.

Best,
Kasper
#
I would say that it's unsupported, and surprising that it actually worked
for you. It would require some non-trivial engineering to get it fully
working on Windows. For example, to access data via HTTP, one would need to
implement or integrate an HTTP client based on Windows sockets, as well as
abstract Jim Kent's code, which is based on POSIX-specific things like
integer file descriptors.


On Thu, Sep 17, 2015 at 1:09 PM, Kasper Daniel Hansen <
kasperdanielhansen at gmail.com> wrote:

            

  
  
#
Thanks.

While fetching data over the internet is pretty nice, for me there is a big
difference between being able to read local BigWig files and not being able
to read anything at all.  Wrt. your answer I am a bit unclear whether you
expect it to work on a machine local file on Windows.

Best,
Kasper

On Thu, Sep 17, 2015 at 6:14 PM, Michael Lawrence <lawrence.michael at gene.com

  
  
#
I've never tested it, and reports up until now have been negative. It would
help to know the error message (if any) and characteristics of the systems
that work or don't work. Given that it works in some circumstances, it
sounds like we could make it work more reliably.

Michael

On Thu, Sep 17, 2015 at 8:16 PM, Kasper Daniel Hansen <
kasperdanielhansen at gmail.com> wrote: