Repeated make code about python-matlab-bridge HOT 10 CLOSED

Sorry for the delayed response. Users are not supposed to run these make
scripts. The intention of these scripts is that we ourselves run them on
each of the platforms and provide an already compiled binary. If we change
the name of these scripts would that change your expectations about them
and be less confusing?

On Mon, Apr 14, 2014 at 8:10 PM, Aron Ahmadia [email protected]:

In the messenger directory, the file make.py appears to be copied (with
very minor modifications) separately into the Linux and OS X build
directories. It's not copied into the Windows directory, making for an
inconsistent build experience.

The intention and usage of the make.py bootstrap script should be
clarified.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/66
.

from python-matlab-bridge.

I'm a "user" in the sense that I need to build these binaries myself. For maintainability purposes, it probably makes sense to only have one of these. I'll submit a PR if I've got some reasonable changes.

from python-matlab-bridge.

Yes - I agree that a reorganization of that code could easily remove some
redundancy.

Why do you have to build these yourself? Can't you just use the binaries we
compiled?

On Sat, Apr 26, 2014 at 6:19 PM, Aron Ahmadia [email protected]:

I'm a "user" in the sense that I need to build these binaries myself. For
maintainability purposes, it probably makes sense to only have one of
these. I'll submit a PR if I've got some reasonable changes.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/66#issuecomment-41485668
.

from python-matlab-bridge.

@cekees has been recompiling them in our stack. I assume the mex bindings are specific to the libraries being used.

from python-matlab-bridge.

Yes, I wanted to make pymatbridge a fully functional package within
hashdist so I was building the messenger extension so that it points to
libzmq, etc. from the active hashdist profile.

On Sun, Apr 27, 2014 at 12:13 AM, Aron Ahmadia [email protected]:

@cekees https://github.com/cekees has been recompiling them in our
stack. I assume the mex bindings are specific to the libraries being used.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/66#issuecomment-41488678
.

from python-matlab-bridge.

What were you doing to build? Were you using our make.py scripts? Was it
fairly straight-forward?

On Mon, Apr 28, 2014 at 7:51 AM, Chris Kees [email protected]:

Yes, I wanted to make pymatbridge a fully functional package within
hashdist so I was building the messenger extension so that it points to
libzmq, etc. from the active hashdist profile.

On Sun, Apr 27, 2014 at 12:13 AM, Aron Ahmadia [email protected]:

@cekees https://github.com/cekees has been recompiling them in our
stack. I assume the mex bindings are specific to the libraries being
used.

—
Reply to this email directly or view it on GitHub<
https://github.com/arokem/python-matlab-bridge/issues/66#issuecomment-41488678>

.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/66#issuecomment-41567506
.

from python-matlab-bridge.

Yes, the only thing that hasn't been straightforward is the cygwin build,
because I was trying to use gnu compilers and matlab doesn't have defaults
for building mex files with gnu on cygwin, though it can be done.

On Mon, Apr 28, 2014 at 9:53 AM, Ariel Rokem [email protected]:

What were you doing to build? Were you using our make.py scripts? Was it
fairly straight-forward?

On Mon, Apr 28, 2014 at 7:51 AM, Chris Kees [email protected]:

Yes, I wanted to make pymatbridge a fully functional package within
hashdist so I was building the messenger extension so that it points to
libzmq, etc. from the active hashdist profile.

On Sun, Apr 27, 2014 at 12:13 AM, Aron Ahmadia [email protected]:

@cekees https://github.com/cekees has been recompiling them in our
stack. I assume the mex bindings are specific to the libraries being
used.

—
Reply to this email directly or view it on GitHub<

https://github.com/arokem/python-matlab-bridge/issues/66#issuecomment-41488678>

.

—
Reply to this email directly or view it on GitHub<
https://github.com/arokem/python-matlab-bridge/issues/66#issuecomment-41567506>

.

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/66#issuecomment-41567736
.

from python-matlab-bridge.

This #73 should help. Could you take a look?

from python-matlab-bridge.

I'm calling this one fixed by #144. Feel free to submit any cygwin specific changes @cekees.

from python-matlab-bridge.

Thanks. I haven't gotten around to updating pymatbridge in a while. I
appreciate all the effort you guys are pointing into it and will try to get
our stack updated to a recent master at some point in the next month or so.

On Wed, Feb 18, 2015 at 9:44 PM, Steven Silvester [email protected]
wrote:

I'm calling this one fixed by #144
#144. Feel free to
submit any cygwin specific changes @cekees https://github.com/cekees.

—
Reply to this email directly or view it on GitHub
#66 (comment)
.

from python-matlab-bridge.