On 16/06/18 13:59, Dennis Lee Bieber wrote:
> On Sat, 16 Jun 2018 12:49:30 +0100, RobH  declaimed the
> following:
>
>
>>from os.path import join < I added this
>>
>  Why? Was typing os.path.join too much effort?

No it wasn't actually, as I said in another reply on here, I have been
reading up on the shutil module, and I thought to try some of the
examples I looked at.

>
>> src = ("RAMname")< and this
>> dst = ("NASname")< this
>
>  These lines do nothing useful since you never use the names src and dst
> later (except to replace the contents and again not use them)

No I have removed them now
>
>
>> while True:
>>      # Wait for motion being detected
>>      pir.wait_for_motion()
>>      triggerTime = time.time()
>>
>>      #create base filename at time of motion trigger
>>      fname = time.strftime("%Y-%m-%d_%H.%M.%S.h264",
>> time.localtime(triggerTime))
>>
>>      RAMname = os.path.join("/run/shm", fname)
>>      NASname = os.path.join("/mnt/CCTV/PiZero" , fname)
>>
>>      camera.start_preview()
>>
>>      #capture video to RAM device first
>>      camera.start_recording(RAMname)
>
>  I suggested trying NASname in that call, to directly capture to the end
> device rather than buffering in RAM first and then copying. If that fails,
> then the problem is not inherent to the script but is something wrong in
> the environment. RAMname (and later the shutil copyfile) was proposed to
> let the RPi capture to RAM without the overhead of network traffic (given
> that the network interface of an RPi is via the USB system)
>
>>      camera.wait_recording(10)
>>      camera.stop_recording()
>>      camera.stop_preview()
>>
>>      #copy from RAM device to NAS device
>>      #shutil.copyfile(RAMname, NASname)
>>
>>      src = join('/', RAMname) #new lines here < I added this
>>      dst = join('/', NASname) < and this
>
>  Similarly, these do nothing useful since "src" and "dst" are never
> referenced in the code... And what they are doing is just trying to add
> another / to the front of the names generated earlier. But since / is the
> path separator in Linux, and .join() puts the components together by
> putting a path separator between them...

Ok , I've removed them
>
> -=-=-=-=-
>>>> import os.path
>>>> import time
>>>> fname = time.strftime("%Y-%m-%d_%H.%M.%S.h264", time.localtime())
>>>> fname
> '2018-06-16_08.45.24.h264'
>>>> nname = os.path.join("/mnt/CCTV/PiZero", fname)
>>>> nname
> '/mnt/CCTV/PiZero/2018-06-16_08.45.24.h264'
>>>> dst = os.path.join("/", nname)
>>>> dst
> '/mnt/CCTV/PiZero/2018-06-16_08.45.24.h264'
>>>> dst == nname
> True
>>>>
> -=-=-=-=-
>
> ... there is no effect
>
>>
>>
>>      #delete file from RAM device
>>      os.unlink(RAMname)      #os.remove() is same function
>
>  I also asked you to comment that out... Since if the capture is going
> to NASname, there is no file in RAMname to be deleted

Yes I did do that last night, but there was some sort of error. I can't
recall it now as it was late.
>
>>
>>     #set up next check time
>>      time.sleep((triggerTime + 20.0) - time.time())
>>
>
>  While picking up that beginners book on Linux, pick up a book on Python
> also...
>
>

I have now commented out what you asked me to and also:

camera.start_recording(NASname)

The script is running without errors, and producing video files of
around 5 or 6MiB on the PiZero, but nothing of todays date on the NAS box.
So the script is producing the files but not moving/copying them accross
to the NAS box

--- SoupGate-Win32 v1.05