Skip to main content

Posts

Showing posts from 2018

Integrating Firefox and Selenium Webdriver

I am posting this on my blog to remind me what to do if I have to integrate firefox, and selenium webdriver.

firstly, install selenium using pip.

and then, download the geckodriver that should be compatible / working with selenium webdriver.

In your code, if your python/django app is in virtualbox (headless mode),

it should be something like this:

-------------------------------------------------------------------------
#!/srv/onlinesystem/env/bin/python

from selenium import webdriver
from selenium.webdriver import FirefoxOptions

opts = FirefoxOptions()
opts.add_argument("--headless")
browser = webdriver.Firefox(firefox_options=opts)
browser.get('http://google.com')


Screen Recording with Cam Studio (Open Source)

I tried to use cam studio for the first time to do a screen recording with my colleague while he was showing me something.

After our conversation, I quickly checked if the recording was successful.

The video was there. OK.

When I played it, it was not playing.

And then it said, the file was corrupted.

I googled for an explanation and realized that it's better to record a specific region and not the whole screen because the video would become very large. My screen recording video was 4GB. Look here for more details.

Learning TensorFlow

Just recently, I am so inspired to learn Tensorflow and DeepSpeech by Mozilla to work on a personal project.

I learned that to install and use DeepSpeech, it is best to use Mozilla's version of Tensorflow and compile it from source. Link to github is here. Link to DeepSpeech is here.

I would suggest to install them by compiling from source, rather than installing it using pip.
I encountered error messages like,

>>> tf.Session()
2018-01-05 17:19:01.747275: I tensorflow/core/platform/cpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: SSE4.1 SSE4.2 AVX
<tensorflow.python.client.session.Session object at 0x7f02a0b3ae48> I was following a youtube video of a tutorial about Tensorflow and this error message showed up.

Another error message also showed up when I installed DeepSpeech:

 (env) $ deepspeech
2018-01-05 17:58:21.092975: F tensorflow/core/platform/cpu_feature_guard.cc:35] The TensorFlow library was compile…

Popper is not defined

I was studying how to integrate webpack to my personal project which is a Django App. I was able to install bootstrap, and then JQuery, but it also requires Popper. I encountered an error from npm saying,
unmet Dependency Popper^@1.12.9  I googled for a solution but I wasn't able to find one. Until I read this answer from StackOverflow, saying that you need to specify
npm install Popper.js --save  It seems that Popper and Popper.js are not the same.