🚨 🦮 Seven ways to test for accessibility using only what is already in browser developer tools of Chromium browsers https://t.co/C7kdbigHGE

@MSEdgeDev @EdgeDevTools @ChromiumDev
#tools #accessibility #browsers
Also, a thread: 👇🏼

I sympathize with mid-size companies seeking to acquire a premium domain name. Even if they stretch to make the maximum possible offer, it often still can't compete with what a larger company can offer. Their desired domain name remains always out of reach. /1

From the broker:

"The prospective buyer responded and explained that their absolute maximum amount they can spend is $100,000 USD. They are accountants and explained this is an extraordinary amount of money for them." /2

The distribution of how much a domain name is worth to a particular company may look approximately like the right half of a bell curve, with values increasing along the X-axis and the # of companies for which the domain name has that value on the y-axis. /3

Since each domain name is unique, a domain owner wishing to maximize the value of that asset will often wait for an offer in the (light green) rightmost tail of the distribution, from that small percentage of companies for whom the domain name will deliver the most value and /4

as a result can justify making the highest value offers for the domain name. But that means that the domain name will remain out of reach for the vast majority of companies that would benefit from owning the domain name. For even a company at the 80th percentile, in the /5

There has been speculation that changes to the way the internet is implemented will be rolled out to allow access only to authorized sites.

Some think this is impossible

I'm here to tell you it is not, and very likely.

Allowing access only to white listed IP addresses is easy.

The vast majority access the internet via their Internet Service Provider, and ISP's can easily block client access to any IP address they choose.

China has been doing this for years.

This is how firewalls work, and is much lower level than DNS blocking.

I would imagine this will be rolled out to 'Combat the ever growing threat of Cyber Attacks'.

Most people only use a few thousand services like Twitter, Facebook, Amazon, Google, Microsoft, TikTok, Snapchat, Instagram, Reddit etc, Walmart, Costco, Tesco, Morrisons etc. as it is.

It will be easy over time to require 'legal registration' for any and all other services and sites.

We already took a step toward this years ago anyway, so this is just the logical next step.

As 5G technology is rolled out in your city, more and more control is being introduced at the street level, using what are known as 'Edge Servers'.

While making services faster to access, it also allows an additional layer of security and access control at the local level.

I'll touch upon few different types of cyber frauds and few possible ways in which we can avoid ourselves to be duped of money, time & energy.

Will also attach few real registered FIRs showing that most of the times such crimes happen with educated & learned people.

1/16

Cyber Fraud type 1 - Facebook/Instagram

Profile cloning -Fraudster will make exact same profile like yours - using pictures from your own profile. They will then message to your contacts asking for money immediately for the surgery of mother or similar emotional lie

2/16

Always verify before sending money.
Don't verify by calling that person, instead, call a common friend or family member of that person. Don't think you can't be fooled - I know at least 20 people working in Top IT MNCs becoming victim of this fraud. Lock your FB/Insta.

3/16

Cyber Fraud type 2 -OLX/Quikr
This generally happens while selling something on OLX/Quikr & similar platforms. In such cases, people will fraud in the name of paying money by asking to scan the QR code or click a link.
NEVER scan the QR code to receive money, give your UPI

4/16

Cyber Fraud type 3 - Fake SMS & Call from unknown numbers
Fraudster will send fake SMS showing amount credited to the account - with correct account number in SMS. Then they will call from unknown numbers asking for money.

ALWAYS check your bank account first via bank app.

5/16

Here's an overview of key adoption metrics for deep learning frameworks over 2020: downloads, developer surveys, job posts, scientific publications, Colab usage, Kaggle notebooks usage, GitHub data.

TensorFlow/Keras = #1 deep learning solution.

Note that we benchmark adoption vs Facebook's PyTorch because it is the only TF alternative that registers on the scale. Another option would have been sklearn, which has massive adoption, but it isn't really a TF alternative. In the future, I hope we can add JAX.

TensorFlow has seen 115M downloads in 2020, which nearly doubles its lifetime downloads. Note that this does *not* include downloads for all TF-adjacent packages, like tf-nightly, the old tensorflow-gpu, etc.

Also note that most of these downloads aren't from humans, but are automated downloads from CI systems (but none are from Google's systems, as Google doesn't use PyPI).

In a way, this metric reflects usage in production.

There were two worldwide developer surveys in 2020 that measured adoption of various frameworks: the one from StackOverflow, targeting all developers, and the one from Kaggle.