‘Brave’ the new open source web browser based on Chromium. With inbuilt ad-blocking and a whole mindset of privacy, I am wondering if it will organically change the face of marketing on the internet and thus the internet as we know it?
Most advertising on web pages is loaded with tracking. There is nothing all that sinister about it. It’s just people trying to capitalise on free data. However it has made a mess of the usability of the web.
Brave web browser allows you to choose how many adds it shows per time frame and whether those adds are ethical in that they are not tracking your usage in some way.
Many business people I know have given up advertising on the web for it has become too expensive relative to the return. They are going back to other mediums to advertise and doing better there. I am included in this. Bang per buck and per hour effort has not produced the results for me compared to offline methods.
As a friend in marketing put it “It’s difficult now to get past the noise”.
Once again the engineers back in the 60’s did a great job of creating TCP/IP (the language and structure of the web). They created it to survive a nuclear war and thus it easily handles any corporate or government attempts to control it.
So I wonder how the net will look in the coming years as companies stop ‘in your face’ advertising killed by Brave?
All I can say is WOW, NextCloud (NC) rocks.! I wrote the last post in 2018 and it’s now late 2020 and the NextCloud server is still cranking on a Raspberry Pi.
I’ve done one upgrade to NC and that went smoothly. Webdav was a touch slow, so I went back to SAMBA. Especially for communal data running in a NAS. When adding data from your LAN via SAMBA the NextCloud database does not see the new files. Thus I made a simple shell script to run nightly to update the NextCloud database.
--------------------- filescan.sh - START
# Last updated by DP 31st March 2020
sudo -u www-data php occ files:scan --path="shared/files/"
--------------------- filescan.sh - END
Create it with root permissions. Set the permissions to make it executable.
chmod 500 filescan.sh
Should return something like
Starting scan for user 1 out of 1 (%user_name%)
| Folders | Files | Elapsed time |
| 8069 | 186532 | 01:52:36 |
02 03 * * * /var/www/html/filescan.sh
to cron to run filescan.sh at 03:02 daily.
It should take 20-25mins for 5k files’ish.
I still use Rsync to do second and third backups to offsite drives, but as far as a reliable, easy, cross-platform, cross-device system, it’s by far the best I’ve used.
For this project the customer wanted TeamViewer running on a Raspberry Pi Nano so that a programmer he hired from interstate could connect easily to get some scripts running. For the more tech savvy an SSH terminal connection is by far more solid and runs easily on a Pi Nano, but the configuration is harder as a port forward would have to be setup on the customers home or business router and then DNS resolution has to happen. In the end that set up would be more fault tolerant, but then again it depends on how you look at it.
TeamViewer has almost become an industry standard as it is generally fast, reliable and easy to set up. I’ve not done a speed test between TeamViewer and FreeNX, but for ease of use TeamViewer is great. It also bypasses the firewall using tunneling without compromising security.
Back to the job at hand. It turns out that the older older ARM chips found in the Nano and the Pi 1 do not support SIMD (Single Instruction Multiple Data). SIMD is a type of parallel computing involving vectors (beyond the scope of this article). Either way the ARM processors in the Pi3 have an architecture called NEON which supports SIMD.
After 4ish hours of flattening the OS, reinstalling Raspbian, hunting for the appropriate version of TeamViewer host for Linux and then figuring out what the spurious errors meant I finally realised the NEON SIMD issue above. I popped the SD card out of the Raspberry Pi Nano and put it in a Pi3 B+ and away it went. Mission accomplished.
I have to say, I’m impressed so far. A few weeks ago I installed NextCloud server on a Raspberry Pi 3b. So far it is working very solidly. The Android apps and the website interface work very well and are highly polished. The OS X client is very smooth and the Linux Mate client is nearly as good as the OS X one. It’s only missing the little green / blue dot icons in the Caja file browser to notify of synchronisation. It’ll take some months to make a true assessment of NextCloud, but I am fairly confident it will keep out stripping my expectations.
For instance, I wanted to make a standard network share for our client. I was going to set up Samba (windows file sharing on Linux servers). Samba works well. I’ve used it for years. It was to point to the same NextCloud share. However there was no need to set up Samba as NextCloud uses a protocol (communication language) called DAV. Simply create a
The security systems seem good and you can make it as hard as you want. Even to the point of using two factor authentication using SMS. How nice it that?
Why did I install it on a Raspberry Pi and not a more powerful solid server or a VPS?
The client does not add a lot of data per day
The Pi uses less than 2 watts of energy (+ hard drive power)
It’s not a huge expense to make a replicated mirrored backup server with automatic fall over protection.
Having it onsite means the client knows physically where their data is at all times.
Using Bind9 with NSupdate makes it possible to connect to NextCloud from any internet connection, so why not.
It just makes sense. Most data conscious companies and business owners I speak to just want to control their own data. The phrase “just stick it in the Cloud…” is too ambiguous for them. Even though there are so many advantages to the technology of Cloud based systems, they still feel unsettled knowing that the core of their business is reliant on some other data company. Now systems like NextCloud are possibly giving an alternative way to using this tech.