A list of base hack like content grouping, user id, custom user properties and UX metric like page generation time and page load time

Continue discussion about using static files as database instead of default relational MySql/Postgree/Mongo

For a friend I made this: https://en.hayovasweets.com/

It’s mostly list of markdowns and images in folders. Everything else generates during deployment on GitHub:actions

It was simpler to show GitHub desktop rather than installing something like Wordpress machinery monster and spent hours to teach how to use it. Now the process is simple:

1. Create a folder

2. Put there markdown and images

3. Commit

In this short study, we will analyze the Instagram pages of five world-class musicians - Michael Jackson, Amy Winehouse, Freddie Mercury, Bob Marley, Prince.

By the way. Did anyone try to use a large sd-card as storage ?

I meant 256-512GB. In my logic it must decrease power consumption when keeping the same IO performance as external ssd connected through USB.

I am uncertain if USB can be a bottleneck there, but if I am not mistaken, SSDs offer better (more?) cache. This dramatically reduces time needed for non-sequential read/write operations.

USB - 30Mbyte/sec theoretically for RPI, but in reality somewhere around 20Mbyte/sec

SD cards - class 10 tested on RPI show the same performance 21-27Mbyte/sec

So the issue mostly in throughput performance of the interface

Face Palm? Face Down ? Face Epic Fail - after acquisition of Epic Games

I tried but I can't be serious about this news. Th company must name : Epic Face :)))))

Facepic?

IP blocked for non-residential areas eq. VPN/datacenters are blocked by loginwall. This is super stupid idea to fight with web scrappers.

The issue with web-scrappers comes from API limits: none wanted to parse twitter web if you can get everything from API for a reasonable price.

I don't think API access brings lots of money 5% increase in profit costs customer experience on mobile.

Looks like Redis is trying to promote itself as a session data storage. From my perspective:

1. There's no difference between stealing jwt or session_id with al the problems it brings

2. Session in external storage is a solution for small projects like personal blog etc. because you have almost no traffic to handle. FOr example: 1kk session with an average size 1k = 1kk * 1k = 1Gb in memory storage. Not so big but you will meet all the issues of 10k connections problems when trying to build a service witch can handle this also it will consume at least 100Mb/s bandwidth just for getting/setting data.

3. Server side session a simple place to store all unnecessary/secure data and also can grow exponentially. I saw plenty projects when in one moment session data was grown over 1Mb. So the limits of JWT is a huge plus

4. JWT is easily scalable solution. It works great with 1k session and with 1kk - 100kk - 1000kk sessions.

5. Migration. Because of limitation of server side session - plenty projects started with DB as session store just because they already have DB. It's way of pain of migration from DB to external storage memcache/redis then to JWT.

From my point I don't see any reason to keep using server session instead of JWT.

6. Microservices. It's not unusual to have several Microservices consume a common entity, like say, User. If every microservice has to authorize each request doing a db lookup, it doesn't scale either.

He does bring fair points that it's important to understand before using JWT. If your use case includes immediate logout, or to avoid stale data, then it might better to not use JWT.

WHAT! It's $750 bard based on upgraded raspberry pi ? For the whole controllers I don't want to pay more than $200& Otherwise it's better to purchase a mac mini and use in as rack mount.

You're talking apples and oranges. The LX2160A is an "edge computing part", with a bunch of RAM and I/O bandwidth, and enough cores to do something with it. Its not a workstation or a jellybean controller part.