Using #git is not the whole picture on #modernhw version control landscape. Git is great when one decides to locally follow changes, take diffs, create branches and so on. When it comes to collaboration with other people or to create a community around a common project, the need for extra tooling arises, and it becomes evident that git alone is not enough. A #gitforge fills this gap.
Git bare repositories are a means of sharing the local git history remotely. Bares doesn’t show the worktree, as they are used solely as a common exchange place. This might be a remote server accessible through ssh, for example. Several different users may collaborate this way, provided they agree on a common workflow. Bares are more than enough for some needs. A front end on top of it may help to get an overview of what is going on and to take a look at branches, users and the like. All it takes to make this workflow useful is a little management, as git was designed with a fully distributed architecture in mind. Check the docs for more details.
Now, this approach is a bit too bare bones for most people. On top of bare git repositories, some decided to add extra functionality to ease using git remotely, calling for contributors attracted by buttons, colors, menus and most generally, being used to web frontends. Web forges include all usual suspects (project creation and configutation, markup rendering, user account and authorizations, project overview, etc.), as well as more advanced features (continuous integration, #ci, for testing and deployment with git hooks, wikis, code linters, built in actions, issue tracking, etc.). They abstract the use of git showing diffs, logs, issues threads, etc. As any other web gui tool, they come with its own set of inconvenients in what concern user freedom.
Popular examples are all around. #Gitlab may be deployed as a custom (not federated) instance, and is commonly found in research and public institutions; codeberg, based on forgejo, is a great example of how to deploy a lightweight #freesoftware instance of a collaborative forge (and the promise to federate on the fediverse). Many others exist, which more or less features, bells and whistles. You always have the choice.

sourcehut

#Sourcehut, as a collaborative platform, deserves special attention. It departs from mainstream forges, following a different paradigm based on the most robust, distributed and flexible technology at our hands since decades, plain text #email. Git, since its origins, includes a close integration with email, as they both share a distributed philosophy, avoiding central point of failure silos (surprising how mosft git forges tend to concentrate in silos). Sourcehut core architecture is based on mail exchange, patches and #maillists, which turns out to be a much more flexible approach than that of what most forges propose. Their concept of project goes well beyond that of usual workflows, integrating nicely git with email, wikis, bug trackers and build features. They’re still in an alpha stage, so expect the best still to come.

I finally decided to move my #NextCloud instance from one that I had been operating on the #Vultr hosting service to my #HomeLab.

A note on Vultr: I am impressed with this service. I have used them for multiple projects and paid with various means, from credit card to #cryptocurrency for about 10 years and I cannot even remember a downtime that impacted me. (In fact, I think there was only one real downtime, which was planned, well-communicated, and didn’t impact me because my setup was fairly resilient). With a growing volume of data, and sufficient spare hardware that wasn’t doing anything, I decided to bring it in-house.

This is not going to be a full guide, as there are plenty of those, but I did run into some hurdles that may be common, especially if a pre-built Nextcloud instance was used. So this is meant to provide some color and augment the official and popular documentation.

Getting started

Plan out the migration

Migration Overview

Essentially, there are three high-level steps to this process 1. Build a new Nextcloud server in the homelab 2. Copy the configuration (1 file), database (1 backup file), apps (install apps), and data (all user files) over to the new system 3. Restore all the copied data to the new instance

Preparing to Migrate

1. Start with the NextCloud official documentation for migrating to a different server as well as:
   1. Backing up Nextcloud
   2. and the restoring a server doc
2. Check out Nicholas Henkey’s migrate Nextcloud to a new server blog post. This is very thorough and has some great detail if you’re not super familiar with Nextcloud (because you used a pre-built instance)
3. For the new build:
   1. A full set of installation instructions, placing [Nextcloud behind an Nginx proxy](https://github.com/jameskimmel/Nextcloud_Ubuntu/blob/main/nextcloud_behind_NGINX_proxy.md.
   2. An older install document for Installing Nextcloud on Ubuntu with Redis, APCu, SSL & Apache

Migration

While the official documentation describes the basics, the following is the steps I recommend following. This is at a medium level, providing the details, but not the specific command-line arguments (mostly).

1. Build the new server
   1. Use your favorite flavor of Linux (I used Debian, and these notes will reflect that)
      1. install all updates,
      2. install fail2ban or similar security if you’re exposing this to the Internet.
      3. name the new system the same as the outgoing server
   2. Download the Nextcloud install from the nextcloud download site and choose either:
      1. update the current system to the latest version of whatever major version your running, and then download latest-XX.tar.bz2 where ‘XX’ is your version
      2. identify your exact version and download it from nextcloud
   3. Install the dependencies (mariaDB, redis, php, apache, etc. etc.)
      1. note: if the source server is running nginx, I recommend sticking with that for simplicity, keeping in mind that only Apache is officially supported
   4. Unpack Nextcloud
   5. Validate that it’s working
   6. Place it into maintenance mode

2. Backup the data

   1. If using multi-factor authentication, find your recovery codes or create new ones
   2. Place the server into maintenance mode
   3. Backup the database
   4. copy the database backup to a temporary location on the new server

3. Restore the data

   1. Restore the database
   2. copy /path/to/nextcloud/config/config.php over the existing config.php
   3. rsync the data/ directory to the new server
      
      1. you can remove old logs in the data directory
      2. you may need to use an intermediary step, like a USB drive. It’s best if this is ext4 formatted so you can retain attributes
      3. the rsync options should include -Aaxr you may want -v and/or --progress to get a better feel for what’s going on
      4. if rsync-ing over ssh, the switch is -e ssh
   4. If you have installed any additional apps for your Nextcloud environment, rsync the apps/ directory in the same way as the data dir above
   5. Validate the permissions in your nextcloud, data, and apps directories. Fix as necessary, see the info Nicholas Henkey’s post (linked above) for commands
   6. Redirect your A or CNAME record to the new system
   7. Configure SSL on the new system
   8. Turn off maintenance mode
   9. Log in and test! :fingers-crossed:

Troubleshooting

Hopefully everything is working. Make sure to check the logs if something is broken.

log locations – the nextcloud.log in the data/ directory – the apache logs in /var/log/apache2 – the redis logs in /var/log/redis – the system logs, accessible with journalctl

Reiterating: Remember or check for these items

These are the specific notes I took as I ran into problems that I had to work around or solve. These are incorporated in the above, so this is basically a restatement of the gotchas I ran into:

• upgrade the current one to the latest version of the current release (i.e. the latest of the major version you are on, so if you were running 29.0.3, get to 29.0.9)
  • this makes it easier when you download <version>-latest.tar.bz2
  • If you’d prefer to skip that, use the nextcloud download site with all available versions. Make sure to grab the same one and compare the specific version as listed in config.php. Example: 'version' => '29.0.9.2',
• use the same name on the new server
• use the same web server. Apache is officially supported, but if you’re using nginx, it will be easier to stay on that.
• Most multi-factor authentication, like WebAuthN, FIDO hardware keys, etc. will not work over HTTP in the clear.
  • IOW: make sure you have recovery codes
• If the apps aren’t copied over, the new server sees them as installed rather than installable. I suppose one could “delete” or remove them in the admin GUI and then reinstall, but otherwise, there was no button to force a reinstall.
• Files and data you need to copy over after creating the install. Do each of these separately, rather
  • if you have any additional apps, copy the apps/ directory over
  • copy config.php
  • copy the data/ directory
    
• Is your current install using Redis-based transactional file locking?
  
  • If the previous system was using Redis and it is still in the configuration, the new system will not be able to obtain file-locking and essentially all users will be read-only and not able to modify or create new files.
  • In config.php, you will see settings such as 'redis' and 'memcache.locking' => '\\OC\\Memcache\\Redis',
  • make sure Redis is installed on the new system and running on the same port (or change the port in config.php)
  • Install the necessary software: apt install redis-server php-redis php-apcu
  • Ensure that the Redis and APCu settings in config.php are according to the documented single-server settings

The Memcache settings should look something like the following configuration snippet. Alternatively, you could enable and use the process socket.

'memcache.local' => '\OC\Memcache\APCu',
'memcache.distributed' => '\OC\Memcache\Redis',
'memcache.locking' => '\OC\Memcache\Redis',
'redis' => [
     'host' => 'localhost',
     'port' => 6379,
],

Nextcloud administration notes

These instructions and administrative notes were written for the pre-built Nextcloud provided by hosting provider Vultr. As a way to de- #Google my life and take back a bit of #privacy, I have been using a Vultr-hosted instance for a couple years now and it has run quite well. These notes are really aimed at the small instance for personal use. Please don’t use my notes if you’re responsible for an enterprise server!

Upgrading Nextcloud

#Nextcloud, with all it's PHP-based functionality, can become temperamental if not upgraded appropriately.  These are my notes to remind me how to now completely break things. When upgrading, the first pass will usually bring you to the most up-to-date version of Nextcloud in your major release, e.g. an instance running 27.1.4 would be brought up to 27.1.11. Running the script again would bring the instance to 28.0.x.

To update a Nextcloud server running on the #Vultr service to the latest version, you need to follow the steps below:

1. Backup your Nextcloud data: Before starting any update process, it's always a good idea to create a backup of your Nextcloud data. This will ensure that you can restore your data in case of any unexpected issues during the update process.
   1. Shutdown the OS with shutdown -h now
   2. Power down the instance in Vultr
   3. Create a snapshot
   4. Wait
   5. Wait some more – depending on how much data is hosted on the system
   6. Power it back up
2. SSH into the Vultr server: To update the Nextcloud server, you need to access the server using SSH. You can use an SSH client such as PuTTY to connect to the Vultr server.
3. Switch to the Nextcloud user: Once you are logged in, switch to the Nextcloud user using the following command: sudo su -s /bin/bash www-data.
4. Navigate to the Nextcloud directory: Navigate to the Nextcloud directory using the following command: cd/var/www/html  (could be /var/www/nextcloud or other.  Check what's in use)
5. Stop the Nextcloud service: To avoid any conflicts during the update process, stop the Nextcloud service using the following command (as www-data): php occ maintenance:mode --on 
6. Update the Nextcloud server: To update the Nextcloud server, you need to run the following command(as www-data): php updater/updater.phar. This will start the update process and download the latest version of Nextcloud.
7. Update the OS, as needed, with apt upgrade
8. Start the Nextcloud service: Once the update is complete and verified, you can start the Nextcloud service using the following command: sudo -u www-data php occ maintenance:mode --off.
9. Verify the update: After the update process is complete, you can verify the update by accessing the Nextcloud login page. You should see the latest version of Nextcloud listed on the login page.
10. Assuming all is running smoothly, the snapshot that was created in step 1 can be safely deleted. Otherwise, they accrue charges on the order of pennies / gigabyte / day.

Some other notes

Remove files in the trash

When a user deletes files, it can take a long time from them to actually disappear from the server.

root@cloud:/var/www/html# sudo -u www-data php -f /var/www/html/cron.php root@cloud:/var/www/html# sudo -u www-data php occ config:app:delete files_trashbin background_job_expire_trash

Set files to expire

root@cloud:/var/www/html# sudo -u www-data php occ config:app:set —value=yes iles_trashbin background_job_expire_trash

I need to figure out how to reasonably deal mail and deliveries privately.

How it started

I donated to a local nonprofit in 2024, and I really shouldn't say this, but I honestly wish I never did. However, this is not due to a reason you probably expect.

I started to receive significantly more junk mail from charitable nonprofits and groups, more so than usual (at least since the 2020 COVID-19 pandemic). I won't name specific names, but this was a local nonprofit which has a total annual budget size between the order of $1 million and $10 million.

(To the reader: if we know each other IRL, then I'll tell you who the offending org is; and if your savvy with implementing an actionable fix with the issue below, then maybe we can work out a way for me to get out of this rut of a “situation” — as if this is or should be by highest priority project to take on right now. Let's just say that some of you will be surprised by the org I have in mind, which either intentionally uses the services of data brokers, or at least has some heuristic workflow that is leaking donor info to data brokers. The overall situation has a bit of a tragic irony.)

I'm (usually) not a vengeful person, at least when it comes to nonprofit orgs genuinely acting in good faith; but I am keeping a running list of these others orgs that engage in buying/selling/sharing snail mail lists as orgs I won't donate money to in the future, due to their respective disregard for mail privacy. However, there are 3 national-level orgs that have (so far) never sold out to physical mail lists: the ACLU, including state chapters; the EFF; and the Freedom of the Press Foundation. I am purposefully excluding comparatively technical groups that would respect the privacy and security of others in general, such as the Signal Foundation and The Tor Project.

On the other hand, the only other way to avoid excessive physical mail list tracking is to donate to small local nonprofits. (Any method is fine — if you're super concerned about protecting your membership info, using a PO box for your mailing address and renewing your member dues via paper check is more than sufficient for most local community members.) This is because these groups literally don't have the money to spend for mass mail solicitations or blanket marketing.

After this happened, I expressed to a local activist about how I'm going to go straight for a paid plan on Privacy.com (at least the lower tier) and skip the free plan. Additionally, I commented that I reaction was essentially the “I can't believe you've done this” meme. (Somehow, I was initially confused this with the “Charlie bit my finger” meme.)

How it's going (and the future)

I no longer think it's safe for me to order computers and ship the delivery to my residential address, using my own debit card. (That does remind me – I really should get a credit card for better payment protection and everything else that encompasses.)

I remembered that I ordered the HP Dev One in 2022 and the box's outer shipping box wasn't even taped closed when it arrived on my doorstep. Due to my living situation since 2020, I no longer trust anything that goes through the mail, and after Andrew “bunnie” Huang's assessment of overall supply chain security after the 2024 exploding pager incident in Lebanon, I think it's about high time I figure out the logistics of shipping to a private mail box (PMB) – or maybe I use a friend's address and/or credit card to purchase an online only computer (while I pay my friend for the cost, of course).

However, quite a few large computer manufacturers, who primarily have B2B (business-to-business) though also some minor B2C (business-to-consumer) sales, will tell customers that sending deliveries to a PO Box is not allowed during checkout. This includes Lenovo, HP, and even Framework. (I have to double check for System76.) This is partly why I was sad when Costco no longer sold any in-store ThinkPad laptops anymore (one probable cause might be the pandemic, but that's another matter).

If you have any somewhat serious considerations to become a Linux distro maintainer or even a package manager (such as the AUR/MPR), you should at least consider this while threat modeling. I recall Ariadne Conill tweeting about how a Lenovo ThinkPad laptop that they tried ordering online was suspiciously redirected to Langely, Virginia while en route to their home in early 2022, which was symptomatic of mail interdiction. However, those tweets were deleted around late 2022 or early 2023.

There is always something new to try... https://soapbox.pub/servers/

BUT I am now a concentrate and focus. Too much candy? Too many ideas and possibilities? It all depends on the priorities we need. In other words what is your hat colour? Black, white, grey or red? No hats for me, not even green or hoody.

Security for me is transparency or zero preference. Otherwise I am spending all my time on noise and “AI” generated attempts to fathom my rousing browsing. I am already using too many browsers, except TOR. Which is one rocky peek too many.

Slow too. Too slow. Like my keyboard. Old and clunky. Noisy and dusty. Good enough...

There was a guide from early 2023 on what to change in the default KDF settings of Bitwarden.

(The guide has been saved on the Wayback Machine and archive.today.)

You must log in via browser to edit these settings. (Neither the desktop apps nor the mobile apps can change the following settings.)

1. From the main screen in Bitwarden, navigate through the following menus: Security (vertical menu) > Keys (horizontal)
2. Select Argon2id for “KDF algorithm” and enter 10 for “KDF iterations”.
3. Enter 64 for “KDF memory (MB)” and 8 for “KDF parallelism” (number of threads).
4. If you changed any settings, then click on the “Change KDF” button to save any changes (and Bitwarden will log you out of your account on all devices).
   • Otherwise, if no changes were made, then you can leave the “Keys” menu.

Personal context

I need to make sure I have something I can reference when I set up organization accounts on Bitwarden for colleagues and friends.

I vaguely remember that this was discussed roughly around the same about how the default KDF for LUKS (full disk encryption on Linux) was set up. Back in April-May 2023, the sources for episode 132 of the the Surveillance Report podcast was released during the time when the podcast released roughly biweekly – so the podcast lagged at least 1-2 weeks behind current events.

This forum thread helped to date this news story, as well as this assessment.

Remember KISS? Keep It Simple Stewpit,

We do not have to spread ourselves thinly. We can rely on the wheel being invented. We can focus on less but better and complete and cooperate and merge efforts. That is why I trust my experience and others who are offering real services I need. Real alternatives. Really simple. Really.

This evening it's Old Granddad 114. I picked it up at Costco for under $30. I've heard good things about it, so I thought it was time to try it.

The proof makes it hot, so I'm having it with some ice. On the nose, I'm getting maple, chipotle, and maybe some anise. I don't notice anything right away on the tongue, but the maple shows up with some cinnamon. The heat, along with the flavors, lingers on the tongue. There's briefly a hint of anise later. The heat sticks around and follows the swallow and hangs around.

I'm kind of reminded of whisky's that have been finished in amburana casks, but the maple isn't quite as strong.

We'll see how the second dram this evening goes...

I’ve been a “dabbler” with crewAI for a while now, having come across it in March of 2024 and tinkering when I have some time. I love the idea of task-based LLM actions that are specific enough that you can use lower cost but specifically-trained models for the tasks, even running those models on your own hardware. A few weeks back, my team at work used CrewAI for a hackathon in an effort to automate an onerous process, and it came out pretty well!

So, when I saw that they have a couple official training videos on a new e-learning platform called DeepLearning.ai, I figured I’d check them out. #CrewAI is evolving rapidly, and the some of the notes I’ve taken over the past 8 months aren’t even applicable anymore, so I figured this was a great way to level-set and fill in gaps in my knowledge.

I’m not going to describe CrewAI here, other than it’s a framework for easily building multi-agent teams and requires little to no coding experience. How CrewAI works is well-explained in the first fifteen minutes of the course, so at least listen to that part!

About the Course

The course, Multi AI Agent Systems with crewAI, follows a flow familiar to anyone that has taken online courses, and is taught by the creator of crewAI, João Moura. The lessons, ranging from a minute to 18 minutes, are a mix of descriptive lecture and hands-on coding. For the lessons where coding is involved, the window handily split-screens and on one side is an iPython notebook environment with the code pre-populated, so you can work through it as João explains what the code does.

You can also get an indication of the course and CrewAI by checking out my crewAI experimentation repo on GitHub.

Target Audience Analysis

Professionals working in project management, artificial intelligence, and team leadership can greatly enhance their skills in constructing multi-agent teams. Those keen on optimizing team performance, utilizing state-of-the-art technologies for collaborative work, and streamlining task execution processes would discover value in enrolling in specialized online classes tailored to augment their proficiency in this realm. Addressing prevalent challenges like steering diverse teams, accomplishing project goals in intricate scenarios, and keeping pace with evolving team dynamics is indispensable for professionals aiming to excel in their respective roles.

Content Outline

Introduction to Multi-Agent Teams

Familiarizing with the basics of multi-agent teams and their significance in managing complex tasks effectively.

Importance of Building Multi-Agent Teams for Complex Tasks

Unveiling the reasons why seamless collaboration among agents is crucial for successful task fulfillment.

Strategies for Creating Effective Multi-Agent Teams

Scrutinizing established methodologies for assembling and overseeing high-performing multi-agent teams.

Multi-Agent Crew Performance Examples

The bulk of the course is working through five clear examples of building multi-agent systems with CrewAI. The result is a set of real-world instances where multi-agent teams can perform and achieve remarkable results across diverse problems.

write_article

The first example is super simple: take a task you might do with a single prompt with an #LLM chatbot, such as chatGPT, and have it performed by multple agents, each with their own persona. This performs no research and the output is purely from the LLM, making it an easy entrypoint for anyone. (Hint: I had it write an article about taking an online course for developing multi-agent teams, and even incorporated a bit of it into this.)     – this one required an update from the training to run on the latest version of crewai.     – older versions of crewai used an integer for the verbose declaration, and if you are running the latest, you need to change that to a binary, e.g. True.

This example is definitely one you’ll want to revisit after you learn how to use and assign tools to your agents in the following lessons.

customer_support

This creates a customer support agent that can answer questions about a specific product or service. It does this by accessing the URL with the support documentation.

customer_outreach

This example creates some marketing lead material to reach out to a specific company that is a potential customer or partner for a second, i.e. “your” company.

event_planning

This example uses multiple agents to research, identify, create the logistics, and then create some marketing material for an event. It takes in parameters like city, size, and budget in order to find a viable venue. – I believe it was this one where I had to fiddle with the asynchronicity of the agents, since I understand that CrewAI needs to have the last agent to perform a task be performing that itself. I could have that wrong, but I had to change that to make mine work.

This is actually a super-cool example, but I found that the LLMs did not adhere to the parameters, often getting venues too small or ignoring the input I would provide while they were performing their tasks. That’s to be expected, however, and I think experimentation is the name of the game when it comes to building these systems.

resume_builder

The final one was to have the agents create bespoke resumes, based upon the job one is applying for. As opposed to the event planning exercise, the output on this one was very good, and I was impressed with how well it could craft a resume for the specific job, as well as anticipate some of the interview questions and provide some hints for how to answer them.

Conclusion

This course provides a clear and thorough introduction to crewAI, bringing the attendees to an intermediate level of being able to use the framework. By immersing themselves in the intricacies of multi-agent team dynamics, professionals can acquire the requisite knowledge and proficiency to thrive in today's collaborative work settings. Embracing online classes tailored to address the subtleties of forming effective multi-agent teams represents a proactive stride towards honing essential skills and keeping abreast in the ever-dynamic professional sphere.

• basic intro instructions for whisper transcription application – link
• basic intro instructions for using yt-dlp to download media – link

Começando... Espero entender logo tudo!

Please stand by for unimportant messages.