I did not know about autolinks - thanks for the link!

It is interesting how different parsers handle this exact situation. I usually am cautious about it because I typically am not sure how it will be handled if I am not explicit with the URL and additional text.

I’m curious about this. The source text of your comment appears that your comment was just the URL with no markdown. For your comment about a markdown parsing bug to be true, shouldn’t the URL have been written in markdown with []() notation (or a space between the URL and the period) since a period is a valid URL character? For example, instead of typing https://google.github.io/styleguide/cppguide.html., should [https://google.github.io/styleguide/cppguide.html.](https://google.github.io/styleguide/cppguide.html) have been typed?

Yes, I am using PersistentVolumes. I have played around with different tools that have backup/snapshot abilities, but I haven’t seen a way to integrate that functionality with a CD tool. I’m sure if I spent enough time working through things, I may be able to put together something that allows the CD tool to take a snapshot. However, I think that having it handle rollbacks would be a bit too much for me to handle without assistance.

Thanks for the reply! I am currently looking to do this for a Kubernetes cluster running various services to more reliably (and frequently) perform upgrades with automated rollbacks when necessary. At some point in the future, it may include services I am developing, but at the moment that is not the intended use case.

I am not currently familiar enough with the CI/CD pipeline (currently Renovatebot and ArgoCD) to reliably accomplish automated rollbacks, but I believe I can get everything working with the exception of rolling back a data backup (especially for upgrades that contain backwards incompatible database changes). In terms of storage, I am open to using various selfhosted services/platforms even if it means drastically changing the setup (eg - moving from TrueNAS to Longhorn, moving from Ceph to Proxmox, etc.) if it means I can accomplish this without a noticeable performance degradation to any of the services.

I understand that it can be challenging (or maybe impossible) to reliably generate backups while the services are running. I also understand that the best way to do this for databases would be to stop the service and perform a database dump. However, I’m not too concerned with losing <10 seconds of data (or however long the backup jobs take) if the backups can be performed in a way that does not result in corrupted data. Realistically, the most common use cases for the rollbacks would be invalid Kubernetes resources/application configuration as a result of the upgrade or the removal/change of a feature that I depend on.

There are several proprietary options (many/most of which you cannot host). Looking for Amazon Wishlist alternatives should help in putting together a list of potential options. Some additional projects which are open source and selfhostable that you could also start with include:

• Christmas-Community
• Gifthunter
• Listidge
• Listslist
• PHP Gift Registry
• Pønskelisten
• Wishlist App

Everything I mentioned works for LAN services as long as you have a domain name. You shouldn’t even need to point the domain name to any IP addresses to get it working. As long as you use a domain registrar that respects your privacy appropriately, you should be able to set things up with a good amount of privacy.

Yes, you can do wildcard certificates through Let’s Encrypt. If you use one of the reverse proxies I mentioned, the reverse proxy will create the wildcard certificates and maintain them for you. However, you will likely need to use a DNS challenge. Doing so isn’t necessarily difficult. You will likely need to generate an API key or something similar at the domain registrar or DNS service you’re using. The process will likely vary depending on what DNS service/company you are using.

Congrats on getting everything working - it looks great!

One piece of (unprovoked, potentially unwanted) advice is to setup SSL. I know you’re running your services behind Wireguard so there isn’t too much of a security concern running your services on HTTP. However, as the number of your services or users (family, friends, etc.) increases, you’re more likely to run into issues with services not running on HTTPS.

The creation and renewal of SSL certificates can be done for free (assuming you have a domain name already) and automatically with certain reverse proxy services like NGINXProxyManager or Traefik, which can both be run in Docker. If you set everything up with a wildcard certificate via DNS challenge, you can still keep the services you run hidden from people scanning DNS records on your domain (ie people won’t know that an SSL certificate was issued for immich.your.domain). How you set up the DNS challenge will vary by the DNS provider and reverse proxy service, but the only additional thing that you will likely need to set up a wildcard challenge, regardless of which services you use, is an email address (again, assuming you have a domain name).

Raspberry Pi + PiHole + PiVPN = Network Gateway Drug

Although, PiVPN is winding down so you might want to find something different instead. Setting up a regular Wireguard VPN isn’t so bad, but it may be simpler to setup a Tailscale Tailnet.

Calls made from speakers and Smart Displays will not show up with a caller ID unless you’re using Duo.

Is it possible to use Duo still? Google knows it discontinued/merged Duo with Google Meet nearly 18 months ago, right?

https://changedetection.io/

Change Detection can be used for several use cases. One of them is monitoring price changes.

Some additional ideas for the Protectli device:

• backup/redundant OPNsense instance for high availability
• backup NAS/storage
  • set it up at a family/friend’s house
• a test/QA device for new services or architecture changes
• travel router/firewall
• home theater PC
• Proxmox/virtualization host
  • Kubernetes cluster
• Tor, I2P, cryptocurrency, etc. node
• Home Assistant
  • dedicated local STT/TTS/conversation agent
• NVR
• low powered desktop PC

There are so many options. It really depends on what you want, your other devices, the Protectli’s specs, your budget, etc.

Could you explain further? Wouldn’t this just need to be setup once per server that OP wants to connect?

Could you use symlinks? Not sure what the “gotchas” or downside to this approach is though.

tl;dr: A notable marketshare of multiple browser components and browsers must exist in order to properly ensure/maintain truly open web standards.

It is important that Firefox and its components like Gecko and Spidermonkey to exist as well as maintain a notable marketshare. Likewise, it is important for WebKit and its components to exist and maintain a notable marketshare. The same is true for any other browser/rendering/JavaScript engines.

While it is great that we have so many non-Google Chrome alternatives like Chromium, Edge, Vivaldi, etc., they all use the same or very similar engines. This means that they all display and interact with websites nearly identically.

When Google decides certain implementation/interpretation of web standards, formats, behavior, etc. should be included in Google Chrome (and consequently all Chromium based browsers), then the majority marketshare of web browsers will behave that way. If the Chrome/Chromium based browsers reaches a nearly unanimous browser marketshare, then Google can either ignore any/all open web standards, force their will in deciding/implementing new open web standards, or even become the defacto open web standard.

When any one entity has that much control over the open web standards, then the web standards are no longer truly “open” and in this case becomes “Google’s web standards”. In some (or maybe even many) cases, this may be fine. However, we saw with Internet Explorer in the past this is not something that the market should allow. We are seeing evidence that we shouldn’t allow Google to have this much influence with things like the adoption of JPEG XL or implementation of FLoC.

With three or more browser engines, rendering engines, and browsers with notable marketshares, web developers are forced to develop in adherence to the accepted open web standards. With enough marketshare spread across those engines/browsers, the various engines/browsers are incentivized to maintain compatibility with open web standards. As long as the open web standards are designed and maintained without overt influence by a single or few entities and the open standards are actively used, then the best interest of the collective of all internet users is best served.

Otherwise, the best interest of a few entities (in this case Google) is best served.

Would something like this work for the web or Teams versions of Excel?

Regardless, I agree. The license and remote only execution are horrible.

That’s true. It wouldn’t solve the dependency issue though (eg - I’m using Pandas v1.5.3, you’re using Pandas v2.0.3) and I’m not sure how well it would work for some platforms like the web or Teams.

The code storage for Python is no different than regular Excel functions (eg - VLOOKUP(), SUM(), etc.), meaning that it is stored within an Excel cell. The only differences are that Python code is run remotely vs Excel functions running locally and the location of Python’s code matters vs Excel’s functions are location agnostic (ie - Python code runs in cells located left-to-right, top-to-bottom but Excel’s functions can dynamically determine the calculation order/location).

I’m not sure that this new Python integration changes much about this use case (except for another way to accomplish the same/similar tasks).

I agree with all your points about Excel being capable. However, I’m struggling to think of examples where this newly announced Python integration within Excel would be helpful (with the exception of new/different visualizations) - especially for the reasons you stated about modern Excel.

Are there any use cases that you can think of where someone who knows Excel well would resort to “adding a little Python to patch up any issues”?

This integration won’t allow you to do that. Python will not run locally, but instead on Microsoft’s platform (likely Azure).

If you’re just reading some simple data from Excel, there are several ways of accomplishing this already. For example, Pandas has read_excel() and there is also openpyxl. You could even use those tools to write the results back to Excel. Things get more complicated though if the Excel file is something more than just a simple list.

I agree! I’m not sure why you’re being downvoted either.

This new integration just allows you to do data analysis and data visualization of existing data within an Excel file via Python. The output of your Python scripts is limited to the Excel file. The Python environment itself is also limited as it runs on Microsoft’s platform and is controlled by Microsoft.

The (Excel) problem that people already using Python for data analysis/visualization is that they have to use Excel files. Reading/writing Excel files via Python can sometimes be tedious or limiting. Utilizing Python inside of Excel via this integration may help in some scenarios, but they won’t be able to use custom libraries built internally, control the Python environment (eg - must use specific version of Python or Python library, can’t utilize all Python libraries available on pip, etc.), connect to all necessary external data sources via Python, and utilize proper VCS tools like git.

The problem that people automating tasks via Python have is that there is no Python library nearly as capable of reading or manipulating Excel files as VBA is. This new Python integration does not change that.

The problem that Excel users have is that they want more advanced (or simple/easier) data analysis and data visualization capabilities. However, with Excel’s dynamic array formulas, LAMBDA formula, Power Query, and Power Pivot, Excel is becoming much more capable than it ever was. If those tools cannot meet your needs, you likely need to move to something like R, Python, or some other tool. Embedding Python into Excel like this integration does still limits Python with all of Excel’s current restraints (size, performance, etc.).