TLDR: When working with LLMs, the risks for the L&D workflow and its impact on substantive learning are real:Hallucination — LLMs invent plausible-sounding facts that aren’t trueDrift — LLM outputs wander from your brief without clear constraintsGeneric-ness — LLMs surface that which is most common, leading to homogenisation and standardisation of “mediocre”Mixed pedagogical quality — LLMs do not produce outputs which are guaranteed to follow evidence-based practiceMis-calibrated trust — LLMs invite us to read guesswork as dependable, factual knowledge These aren’t edge cases or occasional glitches—they’re inherent to how AI / all LLMs function. Prediction machines can’t verify truth. Pattern-matching can’t guarantee validity. Statistical likelihood doesn’t equal quality.

AI’s instructional design “expertise” is essentially a statistical blend of everything ever written about learning—expert and amateur, evidence-based and anecdotal, current and outdated. Without a structured approach, you’re gambling on which patterns the model draws from, with no guarantee of pedagogical validity or factual accuracy.

There is no way to link PRs; you have to manually note changes in commit messagesIf changes are made to the parent, you have to rebase all of the child branches

however, due to the arbitrary data structure design, LPGs are not as practical for modelling ontologies and other structured data representations as RDF models.

current system is ‘closed source’, and is carried out by competitive agents that do not share innovations for very long time periods; the competitiveness of these agents requires behaviors that externalize costs

So fork is fast, unsafe, and maybe bloated.

This behaviour may affect one-click unsubscribe links. If clicking the link in an email automatically processes the unsubscribe, then Barracuda may unsubscribe users without their knowledge.

This behaviour may affect opt-in confirmation links.

IMAP's main downside is that you have to trust your email provider to not lose your email.

t may be that in using his system hedeveloped his mind and his knowledge of history to the point wherehe expected his readers to draw more inferences from the facts heselected than most modern readers are accustomed to doing, in thisday of the predigested book.

So you can’t apply a background and know it will cover that whole grid area anymore.

cannot visualize/explain well attributes that are not spatiallylocalized, like size, color, etc. In addition, they can showwhich areas of the image may be changed in order to affectthe classification, but not how they should be changed

con: doesn't let you define after_transaction callbacks anywhere like ar_after_transaction does (outside of the after_commit, etc. callbacks which only happen at certain points in the model's life cycle)

React, on the other hand, often requires a fair amount of boilerplate code, even for simple interactions.

This kind of error handling does express error state (either via HTTP 500 or by the top-level "errors" key), but it doesn’t take advantage of GraphQL’s type system and can only express one error at a time.

That’s not the only way of writing end-to-end tests in Rails. For example, you can use Cypress JS framework and IDE. The only reason stopping me from trying this approach is the lack of multiple sessions support, which is required for testing real-time applications (i.e., those with AnyCable 😉).

network requests are a big deal, and having to deal with this kind of thing is one of the prices of switching away from server-side rendering to a distributed system

Global mocks introduce magic and technically can reduce test coverage.

It might be a pain because every time you modify function's arguments you'll need to recreate/drop types aswell tho.

JavaScript, as a language, has some fundamental shortcomings — I think the majority of us agree on that much. But everyone has a different opinion on what precisely the shortcomings are.

STATSD_SAMPLE_RATE: (default: 1.0)

ENV['STATSD_SAMPLE_RATE'] = 1
config/initializers/statsd.rb:8:in `[]=': no implicit conversion of Integer into String (TypeError)

The Result object that we pass around keeps accumulating data and becomes enormous, so we have to use **rest in our function signatures

The DSL has a weaker control over the program’s flow — we can’t have conditions unless we add a special step

Fans of e-readers and e-books are a point down in the print v digital reading debate. E-books make a poor background in Zoom meetings.

The downside is that if you introduce multiple layers of scrolling (one for the browser, one for the iframe) your users will get frustrated. Like adzm said, you don't want to use an iframe for primary navigation

However, one of the drawbacks of this property is that the line intersects descenders of the characters.

The downside is the installation files are bigger than the traditional Debian package manager (DEB) files. They also use more hard drive real estate. With snaps, every application that needs a particular resource installs its own copy. This isn’t the most efficient use of hard drive space. Although hard drives are getting bigger and cheaper, traditionalists still balk at the extravagance of each application running in its own mini-container. Launching applications is slower, too.

I had one issue with snap and that involved VLC but I can see how it would lead to issues with other packages. I hav the libdvdcss2 package installed to allow me to watch DVDs on my laptop. The snap version of VLC was not aware of that and wouldn’t play the DVD. I had to uninstall the snap and install the .deb package. Just one example, but I know there will be others. Due to the quasi-legal nature of libdvdcss2, I doubt it’ll ever be bundled in a VLC snap package.

Moreover, due to the confinement, snap does not allow Chromium to download by default in another folder than /home : “it won’t let the application see files on the host system (save for a few exceptions, like $HOME)”. Is it a definitive point for snap ? If yes, it means that when all apps will be converted to snap without possible backport to debs (if installed without --classic and perhaps excepted Nautilus), it will be impossible to save files issued from them elsewhere than in /home ? Absolutely all my datas (documents, music, videos, photos) are on other partitions, this would be prohibitive…

If folks want to get together and create a snap-free remix, you are welcome to do so. Ubuntu thrives on such contribution and leadership by community members. Do be aware that you will be retreading territory that Ubuntu developers trod in 2010-14, and that you will encounter some of the same issues that led them to embrace snap-based solutions. Perhaps your solutions will be different. .debs are not perfect, snaps are not perfect. Each have advantages and disadvantages. Ubuntu tries to use the strengths of both.

Sadly, unlike Apt, it isn’t possible to uninstall all of these packages in bulk. You must manually remove each one of these packages by hand.

SVG has the advantage that integrates very well with Svelte, since it’s an XML and the nodes can be managed as if they were HTML. On the other hand, Canvas is more efficient, but it has to be generated entirely with JavaScript.

The main point — each framework/library/instrument has some kind of flaw, something it handles a little worse than others. And Svelte is no exception.

-0.2 Kingmaking

Kingmaking may occur

Combining with minimizing is not possible as minimizers usually only emit a single line.

This prevents you from debugging execution on statement level and from settings breakpoints on columns of a line.

The @import rule has a number of serious issues:

Would style .classInChild from your parent. The only drawback is that you might need an extra wrapping element.

Also, Rollup, which I use in the article as a bundler is pretty slow. Why? Because it needs to re-compile the whole shebang every time you change a file. It produces very small and efficient bundles though.

What i dislike in Svelte 💔

But with Disqus, these advantages are not valid anymore, because your blog text and its comments are hosted separately on different servers.

most com-monly, systems such as PGP are used, which use long-livedencryption keys (subject to compromise) for confidentiality

The biggest advantage of matrix/megolm is that you can always add new client logins, that aren't tied to your mobile phone login (as opposed to whatsapp/signal).

It’s an alternative to using YAML anchors and is a little more flexible and readable

Note that this will pull the container if it exists. Pulling even the small alpine is quite an overhead just to get that information. Not mentioning the several GB-Image one often has in CI-Environments.

using SSH is likely the best approach because personal access tokens have account level access

JAVA based

Thank you for letting me know about this move by Google. Definitely something to watch. While I agree with Google's position from an end user experience perspective, it unfortunately puts Firefox at a further disadvantage since Mozilla does not have its own language translation initiatives.

Advantages and disadvantages

But where is the helper method defined? What’s its visibility? Can I put it in a module? Can I use inheritance? Who can call it? Can I call super from the extracted method? If so, where does super go?

The downside of this is that the Agile Keychain had to keep the titles and the web locations unencrypted. In the OPVault format, we found a way to work with all data encrypted.

One mistake that we made when creating the import/export experience for Blogger was relying on one HTTP transaction for an import or an export. HTTP connections become fragile when the size of the data that you're transferring becomes large. Any interruption in that connection voids the action and can lead to incomplete exports or missing data upon import. These are extremely frustrating scenarios for users and, unfortunately, much more prevalent for power users with lots of blog data.

This situation usually arises from external constraints not design choices such as my example with Sequel. My point is that assigning a value to a constant is allowed by Ruby in certain scopes and not others. It used to be up to the developer to choose wisely when to perform the assignment. Ruby changed on this. Not for everyone's good.

this will only work from the second page view, when consent is already present on the page

you have less direct control as you must rely on the vendor’s adherence to IAB’s guidelines for compliance.

stored in a first-party cookie, which means that users will be requested to give their consent again after deleting browser cookies or switching to another device or browser

To be honest, there’s no good reason to use ga.js anymore. Google is keeping it around for legacy, but hasn’t maintained it since 2014

Probably the greatest disadvantage is the issue of security.  Because  NFS is based on RPC, remote procedure calls, it is inherently insecure  and should only be used on a trusted network behind a firewall.

Also, if a host is having network problems, it may stop logging, making it difficult for the administrator to determine what chain of events led up to the current state. At the very least, the administrator will not be able to access these logs from that host.

Another problem with simple login-based authentication is that there is no way to control how much access an individual third-party application gets: it's an all-or-nothing deal based on whether you are willing to give the program your password.

When third-party software runs amok with your login information for a Web application, the only way to stop it in some cases is to change your password

Using Todo.txt for Android requires a Dropbox account in order to sync your file across devices.

Pro: easy to set up Con: allows brute-force password guessing Con: requires password entry every time

The only real trouble with rsync (underlying), is that it can still take considerable time with large file systems and remote systems over slow links. You might want to think about snapshots as well (not just the rsnapshot ones, but file system ones).

However, in this case you would lose the possibility to render something in between. You are strictly coupled to the higher-order component's render method. If you need to add something in between of the currency components, you would have to do it in the higher-order component. It would be quite similar as you have done it previously by rendering the currency components straight away in the Amount component. If using a render prop component instead, you would be flexible in your composition.