I'm being facetious of course, but this recent rhetorical trend of people confidently vouching for "pet" in "pet vs. cattle" is not a sustainable decision, even if it's admittedly plain practical on the short to medium run, or in given contexts even longer. It's just a dangerous and irresponsible lesson to blindly repeat I think.
Change happens. Evidently, while we can mechanistically rule out several classes of bugs now, RCEs are not one of those. Whatever additional guardrails they had in place, they failed to catch this *. I think it's significantly more honest to place the blame there if anywhere. If they can introduce an RCE to Notepad *, you can be confident they're introducing RCEs left and right to other components too **. With some additional contextual weighting of course.
* Small note on this specific CVE though: to the extent I looked into it [0], I'm not sure I find it reasonable to classify it as an RCE. It was a UX hiccup, the software was working as intended, the intention was just... maybe not quite wise enough.
** Under the interpretation that this was an RCE, which I question.
[0] https://www.zerodayinitiative.com/blog/2026/2/19/cve-2026-20...
Most people seem to see "CVE" and "RCE" and assume the worst here. As you saw though, Notepad is just making totally valid URIs clickable! Web browsers allow it too - why is it not an RCE there? Sure, they usually show a warning when the URI is going to something external but most people just click through things like that anyway.
Web browsers warn you about opening arbitrary protocols. And you have to select the program that will open it.
This Notepad vuln, allows you to click things like ssh://x....
> According to the CVSS metric, the attack vector is local (AV:L). Why does the CVE title indicate that this is a remote code execution?
> The word Remote in the title refers to the location of the attacker. This type of exploit is sometimes referred to as Arbitrary Code Execution (ACE). The attack itself is carried out locally.
> For example, when the score indicates that the Attack Vector is Local and User Interaction is Required, this could describe an exploit in which an attacker, through social engineering, convinces a victim to download and open a specially crafted file from a website which leads to a local attack on their computer.
The low level tool that has served to rescue more systems than I can count does not need to "change" simply because "it happens, bro."
> while we can mechanistically
You can rule it out with process as well. As in "don't change what isn't broken."
> If they can introduce an RCE to Notepad
Then they clearly feel they have no viable competition. This is table stakes. Getting it wrong should lose you most of your customer base overnight. Companies actually used to _work_ this way.
What you find a trusty "low-level" tool is a demo application for a basic WYSIWYG text editor. They modernized it so that it remains being perceived that way, instead of letting it be increasingly misclassified as a legacy product for the enthusiast, like you just did.
For example, Bob's one-of-a-kind trusty server from which Bob is nigh inseparable, vs. a Docker container with a version controlled config you routinely tear down and bring up instances of, maybe even in an automated fashion.
Here this would map to trusty aged codebases you don't touch out of fear and caution, vs. codebases you can confidently touch because the spec, the code, the tests, the tooling, and the processes are solid.
maybe we should separate "real origianl text-only editor" from "fancy text editor"?
windows already got wordpad... why even lay a finger on textpad?
Windows 10 explorer.exe is 100x faster than Windows 11 explorer, it's not even close.
It also signals the death knell for Windows native apps. Microsoft can't make them anymore. It won't be long until even Excel is a Electron sloplication.
I have a hard time believing this. I'm pretty sensitive to performance losses and I haven't noticed any difference between those. It wouldn't make sense either, given they should both host the same shell icon views. Are you sure the difference you're seeing is in explorer.exe? As opposed to something else, like a new shell extension or a new filesystem filter driver on Windows 11?
M$ has now introduced web-latency into the desktop along with their adoption of web-tech into the OS. You gotta get used to staring at that spinning blue circle, counting the many precious moments of your life draining away.
Ultimately, what difference does it make? The file explorer in Windows 10 is much faster than the one in Windows 11, and it's very noticeable. Turn on the old context menus, and try right clicking a file. Instant in Windows 10, visible delay in Windows 11.
It does offer some new features for businesses. Nothing useful for the consumer, and nothing to justify the massive performance loss
The new calculator even manages to screw up basic input. The old calculator accepted both commas and periods as decimal separator inputs. It just worked no matter what I typed in. The new calculator has some sort of "clever" localization where my inputs change depending on the language of the operating system. My language uses commas so of course it only accepts those. Infuriating. Hope whoever coded this is enjoying their promotion.
The slowdown appears to be due to XAML Islands, which allow legacy code to use modern MS UI stuff.
https://www.techindeep.com/why-is-windows-explorer-slow-7289...
20231109 https://news.ycombinator.com/item?id=38212453 Windows 11 Update 23H2 is stealing users' IMAP credentials (666 points, 278 comments)
> the new Outlook is a thin wrapper around the cloud version, so the IMAP sync happens in the cloud, not locally
Literally the best parts of Windows have been the parts they forgot existed for 10+ years and never changed.
Somehow in this timeline AI can only be used to make things worse and sloppier
