Tons of sites are currently offline to hopefully prevent an attack until they can update PHP; just the things I've run into this morning: OBS, TicketTailor, tons of wikis...
I thought they were down earlier, but apparently they were just pre-announcing being down in about 30 minutes, until they finish patching their servers.
I've seen a number of notifications on X and Bluesky of people taking their sites offline, though. Supposedly Cloudflare is providing all customers a WAF rule that will mitigate the exploits they've been able to see so far (though they mention, this is not full protection—patching is still required).
I just moved my last PHP site, my personal blog, to a static site generator. It looks like my timing was lucky!
The only significant thing lost of comments, but they have been getting more spammy over the years. Its actually better now too in many ways and I fixed various issues during the move.
Which is why suggestions like OP’s are never useful or actually serious. I honestly consider them snobby. Like the person who tells you you really ought to check your engine oil, radiator fluid, and brakes everytime you get in a car. It would prevent some awful issues sure, but it’s not gonna happen for 99% of people.
Having static site generator still require an admin endpoint that end users can use to author, publish or otherwise manage the content. Pushing that to a local tool means now you need to somehow synchronize that content across all of the end users devices without git. Not to mention if there is no server side component that means the client needs to have full permission to publish content on the server. Until recently ssh was the only option. Maybe someone could vibe an iroh app that gives you a vpn'd server side app without that now.
Did you try to implement it seriously? Naturally you need an admin endpoint, it just does not have to be reachable for everybody from the internet, just via your VPN. Big difference.
The person you're replying to didn't mention a static site generator. You could simply use Vinyl(fka Varnish), and after you pulled your PHP down, your boring brochure site would still be up.
The first site I put up professionally was a handrolled static generator written in PHP for a state website. Users just edited text files, triggered a rebuild, and then there was a sync to prod if staging looked fine. They could even add new sections by creating new subdirectories, and Imagemagick would automatically conjure new menu buttons and the code would put them in the right place.
> Pushing that to a local tool means now you need to somehow synchronize that content across all of the end users devices without git.
This is not hard. Rsync? Rsyncd?
> Not to mention if there is no server side component
Why would there be no server-side component? What does that even mean? Just give them a script to trigger if you don't want them to have permission to do anything. Give them another script to revert if you don't want to get any emergency calls.
This is not advanced technology. Weird to talk about it like it's impossible.
> never useful or actually serious
So strange. That site was so simple to deal with that it stayed up for a decade after I left. Ancient technology.
Generate static pages from your content system. Even for wordpress there's some projects.
If you need adaptations, AI should be helpful.
You can sell this also to bosses: Delivery of static sites consumes drastically less resources.
Which kind of makes sense, I think a lot of Dutch websites have been taken offline. I think it’s due to recent regulations that have been introduced.
I think, given the (supposed) seriousness of this issue, it’s an adequate response. I suspect Cloudflare et al will have rules added to their managed WAFs within a few hours after release as well.
That said: it seems like the core issue is that they can’t release PHP publicly without revealing the exploit, wouldn’t it have made more sense to coordinate in the background with these organizations, and important distros, so that this could have been patched without revealing the exploit?
But they say it's a stopgap and still advise patching. It makes sense, since their tweet also says "are updating mitigations as new vectors show up" so it seems like there are many ways to exploit this, whatever it is.
Yes our university is doing the same right now! They where thinking of putting the whole network behind VPN access to protect the poor sysadmins in the Institutes that have not updated their website in years…
I know nothing about the Belgian Center for Cybersecurity, but the fact that they are running a platform that has been infamous for its security issues for 25 years, and been considered essentially obsolete for 15 years, immediately makes me not take them seriously.
This is like a Health Ministry recommending leech therapy to its employees.
That's reducing things too far. "What car manufacturer hasn't had car crashes?" All cars are equally safe?
There's a vast difference between different languages and CMSs and how likely it is for critical flaws to occur. PHP has a horrible track record and continues to surprise us negatively.
I fought hard about four years ago at a previous company to upgrade us from PHP 5 to 8 because I think 5 was going EOL or something like that (could be wrong about that now). Based on my experience in software, I have to imagine the vast majority of normal developers do not give a damn about security in being in security patch lifecycles on software.
> I have to imagine the vast majority of normal developers do not give a damn about security in being in security patch lifecycles on software.
I just don't think it's the developer's responsibility.
Depending on the size of the team of course, but the work that developer's do comes from someone else. Empower me to make decisions and I'll make them, but don't blame me for the choices other people make if you won't empower me.
Exactly. It's not the bricklayer's responsibility when mortar fails. It's the person who specified that mortar.
So I'd agree it isn't not a coder or developer's responsibility when they're given a bill of materials and told how to build, but it is an engineer's responsibility. Comp should recognize that as well.
I think I get where you're coming from. I've worked mostly on small teams (sub 30) so it's always been on us. The engineering leadership never prioritizes security unless it's something as big as the log4J stuff because it was on customers' minds.
That said, if they're software packages from something like Node or PyPI, I do think that it falls on the developer. If the point is that leadership needs to give devs the ability to act on that, that's very fair.
I've worked at a company that delivered on-premise software, and the runners of said software were... risk-averse, but in a strange way in that they didn't really want to upgrade software if it didn't need to. So the software was stuck on PHP 5.something, also because newer versions required a RHEL update which they also didn't like to do.
I don't know. Still, at a time like this I would expect a 99% static public website to be up, running and serving all comers. This is a complete mission fail, both in terms of using PHP and in terms of not having a fall-back plan in place.
Due respect, neither of those make sense. PHP is boring software, and this website routinely suggests (1) use it if it is fit for purpose and (2) don't rewrite for the sake of it. For most sites, the engineering effort to have a static backup of their dynamic website (say, generated by Wordpress or Drupal) just isn't possible. They just need to power down and patch.
[EDIT: I see I may have erroneously extended your criticism to sites other than NCSC, for which it does make sense. I would still defend using PHP.]
Because websites with news, continuously updated advice, job openings, etc aren't all that static? This type of site hugely benefits from having a easily understandable CMS that you can teach to more or less everyone in 2 days or less. I suffered through a few of those back in my time in the military, where an instructor would teach a class of 20-30 how to build new pages in the CMS
Now are there more powerful tools out there that result in more scalable websites? Sure. My personal sites are built around static site generators for that exact reason. But none of better tools are as easy to understand for non-programmers as wordpress or one of the other click-and-drag CMS tools.
Not all hacks are used to steal data. It's an authoritative website. If this hack is used to host a fake PHP fix stuffed with credential stealers then you have a problem.
Oh I absolutely agree they should take it offline until the fix is ready, I just mean that the NCSC itself is very much still operational even if their public website is temporarily offline. Like a computer with the monitor switched off.
PHP is going to publish the patch in 5 minutes, right? On their website which runs on... Perl Catalyst? Ruby on Rails? Django? What could possibly go wrong?
Because you need a certain amount of notice and in this case drama to be created so people know to update ASAP, before the exploit is out in the wild. As soon as these patches are out it becomes a lot easier for any actors to figure out exactly what they are patching, and how to exploit it.
Announcing when this patch is going to be released should leave a much shorter vulnerable window with the exploit out in the wild than releasing it immediately and hoping people update before they're a target.
Once the fix is released, people will reverse-engineer it and use the exploit on pages that haven't updated yet.
So they need to give time for infra people in all different time zones to wake up, learn about the announcement, and plan how to patch it.
The intended course of action is 1. take down website a few hours before the fix drops 2. apply the fix 3. only make website public again after it has been fixed
> Today, a remote code execution vulnerability was found in PHP, called php2shell.
[...] Since explain xkcd is affected by this exploit, I have decided to take the wiki down until tomorrow.
PHP is worse in this aspect with its long tail of legacy and staying close to C. It's absolutely possible that memory-unsafe Python interpreters could have something vaguely similar, but it's much less likely. Plus there are Python interpreters in memory-safe languages.
In any case, memory-unsafe codebases will continue to be a significant danger. Be it in interpreters or anywhere else. Just like they have been for the past 20+ years.
>Safely ahead in a digitally resilient Netherlands.
Things being offline seems like a daily occurrence recently. Is it a co-incidence? Is it AI? Are Russia/Iran doing an e-blitzkrieg?
If it's a combination of all three, surely nation states are going to unleash everything possible right now and do as much damage before the world adapts and becomes resilient?
If the latter, would be interesting to put an actual cost on all this downtime and whether it's actually making any impact, or is just noise/free training.
I've seen a number of notifications on X and Bluesky of people taking their sites offline, though. Supposedly Cloudflare is providing all customers a WAF rule that will mitigate the exploits they've been able to see so far (though they mention, this is not full protection—patching is still required).
The only significant thing lost of comments, but they have been getting more spammy over the years. Its actually better now too in many ways and I fixed various issues during the move.
https://sequoia.pub/comments
Having static site generator still require an admin endpoint that end users can use to author, publish or otherwise manage the content. Pushing that to a local tool means now you need to somehow synchronize that content across all of the end users devices without git. Not to mention if there is no server side component that means the client needs to have full permission to publish content on the server. Until recently ssh was the only option. Maybe someone could vibe an iroh app that gives you a vpn'd server side app without that now.
The first site I put up professionally was a handrolled static generator written in PHP for a state website. Users just edited text files, triggered a rebuild, and then there was a sync to prod if staging looked fine. They could even add new sections by creating new subdirectories, and Imagemagick would automatically conjure new menu buttons and the code would put them in the right place.
> Pushing that to a local tool means now you need to somehow synchronize that content across all of the end users devices without git.
This is not hard. Rsync? Rsyncd?
> Not to mention if there is no server side component
Why would there be no server-side component? What does that even mean? Just give them a script to trigger if you don't want them to have permission to do anything. Give them another script to revert if you don't want to get any emergency calls.
This is not advanced technology. Weird to talk about it like it's impossible.
> never useful or actually serious
So strange. That site was so simple to deal with that it stayed up for a decade after I left. Ancient technology.
Thanks for proving my point.
> This is not hard. Rsync? Rsyncd?
How do I do that on an iPad or an Android?
> Just give them a script to trigger if you don't want them to have permission to do anything.
uhhh, what permissions does that "script" have? or is that your /wp-admin equivilent?
> So strange.
Indeed.
Unfortunately my static HTML is pre-rendered by a PHP engine before being served.
These particular vulnerabilities affect PHP scripts accepting POST requests as far as I've been able to gather so far.
It's a static web site generator, powered by php :)
I think, given the (supposed) seriousness of this issue, it’s an adequate response. I suspect Cloudflare et al will have rules added to their managed WAFs within a few hours after release as well.
That said: it seems like the core issue is that they can’t release PHP publicly without revealing the exploit, wouldn’t it have made more sense to coordinate in the background with these organizations, and important distros, so that this could have been patched without revealing the exploit?
But they say it's a stopgap and still advise patching. It makes sense, since their tweet also says "are updating mitigations as new vectors show up" so it seems like there are many ways to exploit this, whatever it is.
Europa.eu and all related subdomains were down very early CEST morning. They run Drupal.
Dark day for Drupal and Wordpress, heh
This is like a Health Ministry recommending leech therapy to its employees.
Anyway.
CCB is overtly a marketing organization, not an IT security one.
IIRC they don't even host their own website, The Chancellery of the Prime Minister does it for them.
That's reducing things too far. "What car manufacturer hasn't had car crashes?" All cars are equally safe?
There's a vast difference between different languages and CMSs and how likely it is for critical flaws to occur. PHP has a horrible track record and continues to surprise us negatively.
I just don't think it's the developer's responsibility.
Depending on the size of the team of course, but the work that developer's do comes from someone else. Empower me to make decisions and I'll make them, but don't blame me for the choices other people make if you won't empower me.
So I'd agree it isn't not a coder or developer's responsibility when they're given a bill of materials and told how to build, but it is an engineer's responsibility. Comp should recognize that as well.
That said, if they're software packages from something like Node or PyPI, I do think that it falls on the developer. If the point is that leadership needs to give devs the ability to act on that, that's very fair.
When a doctor makes a mistake they get to practice medicine in Peru for the rest of their life.
Once it's out it's likely to be an easy "disable this damnable feature nobody needs" as usual.
There's no such thing as PHP 6. There's no such thing as PHP 9.2.
[EDIT: I see I may have erroneously extended your criticism to sites other than NCSC, for which it does make sense. I would still defend using PHP.]
Now are there more powerful tools out there that result in more scalable websites? Sure. My personal sites are built around static site generators for that exact reason. But none of better tools are as easy to understand for non-programmers as wordpress or one of the other click-and-drag CMS tools.
Announcing when this patch is going to be released should leave a much shorter vulnerable window with the exploit out in the wild than releasing it immediately and hoping people update before they're a target.
So they need to give time for infra people in all different time zones to wake up, learn about the announcement, and plan how to patch it.
The intended course of action is 1. take down website a few hours before the fix drops 2. apply the fix 3. only make website public again after it has been fixed
Scheduled release gives the option to turn things off on a set schedule while applying patches.
e: turns out several people answered this. Whoops.
https://github.com/php/web-php
https://yle.fi/a/74-20252600
> Today, a remote code execution vulnerability was found in PHP, called php2shell. [...] Since explain xkcd is affected by this exploit, I have decided to take the wiki down until tomorrow.
In any case, memory-unsafe codebases will continue to be a significant danger. Be it in interpreters or anywhere else. Just like they have been for the past 20+ years.
>Safely ahead in a digitally resilient Netherlands.
Things being offline seems like a daily occurrence recently. Is it a co-incidence? Is it AI? Are Russia/Iran doing an e-blitzkrieg?
If it's a combination of all three, surely nation states are going to unleash everything possible right now and do as much damage before the world adapts and becomes resilient?
If the latter, would be interesting to put an actual cost on all this downtime and whether it's actually making any impact, or is just noise/free training.