Hi,
I just opened voting for the “Remove the links to X.com from PHP.net” RFC.
The RFC: https://wiki.php.net/rfc/remove-link-to-x-from-php-net
The PR: https://github.com/php/web-php/pull/1879
The discussion thread:https://news-web.php.net/php.internals/130610
Voting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thanks.
Jim
Hi
On 5/18/26 20:07, Jim Winstead wrote:
I just opened voting for the "Remove the links to X.com from PHP.net" RFC.
The RFC: PHP: rfc:remove-link-to-x-from-php-net
The PR: Remove X/Twitter link from homepage social media section by castillo-n · Pull Request #1879 · php/web-php · GitHub
The discussion thread:php.internals: [RFC] Remove the links to X.com from PHP.netVoting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thank you for your RFC. I voted against it, because I believe that the RFC is poorly written and executed.
In particular I believe it is asking the wrong question with the "RFC Impact" section also drawing a conclusion that I don't believe follows from the question asked: The presence or absence of a link on php.net does not change whether or not the @official_php account on X is an official presence or not.
Given that php.net linked to this account, I consider it to be an established fact that the account is an official account of the PHP project - which the RFC also acknowledges.
The question to then decide on is whether or not the account should be abandoned, which is something that entails more than removing the links from php.net (e.g. updating the bio). The removal of the link directly follows from the account status, not the other way around.
This is also something that Rowan noted during the discussion php.internals: Re: Re: [RFC] Remove the links to X.com from PHP.net. As far as I can tell this email did not receive a response from you as the RFC author, in fact the entire subthread starting from Rowan's first email doesn't contain any emails from you. That's the “poorly executed” part I mentioned above: I feel you did not properly engage with (what I believe to be) well-articulated concerns regarding the RFC by either replying with counter-arguments or by making changes to the RFC in response.
Best regards
Tim Düsterhus
PS: I'm also just noting a factual error in the RFC text itself: The newest post was made on 2024-06-04, whereas the RFC claims in the introduction that the account has not posted since 2023.
On Mon, 18 May 2026 at 14:10, Jim Winstead <jimw@trainedmonkey.com> wrote:
Hi,
I just opened voting for the “Remove the links to X.com from PHP.net” RFC.
The RFC: https://wiki.php.net/rfc/remove-link-to-x-from-php-net
The PR: https://github.com/php/web-php/pull/1879
The discussion thread:https://news-web.php.net/php.internals/130610
Voting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thanks.
Jim
Two data points:
Python Software Foundation is still on X:
Rust stopped posting to X a year ago and removed links from its website
On Tue, May 19, 2026 at 4:34 AM Matthew Brown <matthewmatthew@gmail.com> wrote:
Two data points:
Python Software Foundation is still on X:
- 2x more followers than in 2021
- 5x fewer interactions on the average post than in 2021Rust stopped posting to X a year ago and removed links from its website
- People still talk about Rust
And they argue against the RFC, not for it.
PSF still on X with 2x more followers -- that's exactly the case for
staying. X-wide engagement has changed: different algorithm, different
audience, different content. That's not "the platform is worse",
that's "the platform is different". Lower engagement per post doesn't
mean lower value if followers and reach are still there.
Rust stopped posting to X a year ago and removed links from its website
Rust has momentum that has nothing to do with social media strategy.
PHP doesn't have that luxury. We're trying to repair perception, not
manage hype.
And these aren't isolated cases. TypeScript, Go, the Apache
Foundation, the Linux Foundation, and many others actively communicate
on X. The pattern across language and OSS communities is presence, not
retreat.
Another data point: we have actual analytics on what platforms drive
engagement to php.net. For April 2026:
| Social Network | Visits | Share |
|----------------|-------:|-------
| YouTube | 5,373 | 33.66% |
| Reddit | 2,915 | 18.26% |
| X/Twitter | 2,102 | 13.17% |
| StackOverflow | 2,068 | 12.96% |
| Facebook | 2,048 | 12.83% |
| LinkedIn | 675 | 4.23% |
| Telegram | 205 | 1.28% |
| Hacker News | 166 | 1.04% |
| Vkontakte | 154 | 0.96% |
| Sourceforge | 98 | 0.61% |
| Instagram | 58 | 0.36% |
| Mastodon | 37 | 0.23% |
| Workplace | 16 | 0.10% |
| V2EX | 15 | 0.09% |
| Bluesky | 14 | 0.09% |
| Threads | 12 | 0.08% |
Among text-based platforms suitable for project communications, X
drives more traffic than LinkedIn and Mastodon combined. An active
account would expand reach.
That's the kind of analysis the policy RFC encourages producing going
forward: PHP: rfc:social-media-policy
-Roman
On Tue, May 19, 2026 at 3:28 PM Roman Pronskiy <roman@pronskiy.com> wrote:
And they argue against the RFC, not for it.
PSF still on X with 2x more followers -- that's exactly the case for
staying. X-wide engagement has changed: different algorithm, different
audience, different content. That's not "the platform is worse",
that's "the platform is different". Lower engagement per post doesn't
mean lower value if followers and reach are still there.
This reads very LLM generated, so it's hard to take this seriously.
And these aren't isolated cases. TypeScript, Go, the Apache
Foundation, the Linux Foundation, and many others actively communicate
on X. The pattern across language and OSS communities is presence, not
retreat.
As someone in the Linux ecosystem I can list big community run projects
that did leave X. Like Debian, KDE, GNOME.
I think PHP as a community run project should follow these steps,
not corporate run projects or projects heavily backed by corporations.
--
Cheers
Ilya
On Tue, May 19, 2026 at 1:12 PM Ilya Orlov <ilyaorlov124@gmail.com> wrote:
On Tue, May 19, 2026 at 3:28 PM Roman Pronskiy <roman@pronskiy.com> wrote:
> And they argue against the RFC, not for it.
>
> PSF still on X with 2x more followers -- that's exactly the case for
> staying. X-wide engagement has changed: different algorithm, different
> audience, different content. That's not "the platform is worse",
> that's "the platform is different". Lower engagement per post doesn't
> mean lower value if followers and reach are still there.This reads very LLM generated, so it's hard to take this seriously.
> And these aren't isolated cases. TypeScript, Go, the Apache
> Foundation, the Linux Foundation, and many others actively communicate
> on X. The pattern across language and OSS communities is presence, not
> retreat.As someone in the Linux ecosystem I can list big community run projects
that did leave X. Like Debian, KDE, GNOME.
I think PHP as a community run project should follow these steps,
not corporate run projects or projects heavily backed by corporations.
I use AI for polishing my arguments obviously. The source of the
strings doesn't change whether the data points are accurate.
Fair point about GNOME and Debian. But I don't see why PHP should
follow. "Community vs corporate" would matter if maintaining presence
costed us something. Automated posting via GitHub actions costs
effectively nothing. GNOME and Debian don't have an unfavorable "you
are dead" narrative to push back against. We do.
Exactly the trade-off the policy RFC makes explicit. The current vote skips it.
Just to be clear, Linux Foundation is almost entirely backed by corporations; a long list of them appears right on the homepage.
I think Roman correctly stated that PHP currently lacks the same news appeal as more visible languages like Rust & Python. Therefore, neglecting a viable way to communicate with the user base seems counter-intuitive to say the least.
Ilia Alshanetsky
Technologist, CTO, Entrepreneur
E: ilia@ilia.ws
T: @iliaa
B: http://ilia.ws
On Tue, May 19, 2026 at 4:55 PM Ilia <ilia@ilia.ws> wrote:
Just to be clear, Linux Foundation is almost entirely backed by corporations; a long list of them appears right on the homepage.
Yes, I consider Linux Foundation to be the "projects heavily backed by
corporations", not much community left there.
I think Roman correctly stated that PHP currently lacks the same news appeal as more visible languages like Rust & Python.
Therefore, neglecting a viable way to communicate with the user base seems counter-intuitive to say the least.
With this I disagree, sometimes it's better to do the right thing
rather than the optimal one
--
Cheers
Ilya
Fair argument, “right thing” is a matter of perspective and opinion of course.
What I’d encourage you to consider is “right thing” for the project trying to maintain relevance & hopefully grow, vs “right thing” of a social statement that may or may not represent the opinions of all people involved in said project.
Ilia Alshanetsky
Technologist, CTO, Entrepreneur
E: ilia@ilia.ws
T: @iliaa
B: http://ilia.ws
On Mon, 18 May 2026, Tim Düsterhus wrote:
In particular I believe it is asking the wrong question with the "RFC Impact"
section also drawing a conclusion that I don't believe follows from the
question asked: The presence or absence of a link on php.net does not change
whether or not the @official_php account on X is an official presence or not.
I honestly don't even understand why this is even a contentious point.
We haven't posted anything for years on that platform. Hence, it makes
no sense to link to it, as it's not of any benefit. If that changes,
it's a one liner to put the link back.
I therefore obviously voted in favour of this RFC.
cheers,
Derick
Hey Jim,
Am 18.05.2026 um 20:07 schrieb Jim Winstead jimw@trainedmonkey.com:
Hi,
I just opened voting for the “Remove the links to X.com from PHP.net” RFC.
The RFC: https://wiki.php.net/rfc/remove-link-to-x-from-php-net
The PR: https://github.com/php/web-php/pull/1879
The discussion thread:https://news-web.php.net/php.internals/130610
Voting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thanks.
Jim
I would support a RFC to simply remove it conditionally based on the activity there. (And automatically re-add upon re-activation.)
That would be a no-brainer to me.
The secondary motivation (a very political one) is nothing I want the PHP project be associated with.
A RFC which seems to try to permanently remove the PHP presence on these grounds too, mainly due to this reason gets a no from me.
Like Roman says, participation on social networks should be about actual and effective reach, not perception.
I’m looking forward to Romans RFC which aims to force re-activation of the @official_php handle on X.
Bob
On May 19, 2026, at 6:25 am, Roman Pronskiy <roman@pronskiy.com> wrote:
Another data point: we have actual analytics on what platforms drive
engagement to php.net. For April 2026:| Social Network | Visits | Share |
|----------------|-------:|-------
| YouTube | 5,373 | 33.66% |
| Reddit | 2,915 | 18.26% |
| X/Twitter | 2,102 | 13.17% |
| StackOverflow | 2,068 | 12.96% |
| Facebook | 2,048 | 12.83% |
| LinkedIn | 675 | 4.23% |
| Telegram | 205 | 1.28% |
| Hacker News | 166 | 1.04% |
| Vkontakte | 154 | 0.96% |
| Sourceforge | 98 | 0.61% |
| Instagram | 58 | 0.36% |
| Mastodon | 37 | 0.23% |
| Workplace | 16 | 0.10% |
| V2EX | 15 | 0.09% |
| Bluesky | 14 | 0.09% |
| Threads | 12 | 0.08% |Among text-based platforms suitable for project communications, X
drives more traffic than LinkedIn and Mastodon combined. An active
account would expand reach.That's the kind of analysis the policy RFC encourages producing going
forward: PHP: rfc:social-media-policy-Roman
Thanks for the stats!
An average of 70 visits a day from X doesn’t seem like a ton of engagement though? I’m reasonably sure that figures from 2021 would be much much larger.
I believe that the era of town-square text social networks is over. There is no longer a single place where I can follow friends, colleagues, experts and funny people all in one place. Some of the exodus was a reaction to the Twitter acquisition, but a lot of it was just a general exhaustion from the dynamics that emerge from such systems.
X still has influence in Silicon Valley due to the large number of tech execs, AI companies and researchers, who continue to post there — but I don’t think there’s much overlap in audience.
Matt
Thank you Jim for the best written no-brainer RFC for years!
On Tue, 19 May 2026, 15:55 Matthew Brown, <matthewmatthew@gmail.com> wrote:
On May 19, 2026, at 6:25 am, Roman Pronskiy <roman@pronskiy.com> wrote:
Another data point: we have actual analytics on what platforms drive
engagement to php.net. For April 2026:
Social Network Visits Share YouTube 5,373 33.66% 2,915 18.26% X/Twitter 2,102 13.17% StackOverflow 2,068 12.96% 2,048 12.83% 675 4.23% Telegram 205 1.28% Hacker News 166 1.04% Vkontakte 154 0.96% Sourceforge 98 0.61% 58 0.36% Mastodon 37 0.23% Workplace 16 0.10% V2EX 15 0.09% Bluesky 14 0.09% Threads 12 0.08% Among text-based platforms suitable for project communications, X
drives more traffic than LinkedIn and Mastodon combined. An active
account would expand reach.That’s the kind of analysis the policy RFC encourages producing going
forward: https://wiki.php.net/rfc/social-media-policy-Roman
Thanks for the stats!
An average of 70 visits a day from X doesn’t seem like a ton of engagement though? I’m reasonably sure that figures from 2021 would be much much larger.
I believe that the era of town-square text social networks is over. There is no longer a single place where I can follow friends, colleagues, experts and funny people all in one place. Some of the exodus was a reaction to the Twitter acquisition, but a lot of it was just a general exhaustion from the dynamics that emerge from such systems.
X still has influence in Silicon Valley due to the large number of tech execs, AI companies and researchers, who continue to post there — but I don’t think there’s much overlap in audience.
Matt
Is this serious?
Hans Z
On Mon, May 18, 2026, at 11:07 AM, Jim Winstead wrote:
Hi,
I just opened voting for the “Remove the links to X.com from PHP.net” RFC.
The RFC: https://wiki.php.net/rfc/remove-link-to-x-from-php-net
The PR: https://github.com/php/web-php/pull/1879
The discussion thread:https://news-web.php.net/php.internals/130610Voting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thanks.
Jim
Hi,
The vote failed to achieve a 2/3 majority of “Yes” votes, so the link to defunct account on the website owned by the transphobic white supremacist will remain on the homepage of PHP.net.
Thanks to everyone who voted.
Jim
On Thu, 4 Jun 2026 at 05:55, Jim Winstead <jimw@trainedmonkey.com> wrote:
On Mon, May 18, 2026, at 11:07 AM, Jim Winstead wrote:
Hi,
I just opened voting for the "Remove the links to X.com from PHP.net" RFC.
The RFC: PHP: rfc:remove-link-to-x-from-php-net
The PR: Remove X/Twitter link from homepage social media section by castillo-n · Pull Request #1879 · php/web-php · GitHub
The discussion thread:php.internals: [RFC] Remove the links to X.com from PHP.netVoting runs until 2026-06-03 17:00:00 UTC. There is just one vote to cast.
Thanks.
Jim
Hi,
The vote failed to achieve a 2/3 majority of "Yes" votes, so the link to defunct account on the website owned by the transphobic white supremacist will remain on the homepage of PHP.net.
Thanks to everyone who voted.
Jim
I have voted No not because I am against removing the link to x.com
from php.net, but because I am against the RFC. If we lost access to
the account, create a new one or if we have nobody to post these
announcements, remove the link from php.net. There is no need for an
RFC. From what it seems to me, x.com is still online, and some people
are using it, so it should not be about who owns that platform but
rather whether the PHP team has the need and capacity to post anything
there. I think social media exposure is good, but not strictly
necessary.
Together with the x.com link there are links to fosstodon.org and
linkedin.com. The latest post on X was about PHP 8.3, but the other
sites have more recent activity. So unless someone decides to start
posting on x.com again, the link should be considered dead and removed
from the site, regardless of the outcome of this RFC. In a similar
vein, we don't have links to Facebook or TikTok because, as I assume,
nobody is posting announcements there. Maybe there should be a social
media person who is in charge of an account on all of these social
media sites and will post regularly. But having a link to an account
with no activity is of no use to anyone.
I think everything you said here is pretty reasonable except for one awkward situation created by the RFC: if someone were to go ahead and remove x.com links because “the link should be considered dead and removed from the site, regardless of the outcome of this RFC” wouldn’t that be explicitly contradicting a decision made by the PHP RFC process? It sure seems like having done just that would have been better, but now that the RFC has failed can that really be considered a reasonable action?
Marco Deleu
On Thursday, Jun 04, 2026 at 10:25 AM, Deleu <deleugyn@gmail.com> wrote:
It sure seems like having done just that would have been better, but now that the RFC has failed can that really be considered a reasonable action?
I would argue it is still a reasonable action. The PR in question should have been as mundane as any SEO-related dead link cleanup action. Fact is, the account is dormant (whether that’s by neglect or sabotage doesn’t really matter after 2 years) and a dead link to a dormant account is prominently displayed on the website.
People chose the PR in question and this RFC to push their viewpoints regarding whether the PHP project should even have that account. And I dare say the voting on this RFC was swayed by that bias, especially given the language of the RFC itself.
The reality is the presence of the link on the PHP website and the presence of the PHP project on (any) social media platform is two separate discussions and should have been approached as such. Instead, we got the mess that has played out over the last few weeks.