Jim M

As of today, there are not any plans to make PHP 8.2 or higher compatible with Invision Community version 4. However, like many things in the software world, this is written in pencil so could change as we progress through the end of version 4's lifetime.

49 minutes ago, Claudia999 said:

...but you have to change robots.txt too – isn't it? There is this entry:

Disallow: /tags/

That is correct as well, if you're still using the default robots.txt

You would need to install them by uploading the files to your server and then going to ACP -> System -> Applications to run the installer. The guide below is regarding the previous license terms where you could purchase individual ones but the process is still the same to install them.

While I get what you're saying here, it is not intended to how we built the system. It is intended to only count answers, rather than voters. However, it still does provide you the voters next to the answers if you want to do your own calculations.

If you would like to see this changed, you're more than welcome to submitting a topic in our Feedback forum for further information.

7 minutes ago, tforums said:

IPB already caches a lot of things

We no longer cache full pages/responses. This was removed some time ago in effort to move customers more towards better solutions that wouldn't utilize so much server resources, e.g. CDN caching. We pass a cache header but it won't cache unless you have a CDN (or similar) servicing your community.

Example of a HEAD request to our all activity stream. As you can see, there are caching headers which will tell our CDN to now cache this page for 900 seconds (15 min).

I've moved this to a ticket so we can look further into this to see what we can do for you here. Please watch your email for further correspondence.

This is added on purpose for the search pages. You can likely change this using the metatag editor though.

You can just comment the lines out or copy/paste them somewhere else. The lines being:

\define( 'REDIS_ENABLED', true );
\define( 'STORE_METHOD', 'Redis' );
\define( 'STORE_CONFIG', '[]' );
\define( 'CACHE_METHOD', 'Redis' );
\define( 'REDIS_CONFIG', '{"server":"127.0.0.1","port":6379,"password":""}' );
\define( 'SUITE_UNIQUE_KEY', '########' );

Would recommend disabling Redis for testing purposes as it looks like you have template cache enabled. Would then ensure that all your folders are correctly writable to ensure everything is able to write/delete/regenerate.

As you’re missing files, that would be the first thing you need to resolve. Is the queue task running and taking a while to complete or are you getting an error?

35 minutes ago, rayzir said:

I came back to check on things and the background processes are really starting to back up.  Right now, I have 26 in the queue. 

I checked logs in the Locked Taks: Queue and there's nothing in there.

When attempting to run that task, I am still getting a 503 service unavailable error. You will need to contact your hosting provider to find out what is stopping that service prematurely. It may be a resource issue, configuration issue, etc...

Do you have an example of the request? This could be the form of an attack, if there are many requests at the same time, which would need to be mitigated at the server or network level for optimal results as the software would just cause more consumption of resources.

45 minutes ago, Svetozar Angelov said:

Can you specifically say what exactly you have seen on our forum that makes you think the problem is not with your software.

At a basic level, we are not seeing any direct logs, errors, brute force attempts, etc... The user in question logged straight in which can only be done with having the credentials. The email on the most recent user which you banned also has had passwords exposed from different other, non-IPS websites. Which is why we're asking for further examples of why you believe it to be an issue with the software. It isn't pointing that way but of course, if you have further information, we're happy to explore it.

22 minutes ago, Randy Calvert said:

they not target important accounts?  Why not target admin or moderators?

Also, what Randy is stating here, they would not just attack normal members hold no importance significance  to the community who can be deleted/banned/etc.... They would want to do more damage, gain more exposure, etc...

12 minutes ago, Svetozar Angelov said:

What we suspect, along with a lot of people, is that the IPS are telling us generalities without really understanding the issue, which is a breach with the March security update, and they're probably buying time to figure out exactly what's going on.

As mentioned, please provide an example and we'd be happy to investigate further your particular case. However, what we have seen on your website is not a breach on your community or in our software but we're happy, of course, to confirm that with a specific example user.

Please be advised I have split this off for you to the Feedback forum.

7 hours ago, Svetozar Angelov said:

We have taken enough measures, I ask that you now take measures and look very carefully at the code from the beginning of March, because we have not had such problems before. It is clear that precisely from this period the problems with spammers on the IPS platform started massively. I can't sit all day and clean the forum of spammers after the version is paid for and obviously the problem is yours.

As Randy, us, and others mentioned several times through these conversations, spam is a walk of life on the internet and odds are you will never be able to 100% combat it. It also comes in cycles. Points that are high, turn to lows as people adapt spam measures and combat spam accounts or our spam defense grows aware of individuals, etc... However, our methods posted here will make it significantly less annoying and to a degree remove it from being a daily hassle.

As Randy mentioned, we take significant measures with each release to ensure that we release the best software possible to our customers. We have mentioned several times in this topic that we're happy to look at what you're seeing but we have not gotten a specific example. We have, however, looked at a recent example of a banned member and they have compromised email/password combinations from other websites around the web (non-IPS) so that would explain how a malicious-intended spammer could gain access directly, with no brute force. If this is not an active member, banning them, as you've done, is the right measure. If it is an active member, you may wish to change their password and contact them. However, it could be that the spammer/attacker has access to their email inbox so be mindful of that.

If you would like us to take another look at another example or have a further complex example than just a username, please feel free to us the Contact Us form at the bottom of each page and we'll be happy to take a look.

4 minutes ago, CheersnGears said:

Suddenly it's 2018.

This isn't directed at you Jim, but at the decision-makers at IPS as a whole.

The fact that a vendor of web-based photo gallery software doesn't officially support WebP in 2024 is, frankly, absurd.

And it's absurd for several reasons:

1. It clearly works. I wasn't even aware that webp wasn't officially supported (or I didn't remember) because I've been uploading webp to my galleries for a while now, and it has been working nicely, and will continue to work nicely in their new S3 home.
2. The standard was announced in 2010 and went to stable release in 2018. There's been 8 years to get onboard.
3. Much of the web has been using it for years, especially the big, picture heavy sites.
4. The search engines penalize us for using heavy images and I personally see a significant performance difference on my own site with webp.
5. The only thing that broke here was that I moved storage locations, and the url update task isn't processing the new location properly. The images are still attached properly to the article, they just aren't embedded correctly.

 

 

While I understand what you're stating here, I would advise putting in a suggestion in the Feedback forum. As this is the support forum, there is very little from a standpoint of support we can do to assist as currently, it is not supported.

11 minutes ago, Svetozar Angelov said:

Despite all the measures taken

Unfortunately, I do not see that you have seen all measures taken.

13 minutes ago, Svetozar Angelov said:

I'm sure you have a bug in IPS that occurred after an update from the beginning of March. Our problems continue. 😣

Unfortunately, without an example, we cannot review that. However, I looked at the user who you just recently banned in your administrator log, and they have indeed been a part of a data breach of non-IPS sites. You can use https://haveibeenpwned.com/ to check their email and see if their password(s) have been exposed from other website breaches. 

10 minutes ago, CheersnGears said:

Huh? I'm not using any third party here for this.

1. I upload them as webp to a gallery.
2. I use the "existing attachment" method to insert the images into my article.
3. Profit.

This has "just worked" like this for a while. The only thing that has changed here is where those images are stored on the back end.

I agree that this only seems to happen with webp images, however, the articles were fine until the S3 move and no third party was involved.

webp is not a supported extension of the Gallery by default.

These images are .webp extensions. Our software, by default, does not support these as embedded images in content. They appear as downloaded files. If you're using a third-party service/application to convert these, you would need to work with them to assist you here.

Keep in mind that we only support Amazon S3. Using another service's S3 may not work completely as intended as their API may be just slightly different and cause different results/issues. 

If you use any CDN or other server caching services, you will also want to clear these after doing a move of this magnitude.

The current page is rendering a 500 Internal Server Error so I cannot access it currently.

You can make a test installation of your community for you and your staff to test by appending -TESTINSTALL to the end of your license key. You can read more about it here: 

3 hours ago, DawPi said:

To be honest, it's quite easy to reproduce. See this:

Someone or something is attempting to "hack" or something on the board. In the code, you may try to find it at applications\calendar\modules\front\calendar\view.php:

	/* Pagination */
		$offset = isset( \IPS\Request::i()->offset ) ? min( array( \IPS\Request::i()->offset, \count( $events ) ) ) : 0;

As you can see, there is no check to ensure that the offset value is an integer. Please try changing it to:

		/* Pagination */
		$offset = isset( \IPS\Request::i()->offset ) ? \intval( \IPS\Request::i()->offset ) : 0;
        $offset = $offset ? min( array( $offset, \count( $events ) ) ) : 0;

Should help.

Thank you. I have reported what you mentioned here to our developers. However, I am unable to reproduce the exact error which has been originally reported. Therefore, we cannot guarantee this will resolve the OPs issue without access to their instance.

5 minutes ago, Jipa331 said:

Thanks for the suggestion. it would help to solve this issue.
Where can I find this option in IPS ACP? (logout all users at once and request all of them to reset their PW)

 

ACP > Members > Force password reset

Keep in mind that the biggest hole in any authentication/identity system is the human using it. Odds are that if that user setup several accounts around the internet with the same credentials, their email is more than likely also to be one of those. Your solution may solve the issue in some cases but odds are likely not in its favor. As the attacker, likely has access to their email as well.

Which is why using a non-email source, like a Two Factor Authentication code generation with a cell phone app, is generally more secure. As an attacker obtaining access to that 2FA source is harder.

The best case, would have been requiring it from the start of any community. That’s not always possible but the good news, you can require 2FA starting today and any new members or members who login will have it implemented.

You can also use the logout all members and change password requirements to ensure that users need to reset their password prior to logging in again. In conjunction with requirements around password difficulty, this will help hopefully change passwords for your users.

However, if you feel strongly about the code generating link to an email to login, you’re more than welcome to suggest that in our Feedback forum for further evaluation.