I'll be upgrading CentOS today.

[Josh @ Dreamland]

Expect the forum to enter maintenance mode. You will be unable to post for some window of time during this migration.

Do not be surprised when this happens. I will only activate maintenance mode during the database migration. This should last something like half an hour—The export will take fifteen minutes, but migrating the ENIGMA DNS to point to our new IP will take longer. I'm going to see if I can change the assigned IP over instead of updating DNS records. I will keep this thread up to date.

[Josh @ Dreamland]

As expected, that took fifty times longer than it reasonably should, and the fallout is pretty ugly. I spent eight hours today cleaning up after that migration.

Here's what I know is broken:

• The Wiki has some E_STRICT output at the top.
• The EDC is somehow including literal PHP files in the page output (Don't worry; this whole site's open-source; this is not a security problem).
• The email server is basically guaranteed not to be working.
• If the server reboots, this entire site will fall apart until I manually run some iptables calls and restart MariaDB. Neither of these should be necessary.

Sorry to leave the site in this state, but it's 2AM, and I have to work tomorrow. Please report any additional problems you see, here in this thread.

[hpg678]

Make sure you get enough rest and take your time to iron out the problems. Appreciate the time and time effort you've put in. Keep up the good work!!

 

[Josh @ Dreamland]

The plan for now is to update the Wiki software to the latest version and install a new SMF adapter (the current one's pretty wonky).

The only other problem we've found is that I've broken the LateralGM.org ErrorDocument handler.

[time-killer-games]

Thank you so much for all your hard work Josh. We all appreciate it!

[hpg678]

The plan for now is to update the Wiki software to the latest version and install a new SMF adapter (the current one's pretty wonky).

The only other problem we've found is that I've broken the LateralGM.org ErrorDocument handler.

Can one now upload their own examples and mini-games to the EDC now since you did updates?

[Josh @ Dreamland]

That will require a little more work. My biggest priority for the site right now is to get the email server working—I started a thread on the SMF support forums; feel free to follow along, there. If they can't give me a way to debug SMF's direct interaction with Google's SMTP server, I will have to change it so SMF contacts localhost, and my own MTA contacts Google.

Once that's done, I can take another look into the EDC, I suppose.

[hpg678]

That will require a little more work. My biggest priority for the site right now is to get the email server working—I started a thread on the SMF support forums; feel free to follow along, there. If they can't give me a way to debug SMF's direct interaction with Google's SMTP server, I will have to change it so SMF contacts localhost, and my own MTA contacts Google.

Once that's done, I can take another look into the EDC, I suppose.

When ever you get the chance!    No rush!

[Josh @ Dreamland]

The old server has been decommissioned—this migration is officially complete.  I am still investigating the email problem, but otherwise, the site should be at 100 percent.

EDIT[2017-10-21@UTC+20:08]: Email is fixed.

[Josh @ Dreamland]

We had another forum outage today. I've repaired the tables and upgraded the forum to the latest SMF. I believe that should be the end of it. Time will tell.

[hpg678]

We had another forum outage today. I've repaired the tables and upgraded the forum to the latest SMF. I believe that should be the end of it. Time will tell.

Thanks. I was wondering what was happening.

On another note, Here's wishing you a blessed and prosperous New Year and we move on into 2018.

 

[Josh @ Dreamland]

In what has got to be the last outage for a while, the EDC went down when I upgraded SMF to fix the tables. I am proud of SMF team's work in maintaining a stable API, despite PHP's maintainers' heads being lodged insuperably far up their respective asses. The bottom line is, some minor changes to the EDC's client code brought it back into compliance with new PHP APIs. No changes were required to continue using the SMF API. Let me know here if there are any problems with that.

We're discussing migrating to a dedicated server. It should mean less future headache, as I would probably migrate us to Arch to take advantage of rolling releases.

[Josh @ Dreamland]

Fixed the commit ticker at the top of the site.

Something I didn't mention earlier is that I have issued redirects from http:// to https:// in the interests of user privacy and security. I hope you were all using HTTPS before that, but now it's required. Anyway, this had the side effect of breaking the commit ticker, which did not understand the HTTP 301 response. I've fixed that by adding an S to the GitHub URL.

[faissaloo]

Fixed the commit ticker at the top of the site.

Something I didn't mention earlier is that I have issued redirects from http:// to https:// in the interests of user privacy and security. I hope you were all using HTTPS before that, but now it's required. Anyway, this had the side effect of breaking the commit ticker, which did not understand the HTTP 301 response. I've fixed that by adding an S to the GitHub URL.

HTTPS Everywhere FTW

[hpg678]

@faissaloo

Thanks for the tip.!