Website maintenance thread

So the answer is: Because we are monkeys? :monkey:

1 Like

That’s a mistake we made. IMO that’s why we are in this situation.
Don’t misunderstand, it’s not about this or that software, it’s about commitment and availability.
The thing is, most of us are web devs in our daytime job, and we probably could easily handle this… the problem is… we don’t have the WILL to do it.
We are making an open source 3D engine in Java… that’s what we want to spend our spare time doing… not administrating a website or developing a web app as we do at work.

IMO We should defer this effort to someone/something else. So we have several options :

  • Use a rock solid paid infrastructure with Wiki/forum/landing page services, that would require just some admin customizations and that’s it. But usually to have exactly what you want… well you have to pay, it can get expensive, and we don’t have regular income.
  • Hire a web dude… pay him so he can handle those issues on the spot. It could be even more expensive though…
  • Hand this to the community… put the wiki on github, have a landing page on github.io (basic html one, no fancy app), at the expense of poor customization and basically no branding. And… well, people from the community will have to contribute… The forum on digital ocean seems fairly stable now so let’s cross fingers…

In any case it’s a tough choice, but even when it’s decided… the move is already a lot of work… so we’ll need you monkeys…

1 Like

Ok so I guess I will break out of my usual lurker self and go ahead and say I would be willing to help in any way I can when putting the wiki on github

We could devide the work among us. I would volunteer to put some of the beginners page to jmonkey wiki. depend how mich work this really is. But then you guys would have to give access to the github wiki to some of us (or give me some pointers how that is done with a PR as I have no clue). I’m already familiar with this wiki thanks @pspeed zay es wiki :slight_smile:

Why not just find a largish number of people from the community and give them all ssh access + setup a Slack to log all changes? That way, there should always be someone available to fix it if it breaks. If you’re concerned about someone going in and causing problems, then only pick trustworthy people and probably have automated backups as well with more restricted access. This probably woulden’t cost much, if anything, more than it does now to get the actual server and it should give high reliability as there would be someone available almost constantly.

Thats what we do/did, theres at least five people who have access and know the infrastructure somewhat.

Then why is there still downtime (I don’t mean this as an attack at whoever those 5 people are, but, rather as a question of problem solving)? If it is because all 5 people are consistently unavailable, then perhaps adding more people would be a good solution?

I don’t think that would solve the problem. Actually adding more people got us in the situation of having a dodgy server/service.

If you want to limit the number of people, then perhaps swapping out some of the existing people for others who are more consistently available would be best. On a Minecraft server I ran two years ago, I had moderators from nearly every time zone except those that are mostly uninhabited. As such, going on it at just about any time of day one could find live support. I think that a good plan + multiple people would fix the issue. TBH, having a dodgy service is probably more due to poor planning then having too many people. There are lots of very large companies that have very stable websites.

Additionally, to solve the “too many chefs in the kitchen” problem, it should be fairly easy to just have one person who makes the decisions on how things run and needs to approve any sort of changes, but then have multiple people who are able to go in and fix things if/when they break.

Again… we are not administrating a web site we are making a 3D engine. The 5 people that worked on the web site weren’t there for that in the first place, they were there because they are JME users. They nicely proposed to attend the web site, but, as it oftent the case in open source their availability decreased and they ended up not showing up. I don’t blame them be cause… life you know…
We could gather another bunch of people, organize them to do what you suggest… but tbh I’d rather have them work on the engine than on the web site of the engine.
We need something that need only minimal administration.

Just put it on github. As I already said I volunteer to help with that wiki, but I would need access (I guess I only have access to the zay es wiki, I did not check so far as there is still not plan). On github, github will care about uptime of the page. Just a simple landing page and some links to the wiki/forum/download and that’s it. No?

But as long there is no decision I can volunteer as long as I want it will not help…

1 Like

Thanks, we will make a decision.

I want to know honestly: is the website dying? If no one is working on it and we’re not interested in getting people to work on it, is it just a matter of time before it fails completely and never returns?

I signed in thinking of asking a question about a project I hope to work on, but I discovered that this site has far more important things to deal with than my insignificant questions. If the website does not work, then the engine may as well not exist for me since I have no hope of using it effectively without the website’s help!

You have the most important bits of the wiki right in the SDK. Just press F1.

In the mean time, you also have the forum.

Just be patient with us. Not only do we not have access to the wiki machine but apparently no one who wasn’t abducted by aliens still has a backup of the site. We are trying to get a copy.

1 Like

We are not dying, at worst the current machine hosting the wiki is.
We could find another machine and put the wiki on … but we’d end up with the same issue in 2 months eventually…
That’s the reason we are looking for an alternative, to get rid of the problems altogether.

If you have questions, even basic ones, please ask on the forum, questions are usually answered pretty fast.

1 Like

Wait a minute, how are the forums still up? Separate host? Why not just move the entire thing to that one?

The forum has been moved on digital ocean

@nehon I am already paying for hosting of my game web site and have the ability to host unlimited domains for free on the server. No bandwidth or storage limits on it and it’s a fast box. You can test out the speed if you want. My game site is www.quantumgamingstudio.com

I am willing to set you guys up with your own user account and give you access to cpanel if you are interested … for free of course. Of course you would need a backup of the old site and you would hopefully still have access to your DNS provider to redirect the IP.

Seeing as how I am now married to JMonkey for my game and it’s not just a side project it’s the least I can do for all the help you guys provide.

4 Likes

Tutorial
https://web.archive.org/web/20160306073823/http://wiki.jmonkeyengine.org/doku.php/jme3#tutorials_for_beginners

1 Like

Something I think is also important is whoever gets the wiki going again should document how all the pieces fit together. That way if something goes wrong again, we can get things going again quickly on another server. No more single point of failure.

I think what went wrong here isn’t that servers failed or that system access was revoked. Its that there is almost no possibility of recovery in the event of disaster. We need to keep that in mind with whatever solution we come up with.

We just very fortunate that the same thing didn’t happen to the forum… yet.

I tried building the wiki from the github repo the other day. Maybe we should focus on getting that working first?