News State of Orbiter Forum, Orbit Hangar Mods

[IronRain]

Hi everyone,

We thought it might be nice to do a “state of Orbiter Forum” post to talk about where we stand and what we’re working on

Orbiter Forum
Some 2 years ago we migrated to XenForo. This step made our live a lot easier and we’re running up-to-date forum software.
This comes at a cost (but what doesn’t these days). Every year we evaluate the renewal of the license and so far we don’t see any reasons to not renew it.

The forum is currently running (well updated) on a CentOS 7 machine. @Xyon and I (but mainly she) are working on migrating it to either a Rocky 8 installation or the app-service that Digital Ocean is offering and we're pushing on the latter. We’ll keep you informed.

Orbit Hangar Mods
The current version of OHM is… less than ideal. @Xyon developed a new API, I’ve been working on a React-based frontend and we progressed nicely. Eventually, life caught up and development slowed down. We regularly talked about things that needed to be done, but both myself an her are busy with our personal lives as well.

With Orbiter being open sourced (more on that later) we saw a rise in interest in the project. This means there’s also a rise in interest for add-ons and OHM.
Therefore, we were considering the following options:

• Continue development on the API and frontend
• Integrate OHM functionality into OF

After some considerations we've decided to go-ahead with integrating OHM into OF.
There's a lot todo there in terms of migrations and planning but it seems like this is the best option. We're pretty excited about this.
This will also return some of the functionality we used to have on the vBulletin board (post creation, for example).

Orbiter Open Sourced
This one is a bit complicated

Martins open sourced Orbiter about a year ago on GitHub.
A couple of us (both OF admins and none admins) are members of the organization but not an owner or maintainer. This limits us in what we can do in terms of CI/CD, setting up pipelines, creating releases, etc.
Also, the main orbiter site is not maintainable by any us, as far as I’m aware.

With the source code now open, discussions can start on how to continue development (in fact, there’s already a thread for this). There are a lot options here so the discussions are good.
The short-term plan is to get a steady CI/CD pipeline working and then start merging PRs. There also needs to be a roadmap, a discussion on what to add to the releases out-of-the-box, how often a new release is published, etc.

Currently we’re setting up a new website, targeted on a fork called “open orbiter”. However, if we can get the right permissions for the GitHub organization I think we can drop the “open” part and just renew the current Orbiter site with what we’re building.

Like I said, tons of options just need to decide with each other on how to proceed.
If you have any valid input, please join the thread and add it there For those who've already shared their ideas - thank you!

That's it for now. I'll try to do these posts more often.

 

[4throck]

Orbiter Forum is good and works well both on desktop and mobile. Don't waste more time with it, just do maintenance as needed.

Orbiter Hangar has a two crucial usability problems:

• version and date information are missing;
• no categories (is multistage2015 the same as a DG repaint ?)
• search gives strange results;

I have no clue about what the exact problem is, but I have some experience. Being involved in the development of a retrogaming site (spectrumcomputing.co.uk) I know the frontend is relatively easy, if the database has the correct information.

If not then the database must be fixed. Fixing and adding missing data can be handled by the community using a few forms.
On the retrogaming site I mentioned we have this: https://spectrumcomputing.co.uk/zxdb/add/
It's a slow process but people use it and hundreds of titles now have proper release dates, genres, authors, etc.
For Orbiter Hangar it would be much simpler, but the principle would be the same (open an existing entry, add or correct data).

PM me if some help is needed, I'm more than happy to share my experience.

Orbiter Open Sourced
Don't think you should divert efforts for now; let's fix OH and that will take care of everything. We can have addons for OpenOrbiter on OH, or even a full OpenOrbiter release on OH if the license allows... If now we can always have a description-only entry with a link... It's all a matter of providing the right information - version, author, date, license, link, etc.


In short, focus and fix what we have right now, and deal with the future when we get there
Just my 2 cents hope it helps

 

[Xyon]

Orbiter Forum is good and works well both on desktop and mobile. Don't waste more time with it, just do maintenance as needed.

Orbiter Hangar has a two crucial usability problems:

• version and date information are missing;
• no categories (is multistage2015 the same as a DG repaint ?)
• search gives strange results;

I have no clue about what the exact problem is, but I have some experience. Being involved in the development of a retrogaming site (spectrumcomputing.co.uk) I know the frontend is relatively easy, if the database has the correct information.

If not then the database must be fixed. Fixing and adding missing data can be handled by the community using a few forms.
On the retrogaming site I mentioned we have this: https://spectrumcomputing.co.uk/zxdb/add/
It's a slow process but people use it and hundreds of titles now have proper release dates, genres, authors, etc.
For Orbiter Hangar it would be much simpler, but the principle would be the same (open an existing entry, add or correct data).

PM me if some help is needed, I'm more than happy to share my experience.

Orbiter Open Sourced
Don't think you should divert efforts for now; let's fix OH and that will take care of everything. We can have addons for OpenOrbiter on OH, or even a full OpenOrbiter release on OH if the license allows.... It's all a matter of providing the right information - version, author, date, license, etc.


In short, focus and fix what we have right now, and deal with the future when we get there
Just my 2 cents hope it helps

Appreciate your thoughts and insights. I'll expand a bit on what @IronRain laid out above just to clarify that the intention is to close the current bespoke OHM site entirely, and to use the XenForo "Resource Manager" system (which we obtained when we got the XenForo license) to perform the work that OHM has been doing.

The main reason is time. All the problems OHM has can be fixed with sufficient time and coffee, but as @IronRain said above he and I have both had life catch up with us a bit and the project has stagnated. It's a difficult thing to decide to retire something you've worked hard at for years now (the original API was laid down in 2017 iirc) - the backend represents a significant investment of my time and effort, and I know the front-end (which hasn't even been released) has been a similar labour of love for @IronRain. But we have to be realistic about our capacity to manage those projects going forward, and consider what the best option for the community is. Right now we have a wave of new enthusiasm for Orbiter in the wake of it going open-source, and I think it'd be a crying shame if a clumsy broken website were to dampen that down.

That said, there's a lot to like about what I'm seeing with the resource system, and I think it covers basically everything we would want out of an add-on repository. Alongside that, it'll house the old versions of Orbiter that OHM currently holds, and in the future releases of the open-source version can also be distributed via there. Our intent is to make this as straightforward as possible both for add-on developers, consumers and forum maintainers alike - which is definitely not what I would say the state of the current OHM site is at present.

 

[GLS]

I'm not sure if this is being worked on, or has been forgotten, but there were several (IMO) important Orbiter tickets in the old forum, that should be "visible" somewhere so people can now fix them in Orbiter.

 

[4throck]

Appreciate your thoughts and insights. I'll expand a bit on what @IronRain laid out above just to clarify that the intention is to close the current bespoke OHM site entirely, and to use the XenForo "Resource Manager" system (which we obtained when we got the XenForo license) to perform the work that OHM has been doing.

If an existing tool does the job then use it!
As long as the current add-ons aren't lost, no problem. If extra work is needed for DB migration, my remarks about legacy DB cleanup and eventual community help are still valid. Moving to a system that uses current forum member login/authorizations is actually well suited for that.

 

[jedidia]

@IronRain, you didn't mention the financial situation, so I assume you're doing good on that front?

 

[Xyon]

I'm not sure if this is being worked on, or has been forgotten, but there were several (IMO) important Orbiter tickets in the old forum, that should be "visible" somewhere so people can now fix them in Orbiter.

Nothing from the old forum was lost. All the tickets remain, but we as yet have not arrived at a solution for managing them.

I might convert them into issues for the GitHub project, actually.

If an existing tool does the job then use it!
As long as the current add-ons aren't lost, no problem. If extra work is needed for DB migration, my remarks about legacy DB cleanup and eventual community help are still valid. Moving to a system that uses current forum member login/authorizations is actually well suited for that.

Yes, the existing addons (all several thousand of them) will be moved to the new system.

@IronRain, you didn't mention the financial situation, so I assume you're doing good on that front?

As of this moment the forum's liquid funds are just slightly in excess of 550 USD, with an additional ~80USD "credit cushion" which we maintain with our provider, Digital Ocean.

The month-to-month running expenses of the system are, broadly, covered by your generous donations.

Our adjustments here will, on the whole, not significantly change our outgoings, though while we're prototyping there has been a slight additional cost as we're running new services up side-by-side with the current ones. I would consider the financial situation "stable".

 

[GLS]

I might convert them into issues for the GitHub project, actually.

That would probably be the best solution: put there the original post and any relevant subsequent posts. I can do that, if I get with the tickets.

 

[Xyon]

That would probably be the best solution: put there the original post and any relevant subsequent posts. I can do that, if I get with the tickets.

I'll see what I can do, but I will prioritise completing the migration of the forum to the app platform in the first instance.

 

[Max-Q]

What will happen to all of the links to addons on OHM? Some kind of deal to redirect then to the new location on OF? Else there will be so many dead links that it wold be almost impossible to sort out what addons require what dependencies and where to get ’em.

 

[N_Molson]

There's a real lot of work to do, but it definitively needs to be done.

 

[steph]

OF is just fine. I don't mind Hangar being how it is, but perhaps that's just because I already knew it since way back. OF is good for refreshing the 'good' add-on list

 

[Xyon]

What will happen to all of the links to addons on OHM? Some kind of deal to redirect then to the new location on OF? Else there will be so many dead links that it wold be almost impossible to sort out what addons require what dependencies and where to get ’em.

This is harder to fix, but will be dealt with during add-on import. We'll have to have some kind of mechanism listening out on the old OrbitHangar URI which can take the add-on's old URI and send you to the new page for that add-on.

If you recall how OHM used to work before it had an API, I've already handled this once before when we moved to the API and each add-on which existed prior to that update still links into its old URI. So something like that will be the sort of solution we're after here.

 

[Majid]

Thank you for providing years of free support to the community. It can't be appreciated enough. I was looking into the XenForo Resource manager and wonder if there is a way to serve assets via an API or publicly somehow. OHM to my knowledge has never required a login to allow for downloads, and that's always been appreciated. Plus requiring a login could potentially kill (or make it problematic to implement) some really cool use cases like in game one click to install add-on management systems.

 

[IronRain]

Thanks!

Good question and a very important one to note.
The resources manager wil use XFs permissions system and I'm also sure we can allow anonymous access to these areas.
We need to investigate this, but the focus for now is OF migration to new infra.

As for APIs, XF has APIs which are nicely documented. I've been looking into developing an add-on manager after the OHM stuff so we might make it a community effort if there's enough interest.
One thing to note here; the API is not open to use at this time on this board. If there's interest in this, like a specific well defined use-case, we can talk about it. The API can be scoped in terms of permissions but still give a great deal of power so I'd like to be careful with it.

 

[Majid]

......

As for APIs, XF has APIs which are nicely documented. I've been looking into developing an add-on manager after the OHM stuff so we might make it a community effort if there's enough interest.
One thing to note here; the API is not open to use at this time on this board. If there's interest in this, like a specific well defined use-case, we can talk about it. The API can be scoped in terms of permissions but still give a great deal of power so I'd like to be careful with it.

The specific use case I am interested in is I'd like to create the AUR of Orbiter. Except restricted to a set of curated, very high quality add-ons, with their dependencies expressed, and some lua scripts to allow for 1 click to install. 1 click to install means that the add-on has to be downloaded onto the client system automatically, and for that it'd be super convenient if the add-on files were available publicly with stable links.

The mod repo I'm working on (orbiter-mods.com) will ofc also allow add-on developers to upload their files directly, but in case the files aren't there I'd still like for people to be able to have 1 click to install experience by just providing the metadata/scripts needed to fetch the dependencies and install the add-on.

 

[Xyon]

Thank you for providing years of free support to the community. It can't be appreciated enough. I was looking into the XenForo Resource manager and wonder if there is a way to serve assets via an API or publicly somehow. OHM to my knowledge has never required a login to allow for downloads, and that's always been appreciated. Plus requiring a login could potentially kill (or make it problematic to implement) some really cool use cases like in game one click to install add-on management systems.

Yes, addons hosted on Orbiter-Forum will be downloadable without authentication just as OHM permits.

API access is offered via API keys, but you can definitely put in a GET request to be served the resource file as far as I can see from the API spec. We haven't opened up the API to users really, at this point, and it might well be that it's permitted only for some users on a case-by-case basis depending on your use case.

 

[Max-Q]

So, the other day, I tried to log into OHM, and it won’t let me in. Is it now impossible to get into OHM? Posted a thread in the OHM Bugs section, but no response.
https://www.orbiter-forum.com/threads/cant-log-into-ohm.40512/#post-595503

 

[Xyon]

So, the other day, I tried to log into OHM, and it won’t let me in. Is it now impossible to get into OHM? Posted a thread in the OHM Bugs section, but no response.
https://www.orbiter-forum.com/threads/cant-log-into-ohm.40512/#post-595503

Yeah I think it's broken, I don't know quite why. But given the site will be retired very shortly, it's unlikely to get fixed.

 

[jacquesmomo]

Hi

I may be a little late to this discussion...

I totally agree with 4throck (and also with the others) :
Orbiter Hangar needs a complete and global reworking and above all a better advanced research function

For the forum, I think it's very good.

so yes for OH....