Oolite Bulletins

maik wrote:

In our case we have >2000 content pages (>5000 pages if you take redirects, user talk pages etc. into account). Given that templates cannot be migrated, history is lost, and markup will need to be converted partially by search-replace (which still needs manual quality assurance), and partially manually, the effort needed to migrate those sounds prohibitively high to me.

I don't think that losing the history will be much of an issue during a migration. Yes, the work needed could/will be quite high, but if the wiki maintainers pool their resources for the common good I think it could be done quite easily and relatively quickly and give us a clean wiki that will be more or less future-proof (definitely more future proof than the current incarnation). Yes, my preference for pmWiki is clear, but the reasons for that, as stated, are that it's single-tier (no clunky DB connections, no reliance on a second remote server) and thus far more portable. I also propose a regular backup via the web client that will create a copy of the site that can be installed quickly and easier on any server with PHP support without the need for an accompanying MySQL server.

If the wiki admins consider the workload prohibitive then fair do.

I have done 80% of fixes to the OXP List page now - it's readable now.

maik wrote:

Looking at this thread I don't hear from anyone though who would be interested in taking a copy of the keys to a new server though... are we wasting our time?

I've got a fair bit of experience with Linux servers, Apache, and MySQL, and I've used MediaWiki enough to be able to admin it a bit. I wouldn't be able to put a lot of time towards it but I could help with any initial setup rush and cover for you if you needed an extended break. (I don't have a scanner, but I can send you a GPG-signed email which should let you verify who I am easily enough)

cim wrote:

maik wrote:

Looking at this thread I don't hear from anyone though who would be interested in taking a copy of the keys to a new server though... are we wasting our time?

I've got a fair bit of experience with Linux servers, Apache, and MySQL, and I've used MediaWiki enough to be able to admin it a bit. I wouldn't be able to put a lot of time towards it but I could help with any initial setup rush and cover for you if you needed an extended break. (I don't have a scanner, but I can send you a GPG-signed email which should let you verify who I am easily enough)

Thanks, cim. This is exactly what I was looking for. In general, I'm happy to spend time administering the server as needed. But there are times when real life takes all of my time and this is when there should be someone else available to look after things.

Additional help might be needed with migration activities. I think I can automate most of it after all, even if Winston does not resurface. Currently I am looking at scripting options to get the media out of the wiki.

I'll PM you with my email address for the GPG-signed email.

Progress:

• I sent authentication information to cim for access to the server.
• Getting media in an automated way from the Wiki works now, too
• I have a list of Wiki pages that I would exclude from history migration due to their size at first. This is based on page length and number of revisions.

So, shall I go ahead and set it up for good?

I'm all for a reliable and user-friendly wiki, so you get my vote.
I have been thinking, though, and just have one question:- what will happen to the old wiki? We don't really want two in existance at the same time.

My idea for the project tasks looks like this:

• A) starts when we say "yes, we want a new wiki".
  • Migrate content from the old wiki. Actors: me and whoever volunteers and knows his way around MediaWiki to become admin. Duration: until done.
  • Perform quality assurance (check for broken links, missing templates etc.). Actors: everyone who volunteers, no prior knowledge required. Duration: until C concluded.
• B) starts when the content has been migrated. Actors: MediaWiki admins. Duration: until all pages on old wiki have been marked for deletion.
  • Create wiki user accounts for everyone else who either authored pages before or requests one here on the BB
  • Monitor changes to old Wiki and and keep new Wiki in sync
  • Mark pages on old wiki for deletion.
• C) starts when B is done: Point oolite.org website and BB (including wiki keyword) to new wiki. Actors: aegidian(?). Duration: a few minutes.

Update:

• All images are uploaded
• Some pages are uploaded for testing purposes (main page, Activities OXPs, OXP, OXP List). The latter two without history, the others with history.
• Main page is the old Oolite main page
• Skin is changed to monobook (which was default in MediaWiki 1.15 which is running on the alioth.net server)
• Oolite icon is used for main wiki icon

maik wrote:

My idea for the project tasks looks like this:

• A) starts when we say "yes, we want a new wiki".

I haven't participated in this debate so far. The main reason is that I'm fine with the Elite Wiki as it is, and have no desire to start a new Wiki.

Well, it turns out some guys were not fine with the Elite Wiki as it is, did have the desire to start a new one, and created it. And it's up and running now.

So now we have two Wikis. And here begins my problem. I have recently uploaded a new OXP and documented it on its Wiki page. This required me to edit three different Wiki pages: The OXP page itself, the OXP list page, and my user page. This in itself is already a little awkward, but I have learned to live with it.

However, now there are two Wikis. Maik noticed that I had edited a page in the old Wiki, and immediately made an account for me on the new Wiki and sent me the login details, in order to enable me to synchronize the new Wiki with the old one. That's a nice gesture, of course.

However, it means that now I find myself involved with the new Wiki involuntarily (which I have no desire to start in the first place). So now I have to edit six Wiki pages instead of three. I have done it this time, and again for another OXP update I just committed (four pages instead of two). But actually I don't want to maintain two Wikis.

So, as I now became officially involved with this new Oolite Wiki project, I have to ask the fundamental question: Do we want a new Wiki? Do the pros (especially better maintenance) outweigh the cons (especially severing the ties that Oolite has to Elite by leaving the joint Wiki for all things Elite)? Is there a consensus (or something resembling a consensus) for this project? Or do most of the people simply not particularly care either way?

Personally, I am not for the move. But I would follow the masses eventually. However, I do not want the burden of having two Wikis put on me. As long as there is no official decision (whatever "official" means in the context of Oolite) to abandon the Elite Wiki and move to the new Oolite Wiki—specifically: as long as the 'wiki'-tags on the boards point to the Elite Wiki (as in [wiki]Oolite[/wiki])—I will continue to treat the Elite Wiki as the official Wiki. This means that I'll only link to the Elite Wiki, only edit the Elite Wiki, and won't feel responsible for synchronizing my edits on the Oolite Wiki.

This is actually quite difficult. I don't have a strong opinion either way, but, in my view, there are a few things that would need to be sorted out if we are to move.

-First, i would want a "Yes, I'm OK with this from Winston".
-Second, the Oolite part of the Elite wiki should be shut down or we move all of it.
-Third, if we don't take over the whole wiki, it would be nice if the Oolite menu on the first page of the Elite wiki points to the new Oolite wiki.
-The wiki links on the bb should be updated. Monumental task, I'll help, but should be done in my view.

Gimi wrote:

First, i would want a "Yes, I'm OK with this from Winston".

This is what started the discussion for the move for me. Winston has last been seen on the BB on 13.06.2011 (at least that was the last time he logged in). I have two PMs to him sitting in my outbox, the oldest one being >4 months old, meaning he did not open them (difficult as he has not been on the BB since then, but he should have received the emails from the BB telling him there are new messages). I have also tried to contact him via email (managed to find something on the web) but did not get a reply.

The Elite wiki says to contact Winston via this BB to get an account. This is on the main page which is locked. So, if you only do as requested, you never get an account because your PMs are not read. Obviously this does not matter for many of us regulars because we do have accounts, but new people are left out. Not very wiki'ish.

Also, Winston is the only one to have Bureaucrat rights. No one can take over from him if he abandons the wiki (and by all means, it seems abandoned by him to me).

Altogether, this is IMO why it makes sense to move on.

Moving on: The new wiki seems to be stable. I'd like to give it a few more weeks during which there will be double work updating both it and the current one (and if someone does not have enough time to do edits on both I'll pick it up as soon as time permits). But then I would propose to do the move.

As the one who started this topic I have to be honest and say I'm not entirely satisfied with the "end result" of the discussion. The real point of the debate was to consolidate the hosting of the oolite.org site, associated files and the wiki into one site under the oolite.org banner under a hosting package that would allow the distribution of administrator access to apply patches, fixes and addons to make maintenance easier and not reliant on the whim of one person who may or may not be available.

I have to admit that initially, separating the Oolite part of the wiki out from the main one at Alioth was not part of the concept in my head, but since the Oolite content seriously outweighs the rest of the site then it does make sense. Once the two are consolidated (which hopefully won't take long) then the Alioth content would be removed and linked to.

Maik has done a fantastic job moving the content, but in a sense the situation is now worse. Instead of three different sites we now have four - oolite.org, forum, Elite wiki and maik's wiki. Since no-one but us maintains the Elite wiki, I fear that it may fall behind and become the forgotten son of Oolite's wiki - my opinion is that the whole kit and caboodle should either (1) be moved over or (2) the Elite content finalised, Oolite content made to be similar to the Elite stuff (i.e. summarised and made static, linking to the Oolite wiki for more information) then handed off to Winston to maintain since he is reluctant to allow us access to it.

I'm not pushing the concept any more, since the community doesn't seem to bothered by the current state of the online presence, but I want to state for the record that, from a web designer's perspective, something still needs to be done to consolidate the content and make it centralised, rather than relying on three separate entities (and a file hosting service) to make our community work. Oolite is a relatively big thing in computing circles and there is a lot of respect out there for what Giles, Jens and the rest of you have achieved and I worry that it will only take one hosting package to fail and the community could be fractured.

I'm maybe worrying about nothing, but I see too many web sites that hosted good material and good content falling down because all of their files are gone or because their database fails, and I don't want to see Oolite go the same way.

I am with Selezen on this inasmuch as I though the original idea was to 'unify' all the various sites. The current situation is that we have separate wiki, BB and 'oolite,org' and the 'real' problem is that access to two of the three is limited if not impossible.
In a sense I started this debate when I commented on oolite.org and how it seemed rather out of date. The reasons for this are manifold, but a major factor is that nobody has access to it and therefore willing hands are un-able to update it and improve it.
This led on to debate about the wiki. The problem here is lack of access to new members. For those of us with a wiki account, things are fine, but anybody wanting a new wiki account is stuck unless thay are able to catch Winston's attention, which as Maik points out is all but impossible.
The BB is fine as far as I can see.
Personally I have no real problem with the sites all being in different places, although I do like the concept of having the whole thing under one roof. The important issue is one of access though, and Maik's new Wiki will address this as far as the Wiki goes. However, separating it from the main Elite Wiki will clearly have some impact on the Elite Wiki particularly as it seems we are the only people maintaining it.
What it doesn't do, though, is unify everything under one banner, and the original point made about the lack of maintenance of oolite.org is not addressed either.
I'm beginning to think that this is a problem without a clear solution, and that being the case I wonder if we should back-off a bit to give time for more debate.
The truth is, unless we can bring oolite.org and the wiki (and maybe the BB as well) together in a way that those with an interest can both access them and modify them we are not going to achieve what we want to do. And that will require those who currently have control over the various sites to both participate and relinquish some of the control they have.
When we started this discussion I was expecting movement towards an arrangement where oolite.org was the 'shop window', a site where people could find out about the game in general, download it, and see what was on offer in a well laid out and up-to-date gallery featuring both the core game and a wide and representative range of OXPs. It would link to the Wiki (which would be as it is now but was easier for new members to gain access to), and the BB which as I said seems fine as it is. In an ideal world, all three would be hosted together, and there would be a sufficient number of administrators to make it work.
If this cannot be done there seems little point in continuing, but clearly we can't leave things as they are either. I think we need to set some definite (broad) goals before we go much further. We need to agree on exactly what we want first, then set about the details.
So what do you all think, and where do you think we should go from here?

I agree with Smivs, it would be good to have more unification. I demonstrated that we're not stuck with the current wiki if we decide to move things and don't have to loose anything in the process. For the time being, I'm handing out accounts to the new wiki to people who either ask for it or who I see changing pages on the existing wiki, asking them for their time to keep the two in sync with regards to their own edits. If time doesn't permit, I pick up changes, and I have seen cim do the same. I'm happy to do this as long as we don't reach a conclusion on what we actually want to do (as time permits), so that if we do decide for a move, we don't have a stale state on the new one.

Both oolite.org and alioth.net seem to belong to aegidian. Maybe he could chime in with what is possible here: What kind of access do you have to the servers (full root level, or just content)? Would it be possible to share server access with other people here? How about unifying them under oolite.org, bb.oolite.org, and wiki.oolite.org or some such? This doesn't have to result in one machine, they can all happily reside on different ones. Physical hosting is completely separate from logical addressing.

To continue further with Smivs' point about the out-of-date-ness of oolite.org: If we actually talk about changing the underlying technology then I would propose a content management system such as Joomla!. In contrast to the wiki, the website is more static and changes should go through a review process. It also makes sense to have changes only go live in the future and not immediately, e.g. writing the updated portions of the website before a new Oolite version goes live and having all changes appear publicly at the same time when the new version is there. This is nothing that a wiki is made for.

maik wrote:

Both oolite.org and alioth.net seem to belong to aegidian

alioth.net belongs to Winston. Only Winston has access to the back end of his domain and he ain't sharing.

maik wrote:

I would propose a content management system such as Joomla!

With all due respect to other people's opinion, I have used Joomla and it's a bloated, wasteful, slow piece of Zaerlan monkey poop. I used it for about a year before working out that just about everything else in the CMS market is better. Additionally, most of the useful plugins for Joomla are commercial nowadays, and casual developers are abandoning it in droves.

My suggestion for gpEasy was based on the fact that oolite.org doesn't need much in the way of maintenance or content, since the majority of the dynamic content is on either the wiki or the forum. The last thing oolite.org needs to do is be a fully dynamic application with all-singing, all-dancing features. All it needs to be is a more or less static site that can be updated easily (when new screenshots, new versions or new documentation is added, for example). Joomla or WordPress or another of the bigger CMS apps would be ANOTHER database with ANOTHER complex administration system to use. gpEasy allows editing via the front end of the site and doesn't need a degree in web development to use.

Selezen wrote:

maik wrote:

I would propose a content management system such as Joomla!

With all due respect to other people's opinion, I have used Joomla and it's a bloated, wasteful, slow piece of Zaerlan monkey poop. I used it for about a year before working out that just about everything else in the CMS market is better. Additionally, most of the useful plugins for Joomla are commercial nowadays, and casual developers are abandoning it in droves.

Thanks for sharing that. Saves me some probably wasted effort on another private project where i want to replace a site that some Internet agency developed with handcrafted CMS features that inhibit layout changes by the owner. Joomla was really just an example.