Ulthuan

Cor blimy, has it really been 8 months since I last updated the blog? Whoops. Sorry.

Well, as you might have gathered, there's been very, very little done to the site, technologically or similar over the last 8 months. A few new forums for 8th ed, the odd april fools joke, but nothing spectacular.

Despite still being on official sabatical, I've been discussing Province with a few friends again today (one of which may have a job for me involving coding php full time), and I really need to get back onto finishing it. With this in mind, I'm going to try and update the blog again, and keep it up to date with province stuff.

Sadly, I still have a real life, and it's seriously getting in the way of my ability to code. I'll try my best.

In other news, if you want LJ access to the main site news, it has a LJ URL now: ulthuan_news. That'll give you the same news used on the home page, but it only stores stuff for two weeks, so stuff will only appear when there is news to be had.

Have fun

- Sun

Happy New Year on a personal note. Hope you all had a good festive season, whatever you celebrate.

So, today I've done a decent quantity of work on province, and have very little physical stuff to show for it. That's because I've been doing lots of reading on the subject of balance in MMOGs, and discovered that I really need to be implementing the troops and combat system before doing any more balance testing on the building rate of growth. Basically I need to put more of the outputs into the system.

I've also had a few new ideas that might be worth trying. The first is limited storage capacity, so you're limited in what you can hold, and a certain amount goes moldy each night. This requires monitoring the day cycle as well as hours, but could work. It should at least stop the problem I had last with 2.1 billion food. I only stopped there because the storage capacity for a single number on the system is 2.1 billion. (the way I coded it anyway). The second idea is upgradable buildings. So for an extra cost you can upgrade a building to the next type rather than just building another one. The third idea is having a town hall. This dictates what sort of buildings you can build (Currently have 5 levels, and to give you an idea of how much resources are required, upgrading to level three currently costs 13000 Food, 7800 Wood, 1560 Stone and 130 Gold). The Town hall is an upgradable building BTW. The fourth idea is actually one Talh suggested a while back which is to have available buildings based on town level.

I've just had a fifth idea whilst typing, which is to have all buildings have to be researched first, so you have to first work out how to build them, and then actually do it. This could equally apply to troops and anything else.

So, not been idle, just not been coding either.

- Sun

Hey, guess what? I've spent three days reworking part of the home page, and I bet you not a single member would notice without being told.

What you'll see on the home page is that the news has been updated. What you don't see is *why*. Previously, the news on the home page was generated using the software the home page uses, which is separate from the forums. However, most of the Lores don't actually want to be bothered to post news to two places, and as most of the members don't wander outside the forums (pity, there are over 200 pages of information outside the forums), the news gets posted to announcements, and from there read and forgotten.

Yet I care about the home page. It's what the new members see when they start on the site, and really it's out advert. As a result, it should be up to date.

So, over the last three days (I know, I should be coding Province, I got distracted!), I have beat, bullied, and bashed the system into automatically picking up new announcements and displaying them on the home page. It took me three days because there are three distinct challenges to overcome. The first is working out how to get stuff out the forums (RSS), the second is working out how to get it onto the home page (RSS reader), the third is getting the blasted thing to output the right information correctly. Getting the interface working wasn't a problem, that was day one. Day two was spent bashing my head against bbcode. Day three consisted of entirely re-writing a new bbcode parser, and then re-writing the output routine of the RSS reader to output useful HTML rather than useless stuff.

But, as a result, the home page now automatically displays the latest news. There's a few bugs, but nothing serious.

It appears I may have tomorrow off work. Assuming I don't get dragged down by some of the stuff I've been putting off from reality a little too long, I may get a few hours coding done on Province.

Seasons greetings from the Loremasters and other staff of Ulthuan. Whatever you celebrate, let it be a good one.

Due to work and seasonal festivities, all work on Province has paused for the time being. I hope to be able to return to it towards the end of January.

I hope you all enjoy midwinter, 'cos I know we will.

- Sun

Hey look, this is turning into a blog for Province! Well, no it isn't. It'll turn into a blog for whatever project I happen to be coding, which is currently province.

The building system has been rebuilt from the ground up. The database structure has been edited, making access speedier (requiring less and simpler queries) for the simple building stuff. As a result of rebuilding the building system, the resource allocation system (the one that gives you your new resources when you sign in) has also been almost completely re-written. OK, so I kept the population increase code, but everything else really has just been rewritten. That's 150 lines of code on it's own. Nothing in this project is small or short, but that's what I get for trying to code a cgi project in php. I am however happy with the work I've done, it's something to be fairly proud of.

Buildings now have a hidden "production type" value so I can split them up in the database. Type 1 buildings give their database production shown as a simple per hour value. Type 2 buildings will be click use buildings, when I get around to coding them. Type 3 buildings give a % boost to production, and as a result were easy to code (copy paste the type 1 code, boosting up the production already accumulated rather than adding to it). Type 4 is currently labeled "manual coding", which are things that have to have their own lump of code to work. Currently there are two type 4 buildings, the Building Site (yes, it's considered a building!), and the Market (that's only on my version).

The simplified process for dealing with type 1 and 3 buildings means that new ones can just be added to the database with no coding requirements, and will work as expected. I consider this as great an achievement as anything else this weekend, as previously the 3 buildings were all handled in their own code, which was not going to be scalable.

On the other hand, this was not the code I was planning on re-writing this weekend, so I'm slightly annoyed. I need to be programming Type 2 buildings and contemplating the next bit of code.

I also need to remove the sale code at some point, as that's now redundant. The entire project requires a bit of code tidying too, to determine if there's any serious sections that are completely unused and unwanted.

- Sun

In my last entry for "several problems" read "two problems". both of which have been sorted now.

So, I appear to have a working alpha test. This is very much an alpha test, there are several bits of the system that aren't even coded yet. These include clickable buildings, troops, a skin/theme, and means of damaging buildings, to name a few I can see quickly on the main page.

However, the resource collection and spending systems do work, and the whole thing does actually work on the server.

As a result, I'll be PMing all those who requested alpha accounts, and setting up a forum on site for those same people, so we can start doing more heavy testing.

OK, so I've nailed down one set of errors. The local php version and the server version are different. The server is running php 4.x, and my system here at home is 5.x. This makes a radical difference, as I'm using a quantity of php 5 only code. This has resulted in literally hours of work trying to get the coding php4 compliant with no simple way of testing it. It's like bashing my head against a brick wall.

Anyway, after what must be 6 or more hours work, I have the database connection working (and with a darn sight better code than before). There's still several bugs to be sorted out before a test can be announced, but it's getting there.

I didn't think it would take me so many weeks just to transfer the system from my local server to the live one.

- Sun

Right, well I promised an alpha test.... and I've hit a rather major flaw. Having got what little work done this weekend as I'm going to, including coding sale for gold and recoding the registration system, I uploaded the files from my server to Ulthuan's and something in the registration system won't display. I don't know what, I've not had a decent chance to debug it. But the files are there, and not loading correctly.

The whole project needs to work as well on the Ulthuan server as it does on my test one. If the system doesn't load in the first place.... well it sorta stops this project in it's tracks.

Next weekend I'm working again, but I'll try and get an evening or two on this one, work out what's not happening.

I'm tired, so tired. That comes from a variety of reality issues that are clogging up the works. Reality (as always) takes priority, so I've only had an hour or so working on Province this weekend. However, with Tuesday off work, and nothing better to do but stress about reality, I may be able to get a few hours of coding in. If I had a laptop of some sort running Linux I'd be able to get more done, but I'm basically limited to working on my desktop. Not a bad system, but not very portable.

I'm starting to think about the trading system (currently programming trading with the system for gold), and am concerned about how much it's going to strain Ulthuan's already rather strained server. When I started this project, my initial plan was to run it client-side (so it uses the processor on your machine at home, not the server), but it's not turning out that way, and it's currently rather heavy on the database access. Some of the more observant may have noticed we had a couple of hours downtime a couple of days back, when the server froze up and none of us noticed. People, please email, MSN, IRC someone if there's a problem. Or comment here, it's open comments, and it emails me. Once we know, we can get it rebooted.

Having said that, I've set Ulthuan up to do daily reboots now. Not sure if it has been, that's fairly hard to check, but it's set up to.

Yet I somehow don't care if Province strains the server, even past breaking point. Being the sole coder on this project, and the system being fairly independent of GW, the whole system could be transported out and used elsewhere without too many problems. This means in time it could get it's own server, and be run beyond Ulthuan. While the ideas are not all mine (Tahl gets a lot of credit there), the code is, and that means a lot to me.

ETA for a fully working finished campaign is still this time next year, but for one man putting in around 4 hours a week, that represents a rather small 200-250 man hours of work, which isn't much for a MMO project like this.

(So, if I could afford a team of 10 people working 6 hour days, I could achieve much the same in a paltry 5 days. Sadly, I can't afford that realistically)

I got all last week's comments for testers, you're all recruited. Anyone else who wishes to help test, leave a comment, include your forum username, and you too can join in the fun. Exclusive to you nice folks who actually read the blog!

So, I had that week off as I said I would, and went to a very fun party thank you very much. Yes, I got nothing done last weekend. You might also have noticed that I'm *only* getting stuff done at weekends, and that's because I want my evenings free to do other stuff. Next weekend will suffer likewise, as I'm working all day Saturday. I might get something done Sunday if my free copy of Win 7 hasn't turned up by then, but if it has, I'll probably be found trying to get that installed. (Although having said that, when I broke Windows XP two weeks ago, I was suddenly found programming because I do that from Linux)

As much as I can blog about my technical situation, I have a private blog for that, so lets report on the Province state.

Well, I've spent most of today working on the building system. Finding land to build on was harder than anticipated, but does at least work, as does building new buildings. There is also a refresh button for those days that you want to refresh and firefox asks "Do you wish to resend the last data?" (To which the correct answer is normally "no"). A dedicated refresh button that does nothing else makes this a little quicker.

I realise now that a working building system doesn't sound like much, but combined with the working resource system is starting to make this into a playable (if non interactive) game. It's also around 500-700 lines of code, which is not insubstantial. That includes the few bug fixes required in existing stuff too.

There's one critical function required to allow play, once that's coded, I may release the project as a beta for testing by a few people for errors. However, unless I'm enthusiastic this evening, it'll likely be two more weeks before that's occurs.

So, next stages (Useful self reminder):
Code resource sale
Code click use buildings (I'd forgotten about them)
Recode registration, check it still works,
Get a [limited] public beta up and running

Comment here if you wish to have an account for the limited public beta. Include your forum username. It may still be as far as a month off, but it's a useful test.

- Sun