One problem I have with coding for RPoL, besides the general fact I have no idea what I’m doing, is that I bite off more than I can chew.  I had to abort a massive project I undertook for the site in about 2005 because it was too big.  I know what I need to do, but the end of the tunnel is so far away there’s not even the glimmer of hope of some light shining down it.  The more I work on the changes the more I realise needs to be done.

This all stems from opting not to use mySQL (or any structured database) back when I made RPoL in 2000.  I opted for Berkeley databases, some of which are so basic they are actually flat text files, each record being a line in the file.  The sad fact of all this is at the time I knew I should use SQL, but I looked at mySQL and could just not get my head around it (actually, it might have been to the fact that when I was originally learning programming they forced a lot of SQL onto us then and I hated it; I think it took over ten years to get over the trauma).  Later in about 2004 when I revisited it, it all seemed so easy and I had no idea why I didn’t use them before.  I’ve been slowly migrating the databases over to mySQL since.

Using SQL is a.. well, it’s a no-brainer really, but the unfortunate fact is that it requires a lot of work on my behalf for very little difference to the users.  Once it’s done it makes future changes much easier, and makes any data manipulations done much less error-prone (i.e. you can update an individual “cell” (if you thinking about databases like an Excel spreadsheet) rather than manipulating an entire “row” when all you want to do is chance one cell).

Everything was originally in Berkeley.  I moved the game list over to mySQL but everything else in 2005 was still in Berkeley.  The project I started then was to change every other database over to mySQL in one hit.  It was.. massive.  As I mentioned above, I eventually gave up.  Doing so much in one hit was beyond my enthusiasm levels, I do need the rush from releasing a new version to keep me going.  I get the personal satisfaction of a finished version, and the enthusiasm that I then get from users keeps me happy and keen.

I’ve been doing the changes piecemeal since.  I’ve moved the user database and stickylist to mySQL, as well as writing the new (at the time) rMail to use mySQL.  Now I’m upgrading the players (in a game) and characters (in a game) databases to mySQL.  I can’t really break it down any further, but this is a big change, I think the biggest.

I’ve got to change the routines for; how a message is posted (and edited), how the game menu (thread list) is rendered, now the private thread list is rendered, how a message is displayed (I verify that the user has access to view the thread, I don’t assume that just because they’ve somehow got hold of the URL that they have permission to view), how the cast list works, how the dice roller works, how the character profiling works, how the GM menu works.

The first and last ones on that list are the really big ones.  I’ve basically rewritten the GM script, that’s 7,300 lines (which equates to 30,787 words or 256,000 characters).  That’s in the range of a novel.  I’ve now got to battle about 2/3rds that size again with the message posting.  My enthusiasm, is waning again and I’ve started to do some very stupid things to increase my enthusiasm.

More on my stupidity later.

I do some web searches from time to time to see what the general consensus is with RPoL, and to see how it’s received compared to other sites.

Two things to spring to mind when looking at the results.  Firstly, I get the opinion that RPoL is more orientated toward the mature audience and less toward those who possibly expect every web bell and whistle to be available to them, and for everything to be automated.

The second is that I wish I had the time to monitor this stuff more closely, to offer suggestions to visit RPoL and to rebuke stuff like:

I don’t want to push a conclusion or accuse them of fudging, but we’ve always been confused by the way they handle their stats

Thanks kaethre, very kind of you to not not say that.  While it’s vaguely (very, very vaguely) satisfying to know we’re in a position where people want to cut us down, it’s not at all nice to see such accusations flung around.  That’s certainly not the first mythical accusation I’ve seen weaved from a member of that particular site, and I’m sure it won’t be the last.

As I said, it makes me wish I had more time to participate on other sites.. or maybe a small pack of rabid Rabbits Rabbids roaming the Internet.

Back in.. uh… 2000 to 2004? …  Back when I was working at the datacentre where RPoL was also hosted, I found it fantastically quick to do coding.  Whatever I wrote could almost instantly be uploaded to the beta site for testing.  The server was physically about 5 metres from me.

Since both myself and RPoL have left that place I’ve tried to setup similar test beds, but nothing has stuck.  Finally I recently installed Sun’s VirtualBox on my desktop machine, installed (err, grabbed a ready image of) CentOS 5.1 (which the RPoL server runs), installed all the various applications that RPoL uses, and gently applied a hammer to get Berkeley DB to work.  I then grabbed a copy of RPoL’s data and voila, I have my own local copy of RPoL Alpha running on my machine.  Then with the help of FileZilla doing a “Keep remote directory up to date”, anything I edit here is uploaded to the alpha site within a second.  No longer am I forced to wait for 10 or whatever seconds to upload to the dedicated server in America before I can test if a tiny change worked.

I’ve found it amazingly helpful, and really great for speeding through code changes.

It will also let me test code a lot more before I upload it to beta, which used to suffer from very alpha and unfinished code as it was where I uploaded every code change, whether it worked or not.  Beta will get updated at pretty much the same pace, it just won’t get all the intermediate test script uploads that I do that leave the site in a sad state of disarray.

The new private groups are looking good, by the way.

That’s not to say I’ve done anything exceedingly horrid, but sometimes I wonder why I said or did certain things.  Also, looking back, it’s amazing to see the things I intended to do.

I’m not deluded enough to think that everything I say I’m going to do to the letter, but it’s amazing to see how far from the intended path I have strayed.

All I can say is that this particular path to hell is.. mostly paved with good intentions.

Many a year ago I updated this thing very infrequently.  Now I feel like I’ve got more to say, though surely nobody still vists, let alone will want to read it.  But alas, intrepid reader, here’s whatever I feel like jotting down.  Infrequently as ever, I assure you.

I must say, WordPress is being very kind (to me, cruel to you) and is letting me add in backdated posts.  As fun as it was to muck around with coding the old blog, I think the days where I spend my time coding something when another, far superior, program exists are over.

Fortunately for the users of that particular site, RPoL is still far superior to anything else.

There’s so much to do for RPoL, and not just for the site that you’re all used to me referring to, but for other things as well.  People keep on asking about various RP related things that I want to do.  I want RPoL to be to RP like Google is to search engines.

Alas all my grand schemes rely on this new version being done, and although I’ve done some good work on it this weekend (and last week), there’s still so much more to do.

Today I’ve spend about 4 hours on it.  Alas I’ve really lost track of my running tally, probably about 40 hours.

He’ll be sorely missed by everyone who used to come along and tap on his bowl and exclaim “is he dead?”.

Fight long and hard, little buddy.

But enough sharing, that might imply I care.

Nothing really to report, thus the lack of posting.  I’ve been working on RPoL, but nothing very exciting, I’ve been hacking up the visual layout, which is more fun than coding, until you have to get it to work for half a dozen different browsers and for twice as many colour schemes.

I’d guess I’ve spent 16 hours on it all since I last posted.  That’s a lot of html.

This week doesn’t look too consulting-work busy, so hopefully I’ll get quite a bit of RPoL coding done (though hopefully not too much, I’d like to get paid for a bit).

Time spend on RPoL today: 1 hour.
Total time spend on RPoL: 24 hours.

Anyhoo.  Today was a day of… uh… frustration It’s currently 2am.

I got up at 9-something.  A bit late I know. ‘lo any behold, as if to punish me for sleeping in the server was down.  I wasn’t sure if it was the server or Zip, so I contacted support telling them that our server was not available (giving them the details of the server, which included the IP address and host name) as well as, to my folly, the IP address of the last device that responded with my trace route.

Support happily told me that the IP address of the last hop I quoted was responding fine.  Uh.. firstly; can you read?  Secondly, don’t you know the addresses of your own networking equipment?!

eventually they got it right and it was our server, so it was rebooted.  Seven hours of downtime. )c;

I was a little miffed with this and asked what happened to our server monitoring that they were supposed to provide?  We are supposed to be notified if the server’s down for 30 minutes.  It seems they forgot us, but now it’s all good.  So they say.

So that was the morning gone.

Then came client #1.  PC’s running like crap they reckon.  So we load a test document and wow, does that take forever, they’re right.  To cut a boring story short, after 30 or so minutes of playing around trying to figure out why it takes so long to load a document we find out it’s not “a” document, it’s “the” document.  This one document we decided to test the PC with is a piece of crap and takes up to two minutes to open.  Other, large, documents open in a flash.

Client #2 was a server migration.  I was hoping to ghost from disk to disk, but after about three hours of trying I give up, ghost just either crashes or won’t see the new and old disk.  So then I go with a backup and restore option, but by this time it’s past 9pm.  And that’s where I still am now, at 2am, just polishing off the new server, getting Windows happy with the new hardware.  I’m almost done now, almost.

And because this is going to become my running tally of RPoL development;

Time spend on RPoL today: 2 hours.
Total time spend on RPoL: 8 hours.