Mobius Evalon's site

More BGC

Sat, 23 Oct 2010 02:45:40 +0000

After an unannounced hiatus and some secret, unannounced development, I find it time to update the few followers of BGC on the progress of the game.

The existing, working games (chess, checkers, and reversi) have all received upgrades and statistic tracking. This means that all of the game's listed achievements are available to get, and I've even instated a few virals that require you to beat developers or testers to receive them.

As for new additions, backgammon, go, and battleship are in the works. After a run-in with icon-size issues, every game now uses 24x24 icons which means a few boards are a bit larger than I had originally hoped:

Oh yeah, did I mention the game board floats in its own window now?

Now that I've finally deployed a decent-looking Go board, I'm looking into the game rules to make it the next playable game. After that, I'll probably brave Battleship, as its situation is unique -- having to keep track of four separate displays to work nicely.

FFFFUUUU-

Fri, 22 Oct 2010 22:39:06 +0000

So approximately a year into BGC I decide to separate the map control from the remainder of the interface and place it in its' own window when a game is being played. Here I was thinking I could take advantage of the map control's icon-size parameter and change up the tile size based on the game, but it turns out that when the map control's icon-size and world/icon_size are incongruent you get all kinds of lousy-looking results.

My plan was to give large gameboards smaller icon-sizes, allowing something like Go to use 16x16 pixel tiles which would use a modest 320x320 map space while impacting playability little, while allowing chess and checkers to use standard 32x32 for the same screen real estate.

Now I'll have to go back over my entire established display management system and configure the entire thing for 24x24 pixel icons. It's not exactly preferable to make Go occupy 480 pixels square and have chess take up 240, but it's necessary for coming additions to save the extra real estate over 32 pixels -- one such addition being battleship which occupies a 24x12 space which would be ridiculous at 32 pixel tile sizes.

Blah.

Another installment in the awesome series

Wed, 22 Sep 2010 15:31:09 +0000

So for about a decade now I've been in possession of an account on every major messaging service (ICQ, AIM, Yahoo, MSN Messenger/Windows Live Messenger) and more recently an account on Google Talk, and this quickly became cumbersome and irritating to deal with. Naturally, I went in search of a method to combine these all into a single window to regain a sense of simplicity, and in some cases, to avoid the official client altogether (I haven't liked AIM since it went Triton, nor MSN since it turned into Windows Live). It becomes impossible to manage five different client windows on screen, not to mention that until 2004 I was not in possession of a computer that could handle all at once, or some clients altogether (Windows Live Messenger for instance refuses to load on my old desktop PC, likely due to my 400+ contact list).

My first venture into multiprotocol clients was Trillian, and I was using this on an old Compaq Presario for some time. I even actually bought the premium service for a couple of years (what that netted me I can no longer recall) and it worked well enough, though admittedly the computer it ran on was garbage and usability problems were common, ranging from taking several minutes to start up to video or audio chat not working at all. It still combined all of my contact lists into one area, though, and at a time when I was still using the enigmaster2002 moniker concurrently with Mobius Evalon, it certainly bolstered a since of simplicity to bring all of these clients together with a pair of screen names each.

While Trillian gave me a taste of what these types of programs could do, the interface was internally skinned and often overcomplicated and it quickly negated any benefit it might have had. I then moved onto GAIM but it quickly became apparent that the client as it was when I tried it (many years ago) was not what I was looking for -- every contact that existed on both screennames for the same service were doubled on the contact list, and at a time where I was migrating identities it took me about fifteen minutes to abandon it and uninstall.

I continued to use Trillian for a few more years until I decided to take another peek around for clients, finding Miranda IM. This was after I started attending college (mid 2006) but before I procured a laptop (early 2010), so I carried a collection of flash drives around, one of which filled up with PortableApps including Miranda. The program was definitely simple but was rigid on its extensibility, making it impossible to perform the rare video chat or make use of other such fringe features of the individual clients. Nevertheless, I stuck with it until I got my laptop.

After I had bought my Asus laptop I searched around again for multiprotocol clients. I decided to revisit Pidgin (formerly GAIM) since at this point I had long since killed off my older monikers, and Pidgin served me pretty well for the last eight months. It has some odd issues though, such as crashing the moment my laptop went into hibernate.

I'm not entirely sure what prompted me to search for a replacement for Pidgin. I think lately I'm just in the mood for finding things that are better than what I settle for (such as my finding Lunascape), but whatever it was, I stumbled upon Digsby. Admittedly, it has a stupid name and program icon (what is that thing anyways, an egg wearing a baseball cap?) but it's certainly a case of not judging a book by its cover, as the website really seems to undersell what exactly Digsby is capable of.

Digsby comes stock with the basic feature of adding accounts from messenger services including AIM, GTalk and Facebook Chat, but its features beyond that are what make it amazing. Metacontacts are a stock feature and allow you to group contacts together, such as grouping together the same person's MSN and Yahoo accounts. You can add all of your email accounts to it and it will check for new mail at an interval specified by you, in addition to letting you check and delete mail from the Digsby client itself, regardless of it being a custom IMAP account or MSN hotmail. It allows you to keep current on Facebook and even let you like or comment on wall posts right from within the program.

Digsby not only created a competent multiprotocol client, but has gone beyond that to integrate functionality you wouldn't even expect an instant messenger client to provide.

Video game difficulties

Sat, 04 Sep 2010 23:53:36 +0000

I'm trying to finish Gears of War 2's single player campaign on insane difficulty, and it has become apparent that, just like every other game with a difficulty setting, this highest level is simply unfair.

To effectively take down enemies, you need to land headshots; to land headshots, you need to aim; to aim, you have to lean out of cover; when you lean out of cover, you die. What's the alternative? Blind firing all of your ammo away while the enemy forces unload at you with their infinite ammunition.

I just finished Grand Theft Auto 4 this week (not even twelve hours before a lightning strike literally blew up my DSL modem) and that was repeating one mode called Hangman's NOOSE 1200 times. Gears of War on Insane seems like I'm going to see the same parts of the campaign as many times while I carefully memorize enemy spawn numbers.

Who thought this was a good idea?

Thu, 05 Aug 2010 17:41:44 +0000

I'm not entirely certain at what point someone found it a good idea to create services like snipurl, bit.ly, qurl, tinyurl, and others, but it has become apparent to me that everything people use them for is malicious.

Sure, the odd long URL is somewhat annoying (sites like slashdot use them for search engine optimization) but at least you know exactly what you're getting into. Shortened URLs from the above named services completely hide the destination address and you have to be crazy to click them.

Lunascape

Wed, 04 Aug 2010 11:45:52 +0000

I'm not entirely certain how long Lunascape has been around, or why it hasn't made a bigger splash in the browser market, but I'm pretty damned glad I stumbled upon it via a blog "top ten" list.

Lunascape employs all three engines -- Trident (IE), Gecko (Firefox), and Webkit (Chrome) for those unaware -- and can swap them with the click of a button, which means I was finally able to get rid of Chrome and ignore Internet Exploder again. It's also speedy as hell. Regardless of the rendering engine you're using, Lunascape renders pages faster than Opera did on my machine, so I uninstalled that as well.

So what doesn't it do? I was expecting it to use one cache/cookie file for the entire program instead of one for each engine, but was disappointed. This means you have to, for example, login once for each of the three engines to test a website. In addition, any plugins for any one of the engines only works for that engine, as can be expected. Installing Firefox's AdBlock on Gecko won't help you on Trident or Webkit rendered pages.

Lunascape is essentially a browser built on the concept of Firefox's IETab with the Webkit engine included, but it's still a lot more convenient than having all three browsers on your machine. In addition, its memory usage is a fraction of Firefox's when using the Gecko engine, meaning Lunascape has a much more efficient implementation of Gecko while still allowing you to use all of Firefox's plugins.

You don't sacrifice anything switching from Firefox to Lunascape, as I know I found myself right at home with almost zero adjustment time.

Damn my Xbox is annoying

Mon, 26 Jul 2010 16:27:50 +0000

My Xbox suffers from what is referred to as the "open tray error", which is apparently some fault with the laser reader hardware, where it tracks too far inward and attempts to read boot information from the clear center part of the disc. What's particularly annoying about this problem is that I can do absolutely nothing about it except continuously open and close the tray until the Xbox finally gets the idea that it has a disc in there.

I've managed to (somewhat) rectify this by installing the games I play the most to the drive, but this just avoids the inevitable "this disc cannot be read" error -- I still have to open and close the tray until it figures out the disc is in there so I can boot the copy off the drive.

/icon objects versus the raw data

Sun, 31 Jan 2010 22:16:17 +0000

I ran into this problem after creating a cropping proc for user-uploaded icons, and after a bit of experimenting it finally dawned on me that the "raw" file data differs from the content of an /icon object:

//small little test project
#define INPUT "default.input1"
#define OUTPUT "default.output1"

mob
    var
        icon/display_icon = null
    verb
        set_icon_raw(i as icon)
            usr.display_icon = i
        set_icon_object(i as icon)
            usr.display_icon = icon(i)
        speak(t as text)
            world << output("<img src=\"\ref[usr.display_icon]\">[usr.name]: [t]",OUTPUT)

Obviously I'd assume this distinction is intentional, but the kind of situation as above creates problems because I can't find any way to turn the /icon object back into file data. I know it's automagically converted back to the necessary file data when assigned to atom.icon and other vars, but how can I accomplish the same when these vars aren't involved?

This is the cropping proc I was referring to:

proc
    icon_crop(a,m_x=96,m_y=96)
        if(!isnull(a))
            var
                icon/i = icon(a)
                icon_x = i.Width()
                icon_y = i.Height()
            if(icon_x > m_x || icon_y > m_y) i.Crop(1,1,min(icon_x,m_x),min(icon_y,m_y))
            return i
        return null

This function can't be accomplished without turning it into an /icon object, so I run into the problem described above.

Rechargeable batteries

Thu, 21 Jan 2010 11:14:30 +0000

Some time ago, I purchased a Turtle Beach gaming headset, and while its wirelessness is quite convenient, it has eaten standard alkaline batteries like mad since the day I got it. Today I decided to invest in some rechargeable batteries, but about 13 hours afterward I started getting curious as to the specifics of the science of batteries.

When I was in the Wal Mart, I was sold on Energizer E2 NiMH batteries because it boasted 850 mAh. "Hey, that's a lot of milliamps" I said to myself, grabbing a four-pack of 'em for $11 and a charger unit. It never really occurred to me to research what exactly that meant -- sure, more power is always good, but upon some research into the subject, I can't find a plain answer as to how the mAh rating translates into device power time. The best answer I could find is about the "depth of a cycle", and while I generally get the idea of what it's driving at, I have no idea what determines cycle depth. Does the device determine that by it's ampere requirement? Can this information be used to directly calculate the amount of power time a fully-charged 850 mAh E2 can give my headset?

Further research into the subject has yielded more facts but little plain English. Apparently NiMH batteries can effectively be used in either "shallow" or "deep" cycles but have a quite a small discharge rate. Can this adversely affect the battery or the performance of my headset?

On the subject of performance, I've read several things about different kinds of chargers. Apparently a "trickle" charger will continue to supply charging current nonstop while a "float" charger will suspend the charging operation temporarily if the cell reaches capacity. From what I understand, a trickle charger can and does damage batteries and cause failure if left attached to the charger after their capacity is met, but a float charger supposedly speeds up the adverse "memory effect" of the cell. Is it overly important that I discover the charger type that I'm attaching my NiMH batteries to, or are they generally more immune to these problematic effects? This one is important because I'm the kind of person who attaches things to the charger and they can sit there for very very long periods of time -- I have a portable phone in my room that's not seen use for years, but has remained on the charger unit for the last three years nonstop. My XBox rechargeable battery packs remain attached to the quick charger for some time (I alternate between my two controller battery packs, so one sits there for upwards of a day before seeing use again). In addition, I've had dry spells of XBox playing where I'll leave these things connected for 1-2 weeks continuously, and it's likely I'll do the same with these rechargeable AAA batteries.

Anyways, any help on the subject is appreciated. I've learned a lot of new terms and facts about batteries, but the brass tacks of how useful the batteries are in my situation and how concerned I should be about catastrophic failure or unnecessary cell decay is presently what eludes me.

Big icons as screen object overlays

Tue, 12 Jan 2010 01:51:54 +0000

Code:

//this matter has a lot of repeated code, so one of the calls to the following object's New:
if(b.privs&IUPD_HANDGUN) items += new /obj/hud/iupd/weapon_container(null,"handgun","Smith and Wesson .38 special",32,"SOUTH,WEST")
//several more of these to construct the entire display shown below
c.client.screen.Add(items)

obj/iupd/hud/weapon_container
    mouse_opacity = 1
    layer = EFFECTS_LAYER+1
    New(l,p,n,w,s)
        ..(l)
        src.name = n
        src.width = w
        src.icon = Server.resource("iupd_weapon_bg_disabled-[src.width]")
        src.overlays += Server.resource("iupd_weapon_[p]")
        src.screen_loc = s

Problem description:
When the server is initialized, all of the states contained in a pair of DMI files in my project are created as new icon()s and stuffed into a list for future use. I went this route because they have no container /objs and their only purpose is to be an overlay on a button on HUD elements, and I figured I could save some memory by loading them all up once, and using the same /icon object on every player's screen (of course, if this is a bad idea then feel free to suggest better methods). This has somehow netted me a problem, though:

But if I simply maneuver the IUPD logo beyond the map bounds:

My question is: why? What exactly happens when a screen object is displayed over a map void that'd make this display suddenly work?

I'm using the TOPDOWN_MAP format on 462.1056, and each element above is used at their native size (32x32 and 64x32, as applicable).