Sunday, April 28, 2013

RS 3075 Done! (and other musings)

533 pages, wowza, but I've updated all the entries.  Luckily, the second half went far faster than the first, with lots of battle armor/protomechs that are quick to update.  I also kinda slowed down on verifying some data, so there may be some updates from the most recent RS 3075 that were not in previous versions.  Sorry, but they should turn up if people actually start using the site.

Speaking of.  This kinda amazes me.  I read the official forums, and it is amazing how many people are asking for stuff that I have on my site.  Not only are people not telling others about it, but the people searching apparently aren't using google, which my site comes up pretty fast as.  Heck, google "goodsects WHM-6R", or any mech you like, and see how quick you can see the results.  It seems that nobody is spreading the word of all the work I've done, and its discouraging to think I'm only doing this for me, and a couple of my readers, and *I* don't even play it anymore.  The speed of my development is directly proportional to the number of people asking about it.  While I appreciate the few of you who do occasionally comment and visit the site, it just seems there aren't very many folks.

That being said, my "fuel" for development of my Federation and Empire assistant (aka a Cyberboard replacement) is running out, and interest in that has waned as well, so I haven't touched the code for it in I think a month now, even though it was 90% on its way to an alpha release.  Not saying I won't finish it, but again, without that much interest, I get bored and move on.

I have started playing with XNA, which I may be able to port some of my code to and it *might* speed up, but will require DirectX.  This would mostly affect my mapping program, but could also affect anything game related I do.  If I get some more interest I may write a top down, 2D, real time version of megamek that lets you drive around a persistent planet, with other folks, and fight out battles.  Who knows where that'll end up.

Ok, back on track to this site.  Now that I've finished updating my data I am going to start working on a completely NEW site.  I'll copy/paste a lot of the same stuff, but will be starting from scratch.  I'll make it a bit more modular so I can more easily make a few updates.  I've since also figured out a better method to create a dynamic easy to update database which I'll implement.  I'll also fix many of the bugs, and get record sheets for *ALL* units working (with hexes as an option, but circles default) using the same methods I made for the fan made TRO 3063.  This includes updates to all the Quick Strike (soon to be Alpha Strike I think) cards that I've made, and options to use canon data or my own calculations that are far more accurate (as computers so often are).

My current plan, of course subject to change:
Phase 1:  Recreating the search and search results pages
Phase 2:  TRO views of everything
Phase 3:  Record sheets
Phase 4:  Quick Strike style sheets
Phase 5:  Creating the user interaction stuff, like creating accounts and saving unit lists
Phase 6:  Unit validator (enter in your stuff, it makes sure years/costs/BV/availability is correct)
Phase 7:  Online unit designer. It'll be ugly, but it'll be better than *anything* else, and allow construction of *every* type of unit within the game, including things like mobile structures, bases, and even characters

These may be done at any time, depending on mood:
Phase A:  Add support for planets/systems, much like the IS Atlas
Phase B:  Redo the Cartographer app, hopefully HBHK will be out and Volt will be kind enough to shoot me all the updated data
Phase C:  Real time version of megamek, but with you only having a single unit.

6 comments:

  1. I haven't actually seen a link to your site before! I've only been reading this blog for a couple of months...

    Here's a thing that would be really handy, and that nothing offers right now: search by year and/or faction availability. (MUL;info's faction data are woefully incomplete, of course, but my goodness how useful it would be to say "I want 'Mechs generally available to Davion units in 3025".)

    ReplyDelete
    Replies
    1. Unfortunately the MUL is the only source of faction availability, and yeah, its only somewhat complete for the Jihad era, and zilch for any other era.

      I could use Xoti's tables to create values for this, but they are kinda hard-coded to particular eras.

      I thing I wanted to do WAY back, that would still work, is for each faction, for each unit, give a number between 1 and 10 as to the popularity of the unit. Then, I could go through and sum up the numbers and get a basically availability # by faction and unit.

      However, that again limits me by era and stuff. So, a better method I designed, is for each year in the universe, for each faction, for each unit type, show the # of elements that exist, by quality rating (strategic operations). That is a HUGE amount of data to put in, but I could randomize it all in seconds and then just make changes occasionally to specific units. This would actually allow me, for any year, for any faction, to tell you how many of a particular unit are in service, and of what equipment quality. All the math would be done by the computer, and assuming my numbers were right, it'd always be the best way to get random unit types. Instead of BV, a number for commonality could be used. If you want that super rare mech, it may take 5 "points", while the common one that may be more effective just 1 "point". Though this seems hard, I can just apply a base production amount for mechs, from the year they exist (with a few prototypes earlier) with a set period of time until they are no longer produced. For example, mech X starts production in 3000, in 2990 they make 1, in 2995 they make 1 per year, in 3000-3025 they make 10 per year. After this they all decline in quality by 1 each 20 years, and battles would remove a few here and there, and add a few here and there to other factions as salvage. Hard coding generic production info for all units, and static rates of quality decline, along with some random captures/destruction (based on unit survivability, like CASE would increase the chance of salvage), would allow me to quickly, dynamically, create pretty decent availability information. Maybe I'll implement this on the next version as well, though it'll be 100% non-canon, though since there is no conflicting information you can't prove it isn't canon ;)

      Delete
  2. interesting algorithm, how would it go up against available RATs? I was thinking the "availability" would be related to manufacturing capability and access via normal channels, and battlefield salvage and inter-house trading would then be alternate channels.
    So when a system with a factory world changes hands, the faction immediately loses the ability to manufacture that product (therefore being stricken off the "armory/RAT" even though the current armies still field them, making them still available via salvage (from cannibalism or by restoration) and eventually (though much more difficult) by trade agreement.
    The problem is establishing a base "armory" content via RAT because I believe the data is full of holes.

    ReplyDelete
    Replies
    1. Well I'd use the faction for the system building it first, then, when in doubt, look at the RATs.

      Here is the process I'd take:
      Step 1. Get introduction year for all units. If no data exists, determine earliest possible year and use that.

      Step 2. Determine factories that produced the unit, what faction they were in during each year, and when the factory went offline. Also, some will have data on production counts, so use that.

      Step 3. Determine some abstract number, I'm thinking about 480 mechs a year for example, of total mech production in some specific year. Each mech class has a "weight" with it, 1 for assault, 3 for heavy, 4 for medium, and 2 for light. That is the # of the total production for that faction, for that year, as a percentage. So, 480 mechs/year, 24 assault classes being produced means 2 of each class gets produced each year. There are 4 tanks per mech, 1 fighter per mech, and so forth.

      Step 4. Determine a rate of quality loss, perhaps 10 years per year of life a mech losses a level of quality.

      Step 5. Determine a rate of loss during time periods. 3025-3028 may have a high rate of loss compared to 3000-3024. Use available data, extrapolate missing. This is a percentage of total forces lost each year.

      Step 6. Determine the percentage of each unit lost, that can be salvaged, and the percentage that can be captured (both of which reduce quality). Things like CASE would help salvage chances, slow movement may help capture chances.

      Using this data, I should be able to determine how many, of what quality, of whatever kind of unit a particular faction posseses at any given year. From that I can look at the units equipment quality, weight class, and so forth and generate percentages based on those values.

      There are a lot of values I'll just have to make up, or extrapolate, which is very unfortunate. However, since TPTB for BT have stated lots of times they don't ever want to publish numbers (with a horrible reason of not allowing writers latitude), they will probably never really be wrong, just never confirmed.

      From this, I can produce a RAT in milliseconds for any race, on any front, using any unit type, for any equipment rating, in any year, and generate a whole regiment if you want.... in milliseconds.

      Delete
  3. You'd have to tweak the numbers until you don't drive high-loss designs like the Locust extinct before their time, but this seems to me like a very practical approach.

    Way more than I was looking for, but I certainly wouldn't object if you can get it working.

    ReplyDelete
    Replies
    1. Yes, and it could end up breaking. Take the 110/year production line of the VF-QA Valkyrie n New Avalon. The line opened in 2787, and was still going strong in 3025. That is 238 years of production, at 110/year. In 3025 that is 242 regiments of Valkyries, and the FS only has like 80 total. That means at least 220 and probably a lot more of those regiments have been destroyed, over 90% lost, in order to not have them be dominant in every unit.

      So there would have to be a lot of fudging some numbers in order to make it work, without contradicting canon.

      But maybe the 95% or whatever losses the Valkyrie ends up with, may be a good number to base the losses of any design created during the 1st SW that survived through the 4th SW, and with so many extinction rates of high tech stuff before 3025, perhaps 90% of all designs or so get wiped out every 100 years through attrition.

      Delete