May 2026 Version Update

I just noticed SE updated the format of the PoL website... are they slowly thinking about making a comeback with FFXI?

Maybe if we make it to 30 years they will give us Vana'diel: A Realm Reborn? lol

remember that time they deleted Limbus for two months so they could test in prod

Part of me wonders if they released an expansion as a "new game"* if it would be successful.

*basically port everything so the experience as a player is largely the same, but fix all the underlying stuff (directX, networking stuff, get rid of POL). Give people to option to transfer their characters (although maybe start some fresh worlds with no transfers)

Graphic models themselves could be updated (or not, or just slowly replaced with HQ versions overtime) just update some UI/QOL things (which would probably break windower/addons, but many could probably be integrated into the game) just having snappier / easier to use/navigate menus would be nice. Perhaps inventory could be designed so its not so fragmented (and loads faster!)

Doing a ARR style remake would be prohibitively expensive. Doing any sort of remaster would as well. I swear you people have no idea how expensive it is to develop games, especially one this large.

During this process they would almost certainly follow the model of every other fantasy MMO, not letting you change large amounts of gear during combat. It would fail because like 80% of FFXI is making and optimizing gear sets.

Have you ever worked on legacy code before, especially in the days before standardized NoSQL databases? Back when everything was statically defined and you explicitly stated which type you expected every time you did something with it?

You change the field value in the database, then you have to go through every piece of code that touches that value and make sure it's expecting the correct type. Otherwise really weird and unexpected things can start happening.

Does anyone remember the year 2000 issue and the sheer amount of money companies poured into development to ensure the two digit rollover didn't blow stuff up?

Yeah, there was that brief rumored Nexon mobile reboot like a decade ago, and I was cautiously hoping they were just gonna farm it out - give them all the necessary assets, like the animations and navmeshes and such, and be like "okay pay a bunch of people to implement these areas, you can just mess up the gameplay entirely, go wild" - and we could at least get a Fun Vana'diel Theme Park out of it.

But given that that withered on the vine, a proper remake is laughably far out of the question.

You're correct that its not simple but it being expensive does not preclude it from being successful or profitable. (i'm not saying it would be, nor that they should do it, and I'm certainly not holding my breath that it would happen.


That would defeat the purpose.



2000 was 99% hysteria by people who does not understand how computers keep time (which sadly includes some programmers). Of course that doesn't mean strange things can't happen due to ignorance or some assumption someone made that held for longer than expected. I can't wait to not log into my bank account in Feb. of 2038 because some hashing algorithm is using a 32bit number based of epoch time fails to generate a unique one!

33k exp per mob with RoV done and Jubilee Ring during xp campaign. Bonkers. Vana'clock and Kupofried if anyone wants to get technical, no ring.

So you mean like FFXIV... yet another FF MMO for the FFXI players to hate lmao.

This is a misunderstanding of the 2K problem.

It was not that the system could not keep time, obviously it would. It was that beginning of day, end of day, beginning of month, end of month and end of year financial reports that generate very specific calculations and data would crash the global economic system that relies on mainframes, even right now.

In the finance world account ledger transactions were stored in databases as MMDDYY, believe it or not but business ledgers are the reason the US uses MM/DD/YY for dates. At the end of every business day, there is a process called "End of Day" that closes out that days ledger and calculates the transactions along with any accrued interest. Once this happens zero transactions can be done, the books are closed for the day. The next morning another process called "Beginning of Day" happens which generates the ledger for that day and opens the books for business with correct starting values. If the previous nights EoD failed or produced incorrect values, then you can not start doing business until any defects have been remediated and the process rerun. No business, meaning no loans, no payments, no transactions, nothing until those books are fixed. That is serious money.

Similar but more complicated process happens on the last business day of the month, that months ledger is closed out and final reports are generated, sometimes also quarterly reports if its ending a fiscal quarter. The the following business day a new months book is opened with the correct values. Then at the end of the year another massive process that closes out the entire year, generates reports and the next business day opens the new fiscal year.

The Year 2K work was really just going through all the scripts and application code that manages those process's and ensuring they would still correctly calculate ledger values and interest accruals with correct dates. If something was found to be broke, and lots were, they were corrected well ahead of time.

I can say this with absolute 100% complete certainty because I'm now working in the finance industry at a major non-bank financial institution. A years ago we finished a massive project that replatformed our entire loan portfolio because the previous software company went defunct and there was nobody left to support it on the off chance those interest values didn't add up. Executives were not wiling to gamble the future of a 40bn company on the "hope" that nothing broke, so they spent over 50m USD on reverse engineering and migrating to a new platform. We had to reverse engineer it because nobody was around who originally wrote our 90~20's era code to manage all BoD/EoD and monthly stuff. It's all good now but wow, that project was crazy.

People who don't know how to program yet assume it's so easy to do this make so frustrated.

AKA north of 90% of posts on ffxiah.

I'd done some digging recently because I was wondering where the heck they were storing some entity DATs that didn't follow that, and it seems they snuck in some "dynamic" entity DATs (e.g. Maze Mongers/Meeble Burrows/Odyssey/Sortie/a couple of Reisenjima story things) in an entirely different range from those you'd listed, which was neat to find.

[67911 ~ 67919] seem linked with Moblin Maze Mongers
[67920 ~ 67929] seem linked with Meeble Burrows
[67930 ~ 67937] seem linked with Odyssey (2 each for A/B/C/Gaol)
[67938 ~ 67940] seem linked with some Reisenjima story things
[67941 ~ 67943] seem linked with Sortie

Those are the dats loaded when your subzone is set. The subzone is passed at offset 0x9E of incoming instruction 0x00A when you zone into an instance using it. The subzone entity dats are loaded with the calculation subzoneid+66911 (the first known subzone is 1000.. leaving your listed range). Subzone message ids are subzoneid+67511. I think there might be another set of dats in between relating to subzone, because I have a note that there are 300 subzone slots.. but I don't recall the source.

Ohhh, so that's how they resolve it, fascinating.

I noticed there were several runs of DAT ids in use around there that started at ###11, (67911, 68211, 68511, 68811, 69111), and if those all are used for subzones, that'd track with a limit of 300 slots.

Based on that it looks like...?
subzoneid + 66911 = entities
subzoneid + 67211 = JP messages
subzoneid + 67511 = EN messages
subzoneid + 67811 = GER messages
subzoneid + 67911 = FR messages

Veering off topic but it was more than just that. I was working for IBM back then (who probably sold most of those mainframes) they were auditing everything. People who wrote a script to copy stuff from location A to B (nothing to do with time at all) had to sign off stuff. I knew an IBM exec who was paranoid (worried about 9/9/1999 too). Most of the senior programmers were eye-rolling this stuff, but most of those programs printed the date or calculated the runtime if they did anything with date time at all.

There's plenty of bad code out there to give me the Heebie-jeebies, there were some legit concerns for y2k, but also a lot of panic and overreaction.

The scare was about what I posted, global financial system crashing due to major banks being unable to do business due to EoD/BoD not managing ledger properly. Also it had nothing to do with the computer calculating time, but rather values stored in databases and how calculations handled when those values didn't follow a predicted pattern.

This wasn't some application crash or other low impact stuff, this was ancient mainframe scripts / code that calculates daily interest accrual for trillions of USD worth of assets. In 2000 the global banking system was handling 50~60 trillion dollars in "assets", which is financial speak for loans or derivatives (products backed by loans). Having that seized up because some major bank couldn't close out their end of day would be very very bad.

Just trying to point out the source of the panic. A company like IBM folding up and going out of business is small potatoes compared to the global banking system. There was genuine danger that interest accrual calculations would break causing the global financial system to lock up. Banks spent a ridiculous amount of money ensuring that wouldn't happen.