[synth recipes] invalid data from automatic scrape, disabled recipes from lack of data

[TeoTwawki]

• [x] I have paid attention to this example and will edit again if need be to not break the formatting, or I will be ignored
• [x] I have searched existing issues to see if the issue has already been opened
• [x] I have checked the commit log to see if the issue has been resolved since my server was last updated
• [x] I have read the Contributing Guide
• [x] I have specified what branch this happens on branch: All

Additional Information (Steps to reproduce/Expected behavior) : So back in topaz, an automated scrape+ compare of known available data sources was performed and synth_recipes.sql was remade from scratch (with some column order changes too). I believe this attempt at automation came about due to multiple community members back channel complaints and non-reports of the projects synth system being "all kinds of wrong" without any contribution towards correcting it in the way of data, pr, or even issue reports and a hope that by not having any "wrong" recipes it would encourage people to add the missing ones and we could verify the addition was correct and not guessed at.

In my opinion that hasn't worked out. Server owners that actually updated just re-enabled/filled in with their own guess work where they needed to, and little to no retail verification of anything has been performed and pull requested back.

Since then, many previously craftable items cannot be synthed. In many cases we just need the required skill leveled figured out and filled in. But in some others the recipe itself is obviously wrong due to being outright impossible and the column count of the row is wrong. Take the Nagan, which has a duplicate recipe containing a diff skill level and a gold ingot. This recipe uses 9 materials. There are only 8 material slots in game, no recipe EVER uses more than 8 slots and accordingly we only have 8 material input columns in our database structure. This is just one example. not all of the invalid rows are duplicated. The scrape made a lot of mistakes.

Additionally, we did not (and cannot) auto scrape the JP wiki, which can fill in many of the unknowns if someone can read it (I've tried with translation but while I can read a given page well enough, I have a hard time FINDING the page I need in the first place).

We need more hands on this. I am asking YOU dear reader, to consider pitching in if you are able, by checking in game on retail or by comparing our recipes table to the JP wiki and sharing corrections. If you aren't comfortable making a pull request you cans hare right here in this issue thread.

[Skerxan]

Thanks for the explanation Teo. This is such a huge thing for servers and everyone does it differently. I understand and support the sentiment of rather having a slightly wrong recipe than not being able to craft certain items.

I really hope we have enough people who can use the JP wiki. I can't as well sadly. Maybe it would also help if you could copy/paste all the needed items in here? If that's not too much work. Maybe that would more easily incentivize people helping out.

[TeoTwawki]

Maybe it would also help if you could copy/paste all the needed items in here? If that's not too much work. Maybe that would more easily incentivize people helping out.

@Skerxan you can see them all commented out here: https://github.com/LandSandBoat/server/blob/base/sql/synth_recipes.sql

It is a big dang file though.

Anything beginning with 2 dashes is commented out and doesn't get imported to the database.

-- INSERT INTO `synth_recipes` VALUES (4007,0,0,82,0,255,0,0,0,0,0,4099,4241,710,725,728,799,799,2383,0,0,328,328,328,328,1,1,1,1,'Black Hobby Bo');
-- INSERT INTO `synth_recipes` VALUES (4008,0,0,82,0,255,0,0,0,0,0,4099,4241,710,725,728,799,799,2379,0,0,326,326,326,326,1,1,1,1,'Yellow Hobby Bo');
-- INSERT INTO `synth_recipes` VALUES (4009,0,0,82,0,255,0,0,0,0,0,4099,4241,710,725,728,799,799,2382,0,0,330,330,330,330,1,1,1,1,'Green Hobby Bo');
-- INSERT INTO `synth_recipes` VALUES (4010,0,0,82,0,255,0,0,0,0,0,4099,4241,710,725,728,799,799,2381,0,0,329,329,329,329,1,1,1,1,'Blue Hobby Bo');
-- INSERT INTO `synth_recipes` VALUES (4011,0,0,82,0,255,0,0,0,0,0,4099,4241,710,725,728,799,799,2380,0,0,327,327,327,327,1,1,1,1,'Red Hobby Bo');

[RAIST5150]

Only craftables that stick out in my mind from when leveling were the gilded furniture pieces and mohbwa cloth (could make thread, not cloth... so no sashes).

Other than that, the bigger annoyance was lack of desynths. Were a lot of items you could make, but the desynth was not available.

Wish I had kept a list of them. Many of them show up in sparks shop, so the NQ is still readily available from a different source, which can make them less attractive for skill-up targets. Could be a nice QoL improvement on two fronts--for leveling crafts and maybe getting some more +1's on the AH.

Gotta rip the dash out of my car this weekend... d@mn LED's blanking out on the odometer. If I don't waste my Saturday wrangling with that crap... will see if I can make any sense of things enough to at least get a list of sorts started.

[TeoTwawki]

I'll see about making a spreadsheet...its gonna be large

there even 75 cap gear that can't be crafted on stock atm

[TeoTwawki]

broken synth_recipes.csv

There is your list. Excel/libre office comparable comma separated. I took the time to control+V for everyone. But you must understand corrections still have to go back to the sql file anyways.

As you scroll down the list you can see several recipes have to many synth materials, and many others have a skill requirement of 255. These are obviously invalid and the file looks wrong only because the recipe was wrong - hence this reaching out to the community for retail tests and JP wiki readers.

[RAIST5150]

Sweet!

Just tested the Samsung browser's built in translator on a few sites... seems pretty decent so far. The pictures on the anime sites were missed of course, but was pretty easy to work through most the lost in translation elements. Was able to search for some items on one random Japanese wiki with it too.

Got URLs for preferred Asian sources?

[TeoTwawki]

http://wiki.ffo.jp/ is the only one I have checked on - I have no idea what other reliable non English reliable resources there are

[RAIST5150]

Kewl... can actually use their search box there.

Looks like they are linking english names too... so you get both the funky translation as well as the english name too:

cashmilla fabric(Cashmere Cloth)

cloth materialOne. 12stackIt's possible

So this may be very doable... just a matter of the time to look up things up manually.

Here is how the recipe got presented:

［sewing：master（Recipes skills：95）］ soil crystal kashimiya yarn×3

NQ：cashmilla fabric×1個　　 HQによる変化はなし

So it isn't a perfect copy/pasta kinda thing... but I think we can make it work.

[TeoTwawki]

Bonus brownie points to anyone that corrects errors on the English wiki's.

I took care of fletchings..

[RAIST5150]

So far I may have found about a dozen items we may be able to update... managed to round up some info from wiki.ffo.jp on the first ~30 items in that CSV, down to Angon.

That is where I ran into a problem... cannot find the Angon recipe at all. All references I spotted for Angon seem to point to the ability. Tried to go through their guides/discussions of DRG... still can't find a proper alternative word to find that item in the crafting section. Even the link for them in Hagakoff's vendor list that translates to english as "Angon" was just a pointer back to the ability.

Should be able to verify/compare a handful of things with other sources as well as the ID's in our database to mark things up and post details tomorrow night.

EDIT: Grr... apparently, was an issue with the translator. Angon was translating as Nakaniwa in kanji/katakana. When I used the proper characters (アンゴン) for Angon in the search box... it found the recipe. Amusing thing is that google translates those characters to "Anngon" in english. So if this checks out, we may have a working recipe to craft them

[RAIST5150]

OK... was having a little trouble getting to sleep... so went ahead to put some proposals up for review. Copy/pasting from a an Excel and plain text file I was "doodling" in along the way.

I am going ahead with the SQL markup here and now, simply to do it while I have it all organized and all. I haven't tried forking/submitting a PR yet (never comfortable altering/submitting code without a means to test it)... but since this is basically a simple edit I will give it a go if you guys are okay with me doing that (provided you agree with the changes).

Unless someone just wants to do the whole copy/pasta thing... :)

Angon is a match between ffxiclopdia and wiki.ffo.jp, though BG still has a verify flag for smithing on it. The change would be for Smithing at 21:

-- INSERT INTO `synth_recipes` VALUES (3042,0,0,69,255,0,0,0,0,0,0,4098,4240,651,651,713,713,713,820,0,0,18259,18259,18259,18259,33,66,99,99,'Angon');

INSERT INTO `synth_recipes` VALUES (3042,0,0,69,21,0,0,0,0,0,0,4098,4240,651,651,713,713,713,820,0,0,18259,18259,18259,18259,33,66,99,99,'Angon');

Amiga and Amigo Cactus are also in agreement between FFXI and JP, but BG lists earth crystals vs. water, and a verification on alchemy. So it would be crystal element swaps and add the 16 alchemy to it.

-- INSERT INTO `synth_recipes` VALUES (2513,0,0,54,0,0,0,0,0,255,0,4099,4241,769,1774,1817,2563,17868,0,0,0,364,364,364,364,1,1,1,1,'Amiga Cactus');

-- INSERT INTO `synth_recipes` VALUES (2515,0,0,54,0,0,0,0,0,255,0,4099,4241,771,1774,1817,2563,17868,0,0,0,363,363,363,363,1,1,1,1,'Amigo Cactus');

INSERT INTO `synth_recipes` VALUES (2515,0,0,54,0,0,0,0,0,16,0,4101,4243,771,1774,1817,2563,17868,0,0,0,363,363,363,363,1,1,1,1,'Amigo Cactus');

INSERT INTO `synth_recipes` VALUES (2513,0,0,54,0,0,0,0,0,16,0,4101,4243,769,1774,1817,2563,17868,0,0,0,364,364,364,364,1,1,1,1,'Amiga Cactus');

Puk Fletchings were in agreement across all three... but the data was off in our SQL, one needs 82 CL, the other had fields out of whack:

-- INSERT INTO `synth_recipes` VALUES (34017,0,2011,0,0,0,82,0,0,0,0,4098,4240,2145,2145,2145,2145,2145,2145,2148,2148,2148,2148,2148,2148,2291,2291,2291,2291,18,24,30,36,'Puk Fletchings');

-- INSERT INTO `synth_recipes` VALUES (34058,0,0,0,0,0,255,0,0,0,0,4098,4240,2148,2148,0,0,0,0,0,0,2291,2291,2291,2291,6,8,10,12,'Puk Fletchings');

INSERT INTO `synth_recipes` VALUES (34017,0,2011,0,0,0,82,0,0,0,0,4098,4240,2148,2148,2148,2148,2148,2148,2145,0,2291,2291,2291,2291,18,24,30,36,'Puk Fletchings');

INSERT INTO `synth_recipes` VALUES (34058,0,0,0,0,0,82,0,0,0,0,4098,4240,2148,2148,0,0,0,0,0,0,2291,2291,2291,2291,6,8,10,12,'Puk Fletchings');

This was tied into lookups on some elemental arrows in this CSV file, which all three wikis are in agreement on as well. Basically just need to be uncommented.

-- INSERT INTO `synth_recipes` VALUES (2523,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2299,0,0,0,0,0,18699,18699,18699,18699,33,66,99,99,'Earth Arrow');

-- INSERT INTO `synth_recipes` VALUES (2526,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2297,0,0,0,0,0,18698,18698,18698,18698,33,66,99,99,'Water Arrow');
-- INSERT INTO `synth_recipes` VALUES (2527,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2298,0,0,0,0,0,18700,18700,18700,18700,33,66,99,99,'Wind Arrow');

INSERT INTO `synth_recipes` VALUES (2523,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2299,0,0,0,0,0,18699,18699,18699,18699,33,66,99,99,'Earth Arrow');

INSERT INTO `synth_recipes` VALUES (2526,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2297,0,0,0,0,0,18698,18698,18698,18698,33,66,99,99,'Water Arrow');
INSERT INTO `synth_recipes` VALUES (2527,0,0,57,0,0,0,0,0,0,0,4099,4241,728,2291,2298,0,0,0,0,0,18700,18700,18700,18700,33,66,99,99,'Wind Arrow');

There is a bank of round tables that aren't quite in synch across sources, but it may be reasonable to add them with help from the JP wiki data. JP Wiki has them as 67WW, 46 CL, 7 ALCH... FFXI matches except for WW ranging 64-68, and BG only agrees on 67WW but has verify flags on the subs. So, I guess the question is whether to use 67 or 68 WW, with the 46CL and 7 ALCH. This markup is with 67 WW, 46 CL, 7 ALCH.

-- INSERT INTO `synth_recipes` VALUES (3034,0,0,67,0,0,255,0,0,0,0,4099,4241,706,826,826,828,828,2476,2533,2712,410,410,410,410,1,1,1,1,'Red Round Table');
-- INSERT INTO `synth_recipes` VALUES (3035,0,0,67,0,0,255,0,0,0,0,4099,4241,706,826,826,828,828,2476,2533,2706,411,411,411,411,1,1,1,1,'Blue Round Table');
-- INSERT INTO `synth_recipes` VALUES (3036,0,0,67,0,0,255,0,0,0,0,4099,4241,706,826,826,828,828,2476,2533,2707,412,412,412,412,1,1,1,1,'Green Round Table');
-- INSERT INTO `synth_recipes` VALUES (3037,0,0,67,0,0,255,0,0,0,0,4099,4241,706,826,826,828,828,2476,2533,2708,413,413,413,413,1,1,1,1,'Yellow Round Table');
-- INSERT INTO `synth_recipes` VALUES (3038,0,0,67,0,0,255,0,0,0,0,4099,4241,706,826,826,828,828,2476,2533,2709,414,414,414,414,1,1,1,1,'White Round Table');

INSERT INTO `synth_recipes` VALUES (3034,0,0,67,0,0,46,0,0,7,0,4099,4241,706,826,826,828,828,2476,2533,2712,410,410,410,410,1,1,1,1,'Red Round Table');
INSERT INTO `synth_recipes` VALUES (3035,0,0,67,0,0,46,0,0,7,0,4099,4241,706,826,826,828,828,2476,2533,2706,411,411,411,411,1,1,1,1,'Blue Round Table');
INSERT INTO `synth_recipes` VALUES (3036,0,0,67,0,0,46,0,0,7,0,4099,4241,706,826,826,828,828,2476,2533,2707,412,412,412,412,1,1,1,1,'Green Round Table');
INSERT INTO `synth_recipes` VALUES (3037,0,0,67,0,0,46,0,0,7,0,4099,4241,706,826,826,828,828,2476,2533,2708,413,413,413,413,1,1,1,1,'Yellow Round Table');
INSERT INTO `synth_recipes` VALUES (3038,0,0,67,0,0,46,0,0,7,0,4099,4241,706,826,826,828,828,2476,2533,2709,414,414,414,414,1,1,1,1,'White Round Table');

And now to sleep...

Oh wait... gonna upload the excel sheet I imported the CSV too. It's from Office2000, so should be handled in any open source that reads xls. Has notes and date for the ones I looked up today, and highlighted things I checked against and proposed changes for... maybe help us track what still needs to be looked into. Broken_synths.xls

[RAIST5150]

Starting to see a bit of a pattern emerge when cross checking crafting levels...guessing it would be good to establish the guidelines for updatingand enabling recipes?

For an example, Heavy Crossbow is showing mostly similar numbers:

JP: 70 WD, 32 SM, 19 BO XI: 70 WD, 37 SM, 19 BO BG: 70 WD 37SM. ?? BO

Take the higher numbers presented, or the more popular ones for each craft? Give more weight to the JP research?

Or better to leave the synth inactive until things come in to better alignment across sources?

In this case, applying either rule of thumb would result in submitting 70/37/19 if enabling it. BUT, if the smithing were reversed where it was 32, 37, 32 being reported, should we go with the 32, or still submit the higher level of 37?

Or should we commit to at least two aligning completely, or at least one aligning completely with the JP research?

Just want to make sure everyone is on the same page on this front. Making sure we are following the same guidelines can streamline the review process, not to mention the goal of sticking as close to retail as possible.

[Era-Lusiphur]

Take the higher numbers presented, or the more popular ones for each craft? Give more weight to the JP research?

When I did my pass on a batch of recipes, I went with the most popular. If 2/4 sources agreed (and the others were missing info or different), I felt that was good enough. If someone wants to contest something, they can prove it themselves.

[TeoTwawki]

I believe when the automation as done there was a priority / hierarchy of trustworthiness of sources. This backfired though, as it turned out BG had bad data in lots of places.

For the most part things that were commented out we had zero data for, or we only knew "ffxiclopedia is wrong according to multiple people complaining" and not much else.

Corroborating sources is great if those sources didn't just copy off each other.

If we're going to go back to "maybe" correct (which I am ok with since the plan enacted in topaz didn't work out and was given plenty of time) then I think we should put a comment like -- unconfirmed <skill here> level

[RAIST5150]

Sounds good... so in that example, could edit Heavy Xbow from 255 BO to 19 BO (already had 70 WD 37 SM), and append a comment that one source had 32 SM, and another ??BO.

[TeoTwawki]

yeah

the complainers about how our "wrong recipes are a joke" cam open their own correcting pr or shut the F up

[FuzzyWereBeast]

For Altana's Repast, 2 of the 3 sites match a range of (111-113). If I understand correctly we would go with the highest number, 113, and notate -- unconfirmed <skill here> level comment. Correct?

... and where can we find the KeyItem list?

[TeoTwawki]

... and where can we find the KeyItem list?

scripts/globals/keyitems.lua has every keyitem including the craft key items.

As for the skill level I only addressed trusty-ness of source material. Use best judgement. Comment can be whatever as long as you are indicating what is wrong/uncertain.

[RAIST5150]

I am poking around from the top of the file, Fuzzy started working from the bottom. with 800+ records, pretty sure we can avoid overlap... but here is my XLS with notes so far for what I looked at and tried submitting tonight. Broken_synths.xls

[FuzzyWereBeast]

Looking for opinions. the Futhark is a 110-113 Alchemy synth that needs cloth and smithing. The sub levels are unknown. Should I leave this commented out or should I make the subs at level 50 or 60 so the items can at least be made? Thoughts?

[TeoTwawki]

Looking for opinions. the Futhark is a 110-113 Alchemy synth that needs cloth and smithing. The sub levels are unknown. Should I leave this commented out or should I make the subs at level 50 or 60 so the items can at least be made? Thoughts?

if we have no data at all to work with, leave it out. if there's nothing at all on bg, ffxiclopedia, jp wiki, or ffxiah, then leave commented out till someone at least gets a rough estimate from retail.

[Skerxan]

You guys are doing gods work.

[RAIST5150]

think I found some mentions of some SU recipes in some forums... but it is melting my phone trying to load them. seems to be trending around 70 for some sub crafts... provided it is the SU series...

[TeoTwawki]

@RAIST5150 @FuzzyWereBeast I just wanna thank you both for diving into this and taking big bites out of it already.

[RAIST5150]

Got to dig a little deeper this time. Some really old, lower level stuff that should be easily verifiable. Had to go into wayback machine mode to find people talking about some of them.

tried to keep an eye on the PR's related to this and printed up a PDF that kinda tracks what I know has been submitted so far if anyone else wants to pick a chunk to investigate. Or, if preferred, I can post the XLS again. Just chose the PDF since it is lighter weight and all that. Broken_synths.pdf

[RAIST5150]

Once the recent PR's get cleaned up a bit (we had some snafus to address), will have some more stuff to put up soon. Currently in my queue: Eisenplatte set, throwing tomahawks, ingot recipe for mythril bolts, handful of assorted other mythril based items and a few related desynths. PDF of what I've tracked in the pipeline so far: Broken_synths.pdf

[RAIST5150]

Got some pieces that either cite somethimg like 27 with footnote or other form of verify flags noting something like <27, <37, etc.

Main craft is not flagged, and is accepted across sources.

The footnotes and talk pages mention entry points for being able to try the synth, so have SOME idea of where to work from for a starting value... but seeing conflicting notes about that limit. BG says you need to be within 27 to try, Xiclopedia says 15.

Absent of any talk about an HQ animation or when skill-ups stopped, debating whether to alig those numbers, just not sure which rule of thumb to use, or just hold off.

Ie: being able to try with 0 sub skill, use 15 or 27 if no additional info available--or to just leave it disabled?

Any thoughts?

[TeoTwawki]

but seeing conflicting notes about that limit. BG says you need to be within 27 to try, Xiclopedia says 15.

possibly related to SE changing the mange range maybe...I recall SE making some adjustments but its not fresh in my mind

[FuzzyWereBeast]

While going through the list, I noticed that most +1 items do not have a desynth recipe. An exception to this is the Scorpion Gauntlets. There is a recipe there for the +1, but not for the NQ. Should I change this item to the NQ to fit in with the others or add it?

[RAIST5150]

Would hazard a guess that +1 desynth entry is an error. Don't think I was ever able desynth a "+1" item.

Someone may have used the wrong item ID on the desynth entry.

[TeoTwawki]

if HQ can even be desynthed in retail I'd expect it to be identical to the NQ item desynth. In my entire FFXI time I have never thought to attempt desynthing an HQ item ever.

[FuzzyWereBeast]

I'll change it to the NQ then.

[RAIST5150]

committing the cleanup/corrections first in their own PR... currently have just over 40 lines to add. Wanted to make sure this got in first before moving forward with new stuff.

Here is an updated PDF of what I have tracked so far, and what I have pending: Broken_synths.pdf

grrr... just noticed some of the columns are getting dropped in the PDF. Need to scale it down more so it all fits properly on the next one I make. Too much data. lol.

[RAIST5150]

another PR in the queue, #951 ~40 items this time. Skipped around a bit in the list, as I saw a chance to finish a couple armor sets and put up a block of Darksteel items as well. Found a few related desynths as well, not just the synth recipes.

updated PDF showing what I know to be submitted for review so far Broken_synths.pdf :

[RAIST5150]

973 submitted for review.

Some may wish to look at this one closely. A lot of comments.

As noted on the PR... had to dig up really old conversations to make some judgement calls on some of these. If desired upon review, will edit/disable as advised.

updated PDF where I have tracked progress so far: Broken_synths.pdf

[TeoTwawki]

If we're very confident in one of the sources, go ahead and uncomment the row while citing said source. But I think if we're not confident in our data source (obscure forum comments, all sources flagging as needing verification) we should probably leave the row commented out still. We can still update the data from saying ridiculous things like 255, and cite the source we used for the number but we acknowledge we aren't sure by leaving it default off and leave the decision with server operator while still actively seeking new data from retail to at least be sure we're close.

Mainly just concerned nobody will ever update them again or it looks like we made stuff up when a source isn't easily findable,

[RAIST5150]

I have been using JPWiki as first source, as it has tended to have more complete levels cited. Then look at others for alignment. But I have started running into more cases where different/higher values were more common. It is nice when others line up, makes it easier to make a determination.

That is where it has gotten sticky. A LOT of discussion appears to have gotten wiped when sites redesigned a while back. Running into more and more that slip out of alignment. Would only find one additional source in agreement, or where multiple ones were in partial agreement. Have had to skip a lot because there was no discussion about when skillups/HQ's/attempts did/didn't happen.

[mrseatbelt]

Per conversation on Demiurge Arhat's Jinpachi needs to be added to the list of recipes that currently do not work.

[RAIST5150]

Per conversation on Demiurge Arhat's Jinpachi needs to be added to the list of recipes that currently do not work.

Already in there... recipe 34001. Just no one has gotten to it yet.

[RAIST5150]

Samsung appears to have gotten their in-browser translator working again, so was able to look up a handful of things again. Still running into a good bit of incomplete data trying to verify with additional sources, so it is gonna be slow going for a while.

Attaching an updated PDF of the progress I have been able to track on this list so far:

Broken_synths.pdf

EDIT: Oops.... didn't set the print area right in the XLS file... chopped off the left-hand columns. Explains why the file was smaller this time. At least it does have the names of the recipes and the PR progress columns though. Will fix it for the next time I push some changes up.

[Kayanesa]

[Desynth] Recipes Added/Updated Multiple ++#1370 I'm trying to link it here but guess that's not the way.

[RAIST5150]

Oh my... Kayanesa's desynths PR is pushing us close to the 3/4 mark on review of this list!

Looks like we may have been able to enable a little over half of them so far!

Good work guys/gals!

updated PDF attached to show roughly where we are:

Broken_synths.pdf

[RAIST5150]

Managed to find a couple more sources lining up, but it is getting a bit thin. The portion of the list I am looking at is hitting the 95+ main smithing range and has also drifted into some ilevel 100+ items. So seeing a lot of speculation on the crafting levels.

Thankfully, Kayanesa found a nice chunk at the other end of the list. So sombined, there is still a good 50+ items that got reviewed again this round.

Broken_synths.pdf

[WinterSolstice8]

I have 110 smithing on retail, all subs aren't very high. Is there any way I can be of use here?

[Kayanesa]

There is definitely a way for you to help. As long as this info is still valid you can find a smiting recipe whit missing information on sub craft and This is taken from a Bgwiki discussion.

Attempt the synthesis while remaining engaged. You will receive one of two messages: One will say you cannot do that action now. The other will say you do not meet the level requirements.

If you receive the former, you are within the 27 required levels to attempt the synthesis. Your level + 27 = sub craft cap. You will not attempt the synth and you will not lose the items. If you receive the other message that your too low then the synth require a higher sub level.

So depending on how many recipes your ready to help fill in the info on you would make a list of the synth. Then try them with the above method and if you get the "do not meet the level requirements" you would know your sub is too low. And once you got a list big enough for you, you level your sub once and try again until you get the message "cannot do that action now". Once you get that message you will know the sub level cap for that synth (27 higher than your current one).

[Era-Lusiphur]

To help increase exposure to these in the greater XI community, I've created a new category on BGWiki to index recipes with either unverified or contested craft levels. I've started adding some and including initial research from community sites in the talk pages of said items.

If anyone wants to help with adding more items to this index, just edit the item's page and add [[Category:Unverified Recipe Caps]] to the bottom of the item's page. This simple bit tags the item page to include it in the index below, calling more attention to the item needing verification.

https://www.bg-wiki.com/ffxi/Category:Unverified_Recipe_Caps

[lawkill]

I have 110 smithing on retail, all subs aren't very high. Is there any way I can be of use here?

If you could try and verify the following numbers for this synth : https://wiki.ffo.jp/html/8148.html

That would be good, or anyone else if at all possible. I don't have access to retail right now and for while due to financial difficulties.

I would like to be able to have these on the era server I play on.

Thank you.