From what I understood from the original post is that when you compile the lists it will sort out any dupes and such. So it should be good. Thanks for the share.
In the RB window, do you see that it's taking screenshots for "unknown" spells when a telegraphed attack is being cast?
Has it taken screenshots of unsupported spells? What does it say in the RB window when you see a telegraphed attack being cast?
possible bug? EDIT: Can't even recreat it myself after downloading a fresh copy. Don't worry about it. Not sure if I found a bug. When you press compile and you don't have any other xml files inside the spellshapes dir, your database will end up empty. This is the file you get when doing the above. Code: <?xml version="1.0"?> <ArrayOfXSpellShape xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema" />
My spellshapes file that I have been adding to for a LONG time has vanished from the directory. I could cry.
It gets merged with the default one when you put it in the spellshape directory and recompile it; all the stuff you added should now be inside the XDatabase.xml file. Double check that they are there, then make a copy of it and put it in a different directory to prevent it from disappearing.
Okay seems like I can reproduce the error. Test case Code: <XSpellShape> <Id>119</Id> [B]<Flags>NotTelegraphed</Flags>[/B] </XSpellShape> When changing the flag "NotTelegraphed" to an invalid value ("NotTelegraphe" for instance) when the SpellShapes dir only contains the XDatabase.xml file, the database is lost. The same applies when an XSpellShape <Id> tag contains no value. Note ShapeCompiler.cs -> line 30 seems rather dangerous. Please implement a backup routine for the database merging process before removing / overwriting it
Yep, the whole xdatabase.xml file vanished. Odd that i've been using and adding and recompiling this forever and it's the first time it happened.
There is a database file I add spells to as I come across them in my SVN. If you'd like, paste it into your "Spell Shapes" folder (you will have more than 1 database file, but it will be merged in the next step). Once you have your old and new .xml files in that folder, open the plugin options in RebornBuddy and click the "Compile" button to merge them. I believe my database file is up to 283 spells.
Skill #628 (100-tonze swing) cannot be effectively avoided by this plugin. Screenshots of it are useless because the attack choreography is invisible, so it wouldn't help posting them. But the attack acts just like a very large/wide circle AOE attack, so just setting it up as such in the database would work EXCEPT this attack has cast time of about 4 seconds. So the bot will run back far from range, stand there for 1 second, and run back in... but since the cast time is about 4 seconds, your characters ends up walking back into the AOE. And it is a big deal because it's a 1-shot kill, it's impossible to survive. You get hit and you die, that's it. I think the easiest way to handle this attack is to set up a 4 second delay for this specific attack. So: Log shows 100-tonze swing being casted You run out of range You stay out of range for 4 seconds, maybe 5 to be safe. Then you walk back in as normal No idea how to implement that myself though, but just throwing this info out there.
I will add it just in case, but as for the specific delay for this spell, it would likely need to be hard-coded into RB and that would require a Zekken update/replacement if simply adding it to the database isn't sufficient. Added to my SVN.
I don't remember having any trouble avoiding 100-tonze swing (I added it as a circle AoE with the largest range), although I very rarely enable that fate since it is heavily dependent on many other participants, and I can't set a separate fate boss percentage for this and then a much lower one for all other boss fates which I can solo or duo. The way I think Zekken works is that it scans the mob's casting bar and ability name, so that it stays outside of the user-set database shape/range while it is detecting that spell/ability being cast. Once the ability is no longer being cast, either due to completion or interruption, it allows the combat routine to take back over and move back into range. As long as 100-tonze swing has a cast bar, and the spell is cast as soon as the bar completes (no delayed effect), it should be able to avoid it. Since my lag/latency is generally very low, I do not know how latency may effect the spells and how they interact with Zekken. If you are still having issues with 100-tonze swing, I'd like to see how it looks in your database and a snippet of the portion of your log when Zekken detects it and moves out of range then back in range.
Percentage settings wouldn't make much of a difference. I've had him 1-shot me at 20% as well as 99%, literally right before he died lol. I think a better solution for 1-shotting boss fates that depend on having alot of people, is the bot should intelligently detect that you did a % of damage to the boss, and when you die/respawn, simply block that fate for the next 15-20 minutes. Because after you have done a certain amount of damage, you'll get most if not all of the XP. So... You go to boss fate You dps him for 5-10% of his HP He 1-shots you Bot recognizes you already contributed enough damage to get most of the xp bot blocks that specific fate for 15 minutes, then unblocks it You still get the majority of the XP once other people finish the fate for you It would have to be some kind of toggle that you can enable for specific fates, because obviously it's not necessary on boss fates you can solo. I don't think latency would be an issue for me, on speedtest.net I have 9ms and 40mb/s down. I am on wifi though, through ethernet I would have even faster ping/mbs, but 9ms and 40mb/s should be more than enough for an MMO.
Latency is specific to each server, and Speedtest cannot determine that. You need to ping the correct FFXIV server to get the ping snapshot at that moment in time, which is still not necessarily accurate for later on. I ping the servers to test my latency when I am debugging things like Ultima rotations, so thats's how I have a decent idea of my relative ping at various times of the day.
Aren't they all in California? For US servers anyway. Edit: Apparently they're in Montreal, I got the IP info for my server on arrstatus.com C:\Users\Jeff>ping 199.91.189.36 Pinging 199.91.189.36 with 32 bytes of data: Reply from 199.91.189.36: bytes=32 time=75ms TTL=244 Reply from 199.91.189.36: bytes=32 time=75ms TTL=244 Reply from 199.91.189.36: bytes=32 time=80ms TTL=244 Reply from 199.91.189.36: bytes=32 time=92ms TTL=244 Ping statistics for 199.91.189.36: Packets: Sent = 4, Received = 4, Lost = 0 (0% loss), Approximate round trip times in milli-seconds: Minimum = 75ms, Maximum = 92ms, Average = 80ms Not too bad still. And my bot's running around grinding in-game right now so those results are probably inflated.
