it doesn't just "guest" there is some sort of system to it. I'm assuming it takes the time stamp and compares that to the normal real life seconds in between chat log messages
^ this is called 'guessing'. I'm really serious here; there is no fool-proof way to determine double attacks in real time; and even determining it after the fact requires information that a parser won't have, for instance, your exact haste%. (mostly because it's impossible to get this number directly from the client).
the chat log is also subject to lag and client-server trust times, and is not, in fact, timestamped by the game, it's only sequenced. so we don't have an accurate way to conclusively determine the intended time between two messages, just the time between when we actually -saw- those two messages.
there's other reasons why it's guessing, but I don't want to waste space with excess reasons, the above is sufficient.
edit; I'd assume you're using the other parser (not directparse) since last I checked, dp didn't have a feature like this; so I can't actually answer your original question other than to say: these numbers are always going to be suspect. if you want a -real- answer on the DA rate you need to either use the slowest weapon imaginable to make your tests (999delay) or do it the right way; which involves fraps and lots of math.
It is indeed guessing, but hopefully in an 'educated' way. It considers max (normal) attacks per round, average attack delay, killshots, a couple indicators for whether TA is a factor at all, and a few other things, and provides enough detail that you can manually confirm its results, and make certain corrections if needed.
It doesn't need to be tested with huge delay weapons, but a lag-free environment and minimal haste really helps. Figure the game log can lag the specific time of the event by 1-2 seconds even under good circumstances and you'll understand why using a weapon with a delay (including haste) in the 2-3 second range can get tricky. If you'd used the 228 weapon without the haste gear I expect it would have provided pretty accurate results.
Further refinements to the code are still needed, though, which I'll get around to at some point.
Anyone have updated memlocs?
I wasn't aware that they had changed? Just parsed a party last night...
Or updated relative to when? The memloc for the September update is on my google page, and somewhere in this thread.
Tried to get kparser to work, to no avail. I downloaded the lnk from this page and upon opening the .exe file i get this error message.
Unhandles exception has occurred in your application. If you click continue, this application will ignore this error and attempt to continue. If you click quit the app will close.
Exception has been thrown by teh target of an invocation.
I dunno what to do, and seeing and i am as cpu retarded as they come all teh help i can get would be nice lol :D
also i am on vist64x if that could mean/help at all
Did you make the recommended Vista adjustments as indicated in the readme.txt file?
Did anyone try to import DirectParse chat dump to KParser? the output on my PC was only 300KB file data.
Oops. Forgot to update KParser when they changed the database structure. Update made, patch release 1.1.6.2.
Now it stuck on "Saving..." state after reading the lines for more than 3 minutes (this is just <4MB file). It still takes 25% of my CPU load. But it doesn't show up anything when I changed tab during this time. And it completely freeze when my cursor try to click "File". Should I give you the sample file? This one is pretty old. It was from the time when DirectParse still named DvsParse.
Unfortunately, reparsing (and import is a type of reparsing) has a bottleneck that really slows it down. It parses the raw data at about 1000 lines per second (a fast merit party with no filters can generate 12k+ lines per hour), and saves the completed database to disk in very little time (about 4 seconds for said 1 hour merit party), but the stage in between where it inserts the parsed data into the data set in memory is relatively slow (the 1 hr merit party could spend over a full minute in this stage) since it has to make tons of (sometimes very complicated) lookups to make sure the different table data is all hooked together properly.
As such it is not unusual that it would seem to be 'stuck' on "Saving..." for a while, and will max 1 CPU for the duration (I assume you have a quad core if you were stuck at 25%). I would expect a 4MB KParse database to be reparsed within 1 minute. DirectParse doesn't save as much data per line, though, so would have a smaller footprint for the same number of lines, and thus take longer for a similarly-sized database file (not sure how much, exactly).
However it's also possible that there's a flaw in the import for a file that old (I designed the import to handle the database generated after the switch of the name to DirectParse). I would need to look at the file itself to be sure, so yes, if you could send it to me that would be quite helpful. If the PM system won't handle it, you can upload it as an attachment to a bug on the google page, or PM me and I can send you my email address.
It also shouldn't lock up when you click the File menu while the import is running, unless I screwed something up when interacting with the UI thread. I'll have to check on that.
Not sure if this is a recent implementation or not, but I used to have the "Unparsed Data" tab accessible, where it would show information like
So-and-so earns a merit! (# Points)
or
Your Parrying skill goes up .2 points.
All of my recent parses in 1.1.5.0 have stopped having this miscellaneous show up in it.. I don't think I've changed any settings to my knowledge. Do you have any idea what's up?
Edit, just downloaded 1.1.6.2 and this option shows up again; clicked on the Unparsed Data section and then it disappeared. Any clue what's going on?
Are you saying that the data in the tab is disappearing, or the option to open the tab is disappearing? If the latter, do you have the Debug Mode option turned on?
It was the disappearing tab that was haunting me.
Figured it out however,
Debug mode needs to be checked in and you have to select Menu > Window > Unparsed Data in order for it to show.
Thanks Motenten.
Sorry for the bump, but I've been trying to get a parser to work on vista, neither Kparser or Directparser will work.
For Kparser my problem is the same as Shinsei's
For Direct parse, it seems to work, It turns on, I select Start new parse, then when I actually start to fight, nothing is parsed.
This is the debug log
29/11/2008 11:42:09 PM - DvsChatmon_3: Monitor thread started.
29/11/2008 11:42:09 PM - DvsChatmon_3: Attempting to connect to Final Fantasy.
29/11/2008 11:42:09 PM - DvsChatmon_3: Module FFXiMain.dll Base Address = 39F0000
29/11/2008 11:42:09 PM - DvsChatmon_3: Success attaching to Final Fantasy XI process!
29/11/2008 11:42:09 PM - Attempting to read chatlog meta info... (here)
29/11/2008 11:42:09 PM - DvsChatmon_3: Dereferencing chat pointer.
29/11/2008 11:42:09 PM - Error dereferencing first pointer.
29/11/2008 11:44:44 PM - DvsChatmon_3: Dereferencing chat pointer.
29/11/2008 11:44:44 PM - Error dereferencing first pointer. (to here)
The spots I marked just repeated every second between the 2 numbers I posted, didn't feel the need to post it all.
Hit compatibility mode for xp, and ran as admin.
I think I need the latest .net framework... I keep getting errors opening k.parse and now d.parse wont even work for me :/
Version 1.1.7 release
1.1.7
+ Patch: Addition of Salvage Cell listing option in Loot tab. Credit: Valyana of Ramuh
* Patch: Show window title when selecting multiple POL processes. Credit: Valyana of Ramuh
* Redid display of update progress while reparsing or importing files.
- Fix risk of process reentry deadlocks in plugins.
* New base address for RAM reading (12/8/2008 patch): 0x0057A2C8
So if we're trying to update DirectParse, we'd use "0x0057A2C8" as the new memlocs?