16:23 arright! my eyes were crossed and i missed that the merkle_tip / timestamp object is not raw data: it is an object with a `msg` property that contains the raw data. BLARGH. must struggle to continue. 16:24 ! struggle. stuggle x 2. put eyes at terminal. there we go. 16:24 ok goal is what? we have timestamp.msg we need timestamp object. what produces timestamp obhject? find function that returns it. 16:25. -> make_merkle_tree function makes one. it is probably a class from imported library. plan: import manually in pthon repl. read help. also! python trick! `import inspect; print(inspect.getsource(anything))` <- outputs source code for funcs. better than help command. 16:26 . 16:26 . i'm laughing and having trouble controlling my arms and head. onward! i keep laughing instead of coding when i try to code. we are not talking about boss torture. we are just coding. no need for laughter. onward! 16:27 . -> make_merkle_tree is important from library. 16:28 -> it's at github.com/opentimestamps/python-opentimestamps (i had already cloned this when my net was going at 7kB/s earlier today) in opentimestamps/core/timestamp.py ! 16:29 it doesn't instantiate the object, just uses its methds. ummmmmm i kinda want to run it under pdb to find the object but maybe i'll trace it back another step. 16:30 16:31 -> it's made from the `timestamp` of a DetachedTimestampFile object. grep. 16:32 -> timestamp attribute is passed into the constructor and used straight . trace back another step -> found Timestamp class, still 16:32 16:33 -> .msg is a bytes -> __init__ takes only msg -> can hold attestations and ops -> opentimestamps.core.timestamp.Timestamp woohoodlywootywoo! 16:34 dang i'm laughing again. dang. honestly laughing is lots nicer than other things my brain and body can do. but right now: work. 16:34. 16:35 -> okay so now i guess i'd better review Timestamp to make sure none of the fancy attestation behaviors mutate the msg attribute 16:37 blargh i'm confused! my issues are complaining there is too much to track here at once. what if it doesn't ever mutate msg ever, and my original post was flawed? maybe it stores data that can be checked without access to the timestamp doc already! who knows! 16:37 i am really confused! uh-oh! oh btw recently i made an improvement to the GPT language model I called GPTb . it's at https://github.com/xloem/GPTb . It's 16:39 . i'll try to look at this a little bit more