Open dlevenstein opened 5 years ago
see also #86
How will this proceed? I don’t know the first thing about Peter’s system. Will Peter take the lead?
I don't either, or have a sense really of what will be involved in the integration.
To my understanding from the conversation yesterday, the goal is to have a sessionInfo file that maintains the old structure (for backwards compadibility), but also includes the metadata from Peters system. So what is needed is a way that when bz_getSessionInfo is called the first time, it populates the sessionInfo file with the metadata from peter's system, or prompts the user to enter it.
The issue, which I am sure @petersenpeter will show you, is that you want much more information than what is currently stored in sessioninfo. Peter defined a standard that mates with his database (bidirectionally, you can populate it by entering information online). What we discussed yesterday is maintaining the sloppy organization of the current sessioninfo (xml scraper) and then adding his fields as optional with redundant information to preserve backwards comptability. After looking at his organization, and how many times sessioninfo is called in buzcode (<30) I reiterate my original suggestion - abandon the current organization of sessio info and adopt Peter's whole hog. it is conceptually much better, it has more information and can more easily expand without breaking the internal logic, and it has expanded capability (through its readbility with the web database).
That all sounds great. One thing that was nice, was the ability to easily use legacy datasets that only had the xml (when calling bz_getSessionInfo, a sessionInfo file with the info from the xml was automatically created). I just request that this ability will still be there.
Thanks @samamckenzie and of course @petersenpeter it sounds great.
The big question is who will lead this who actually knows Peters stuff? Someone can't start it without knowing his stuff. Is @petersenpeter the only one who can do this? Do others know his system? If we don't find a person to lead this it won't get off the ground.
I will be involved and likely leading this.
Den tir. 29. okt. 2019 kl. 13.20 skrev Brendon Watson < notifications@github.com>:
Thanks @samamckenzie https://github.com/samamckenzie and of course @petersenpeter https://github.com/petersenpeter it sounds great.
The big question is who will lead this who actually knows Peters stuff? Someone can't start it without knowing his stuff. Is @petersenpeter https://github.com/petersenpeter the only one who can do this? Do others know his system? If we don't find a person to lead this it won't get off the ground.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/348?email_source=notifications&email_token=ADLPZOPX2O7V5DGR5FPELD3QRBWGHA5CNFSM4JF47U62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECRLX2A#issuecomment-547535848, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADLPZOJVYFNHVLHBX2MFJKTQRBWGHANCNFSM4JF47U6Q .
Oh well that's perfect then!
On Tue, Oct 29, 2019 at 1:23 PM Peter Petersen notifications@github.com wrote:
I will be involved and likely leading this.
Den tir. 29. okt. 2019 kl. 13.20 skrev Brendon Watson < notifications@github.com>:
Thanks @samamckenzie https://github.com/samamckenzie and of course @petersenpeter https://github.com/petersenpeter it sounds great.
The big question is who will lead this who actually knows Peters stuff? Someone can't start it without knowing his stuff. Is @petersenpeter https://github.com/petersenpeter the only one who can do this? Do others know his system? If we don't find a person to lead this it won't get off the ground.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub < https://github.com/buzsakilab/buzcode/issues/348?email_source=notifications&email_token=ADLPZOPX2O7V5DGR5FPELD3QRBWGHA5CNFSM4JF47U62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECRLX2A#issuecomment-547535848 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ADLPZOJVYFNHVLHBX2MFJKTQRBWGHANCNFSM4JF47U6Q
.
— You are receiving this because you were assigned. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/348?email_source=notifications&email_token=AA26WTN6FNHNSRU2FTDYKITQRBWPRA5CNFSM4JF47U62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECRMARA#issuecomment-547536964, or unsubscribe https://github.com/notifications/unsubscribe-auth/AA26WTJITAY54QH2QIHSPMLQRBWPRANCNFSM4JF47U6Q .
I think the best thing would be for Peter to update bz_getSessionInfo so it
1) Loads his metadata file 2) If the file doesn't exist: -prompts the user to input the necessary info -Pulls info from the xml into the file if it hasn't been done already -saves the file 3) Has option for "noPrompts" in which it just loads from the xml if no file exists
That way, we have one path to all the metadata, (per #360), which makes the file from all the proper information if it hasn't been made yet.
Does this seem reasonable/feasible?
This seems reasonable to me. I have something like this implemented, but not entirely. Let me get back to you with a solution once implemented
Den ons. 30. okt. 2019 kl. 11.01 skrev Dan Levenstein < notifications@github.com>:
I think the best thing would be for Peter to update bz_getSessionInfo so it
- Loads his metadata file
- If the file doesn't exist: -prompts the user to input the necessary info -Pulls info from the xml into the file if it hasn't been done already -saves the file
- Has option for "noPrompts" in which it just loads from the xml if no file exists
That way, we have one path to all the metadata, (per #360 https://github.com/buzsakilab/buzcode/issues/360), which makes the file from all the proper information if it hasn't been made yet.
Does this seem reasonable/feasible?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/348?email_source=notifications&email_token=ADLPZOOMKQPDCYPBMFZLNW3QRGOTZA5CNFSM4JF47U62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECUQYAY#issuecomment-547949571, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADLPZOJZCUP3Z2HHSAC6UJLQRGOTZANCNFSM4JF47U6Q .
If a sessionInfo file exist, it should also translate any information not existing in the xml into the new format, so we need to discuss the current format.
Den ons. 30. okt. 2019 kl. 11.25 skrev Peter Petersen < petersen.peter@gmail.com>:
This seems reasonable to me. I have something like this implemented, but not entirely. Let me get back to you with a solution once implemented
Den ons. 30. okt. 2019 kl. 11.01 skrev Dan Levenstein < notifications@github.com>:
I think the best thing would be for Peter to update bz_getSessionInfo so it
- Loads his metadata file
- If the file doesn't exist: -prompts the user to input the necessary info -Pulls info from the xml into the file if it hasn't been done already -saves the file
- Has option for "noPrompts" in which it just loads from the xml if no file exists
That way, we have one path to all the metadata, (per #360 https://github.com/buzsakilab/buzcode/issues/360), which makes the file from all the proper information if it hasn't been made yet.
Does this seem reasonable/feasible?
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/348?email_source=notifications&email_token=ADLPZOOMKQPDCYPBMFZLNW3QRGOTZA5CNFSM4JF47U62YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOECUQYAY#issuecomment-547949571, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADLPZOJZCUP3Z2HHSAC6UJLQRGOTZANCNFSM4JF47U6Q .
extended bz_getSessionInfo that incorporates Peters metadatasystem for sessionInfo.mat creation and loading