Keep folding history logs

Case number:845813-991188
Topic:Game: Tools
Opened by:ptfrog
Opened on:Wednesday, November 30, 2011 - 14:43
Last modified:Sunday, January 8, 2012 - 11:32

Foldit logs the output of each script run, and perhaps other items as well. This information is saved in the Foldit directory. But the information is overwritten with each new script -- forcing careful folders to create their own tracking logs, often using simple tools such as notepad.

I would like an option to keep the Foldit logs available on my system, so that I can parse and review them. It would greatly facilitate our ability to figure out what worked and what did not over the history of a puzzle. The simplest way to do this, I suppose, would be to simply rename old logs instead of overwriting them -- appending .0001, .0002, etc., or perhaps simply adding a timestamp.


(Wed, 11/30/2011 - 14:43  |  4 comments)

Joined: 09/22/2011
Groups: None

I suspect the reason why logs are deleted/overwritten, is because if they weren't, they would eventually take up a considerable amount of disk space.

I'm not sure how the existing logging system works, but if this change was to go ahead, there should be another option somewhere (not necessarily in the game itself, maybe the "options.txt" file), that allows people to choose how long the logs are kept for, or how much space the logs can use, before they are deleted.

e.g. "Log Keeping" = 30 (30 days, before they are deleted - only in the case of multiple log files)
"Log Keeping" = *200 (total number of MB a single log file can take, OR the total number of log files allowed)

ptfrog's picture
User offline. Last seen 5 years 34 weeks ago. Offline
Joined: 09/29/2011

I agree. Though this feature would only be useful to a certain class of users -- those that know enough to do something with the logs. Even if it is only to upload the logs to a group log repository, that process could also involve deleting them. So a warning that "don't do this if you don't understand the consequences" might be acceptable.

Still, you're right. That's kind of sloppy programming -- they really should have the auto-delete if they are keeping logs. Even without the keeping of logs, Foldit can accumulate 100's of megs over a year of heavy usage. I've often thought about reviewing those files and setting up some kind of archive system for old puzzle info.

Tlaloc's picture
User offline. Last seen 14 weeks 2 days ago. Offline
Joined: 08/04/2008
Groups: Mojo Risin'

tristan has the exact reason I implemented the script logging system the way it is. I figured if you were really using the logs, someone would write a separate program that did an auto archive of the log file. Since all log files end with a consistent XML tag, the program could search for that and when it sees it, move or copy the log file. This program would have to be different depending on whether you were on Windows, Mac, or Linux.

I wanted to get the log file written, because there is much useful information that can be extracted. But I didn't want it to be overly complicated or take up too much disk space, and I didn't have much time to write it. So having one log file per script run was the best compromise.

ptfrog's picture
User offline. Last seen 5 years 34 weeks ago. Offline
Joined: 09/29/2011

Eminently reasonable. I had assumed that the log file was written for uploading back to home base, so that statistics could be kept. I appreciate the fact that you took the trouble to write the log file for us. As you point out, it allows programmers (as well as assiduous and knowledgeable non-programmers) a way to access and archive this info. At least this is true for now; I'm not sure what will happen when the ability to chain scripts is introduced.

In any case, you can't blame a guy for wishing it was easier for "regular" folks to doc+. :-) I suppose I could write something in Java, and offer it out to those who trust me. For the record: what would be really cool is an easy way to allow the log file to be delivered to the original scripter in order to help him or her analyze what each part of a script really does "in the wild."

A facility like this would certainly be open for abuse. Still, it's on my *think about how to do it* list. Among friends -- such as many in my group -- it could be a very useful tool.

Thanks for the response, tlaloc.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons