Mat Kelly

@ibnesayeed I know this per the original message of this issue. Should a WARC be created exhibiting a number of these 0-length response scenarios? I was hoping to establish a...

> zero-length responses are present already with any status code in an existing WARC. There are examples but they're amidst 100+ other records. > a non-redirect code is easier to...

Oh, that's really not good. I must have broken something along the way that is not yet being tested in the Travis config. I will look into this further, @akavel...

@akavel A bit was done with this. Please do a fresh pull and let me know if you continue to experience the above issue.

@akavel Somehow `exc_info` had turned into `exec_info` at some point. This is the reason for the vague, uninformative error message you received. I fixed this in #323 thanks to your...

@akavel I originally was using the `timeout` with `ipfsapi.Client` but ran into some issues that caused the move to SIGALARM. I'll have to verify that reverting to a timeout does...

@akavel At one time we had it as a second argument on the `cat()` function on the ipfsapi.client instance. That was the aforementioned problematic issue I referenced that led me...

Re-tested this in a Windows 10 VM with cb4b5bc3fffa033a1b463ae71847f4acc765aa1e ``` C:\Users\Mat\Downloads\ipwb>ipwb replay Executing first-run procedure on provided sample data. Processing WARC records in salam-home.warc complete IPWB replay started on http://localhost:5000...

Again, the issue seems to stem from `signal.signal(signal.SIGALRM, handler)` On Windows, running: ``` import signal signal.signal(signal.SIGALRM, handler) ``` as pulled from the [signal module docs](https://docs.python.org/2/library/signal.html) produces an error: ` AttributeError:...

An attempt was made to move to the `stopit` module and remove the `signal` dependency in https://github.com/oduwsdl/ipwb/commit/7c3076b31d90789a0a860728e866a195a5e07bdd but the timeout is never executed per the `stopit` documentation. Perhaps a threading-based...