[SciPy-User] Central File Exchange for SciPy
Sun Oct 31 02:23:56 CDT 2010
On 2010-10-31, at 3:13 AM, william ratcliff wrote:
> Would a first version without the gist feature be useful? We could try to add it in later once we see if the service is actually being used. One question about gist though---what is the overall benefit? If the original poster can edit the code is that sufficient?
Basically it outsources the version-tracking burden, in addition to making it possible to use version control tools locally if you want to.
> Also, how does the use of gist effect comments? For example, suppose someone posts some code and people make comments about the original code and those are incorporated--aren't the comments then out of sync with the code? Are the comments then synced with a given revision of the code?
The Github API definitely provides either a SHA-1 hash or a last-modified date, and so recording that when a comment is made would make it easy to flag a comment as being "about a previous version". We'd have to do this bookkeeping for locally stored snippets, too, probably just with a timestamp.
More information about the SciPy-User