[Greasemonkey] Goals for the user script directory (very long)
jdunck at gmail.com
Fri Apr 8 09:16:16 EDT 2005
On Apr 8, 2005 1:38 AM, Aaron Boodman <zboogs at gmail.com> wrote:
> Cool, do it!
> A few comments:
> > There'll be full text search (so you can see how people are using GM_*
> > functions or find a script through a comment)
> So will you cache a copy of the script or something? Seems messy.
How so? Size?
> >, there'll be header search
> As evan was saying before, this is hard to do perfectly and fast. But
> you should be able to hack something that gets close. That should be
> fun. Also, I still think this can be done with just delicious tags.
I understand that @include is pretty loose, causing lots of loosely
Maybe scripts should be ranked by number of non-wildcard chars matched?
Or maybe thresholds a la /. comments which would filter out loosely
As for it being hard to do fast, I guess you're referring to the idea
that a search for scripts matching some URL would have to be run past
a regex for each script. Hmm. That is a problem. I'll have to think
about it some more.
> > Also, particular script versions will be available off of stable URLs
> > on the directory.
> Why? If the author has updated, I think we should only offer that copy.
Well, this is a lot of work and space, and maybe not so useful, but
it's something I'd like to have. I'll keep this idea for version
I just looked at using archive.org as a distributed versioner, but I
don't think its reasonable because there's at least a 6 month lag
between submission and publishing. Too bad. I wonder if there's a
business model in providing on-demand web archive creation? Service
> This is what greasemonkey does to default those values:
> * If there is no namespace, the namespace is the URL the script was
> found at minus the leaf.
Thanks, that's seems sensible. I think maybe userscripts should have
a page on the site, even if they are actually hosted somewhere else.
This page would be an analog to FF's Page Info screen, showing what
the dir knows about the script. Also not version 1.0, though.
> And I think that the versions shoud just be numbers. Eg 1, 2, 3. :)
> you could store the date with them for display purposes.
> > @longdesc so that script authors could write a book (like this one!)
> > about how great their user script is, and the directory would scrape
> > this so that updated functionality just shows up in search.
> Maybe we should just link to their webpage, as we do now.
We could do that instead. ;-) I'd like a way to get to the page
about the script, though. so maybe:
> Consider using existing services instead of building from scratch
> where possible. Think about how to break it up into pieces to do
> little by little. And to get something useful right away.
I think the first release will just be form URL submission,
spider/version hash, delicious for ranking, and search by URL.
More information about the Greasemonkey