[Greasemonkey] Greasemonkey: an Historical Perspective

Matthew Gertner matthew at allpeers.com
Tue Apr 19 14:38:27 EDT 2005


Julien,

My assumption is that these problems will arise at some point in the future.
Greasemonkey simply hasn't been around long enough for the targeted pages to
have had time to change much. But I guess a wait-and-see attitude is
reasonable. If we see that a) scripts are starting to break, b) people are
starting to block scripts and/or c) the whole thing is causing bad buzz for
Firefox, then hopefully this discussion will be revived so that we can take
prompt action.

Regarding the use of versions for webpages rather than specific scripts:
This is actually what I put in the essay initially. I think I changed it
because a change to a webpage might break one script but not another. But as
Mark pointed out, all of these solutions are somewhat fragile. I still
prefer the version-per-script approach because it could lead to a nice sense
of cooperation between content providers and scripters. At the end of the
day, psychological factors may turn out to be more important than technical
considerations.

Cheers,
Matt

> -----Original Message-----
> From: greasemonkey-bounces at mozdev.org [mailto:greasemonkey-
> bounces at mozdev.org] On Behalf Of Julien Couvreur
> Sent: Monday, April 18, 2005 8:46 PM
> To: greasemonkey at mozdev.org
> Subject: Re: [Greasemonkey] Greasemonkey: an Historical Perspective
> 
> We may want to split this thread into multiple threads (even though
> they are related):
> - advantages of structured/semantic data and what can we/GM do to help
> get there (if at all possible)
> - general fragility of GM scripts, means to defeat them, website
> owners fighting user scripts
> - letting site owners "hint" which user scripts work/break on the
> current version of the website
> - security of GM scripts and implications on Firefox
> 
> Regarding Matt's main suggestion: letting site owners "hint" which
> user script version work or break:
> There are very few sites that seem to have reacted to GM yet. I'd
> recommend we wait to get a better feeling of how GM is perceived.
> 
> Have you actually run into a broken script because of a change in the
> website?
> It seems that a better fix would be to have website owners use a
> special meta-tag to indicate the version of their website. GM scripts
> could then say "known to work on website Y version X".
> 
> It's true that website owners can look up scripts in the current and
> future GM directories. What if we actually had a GM header that would
> inform the website of the scripts that will be used on the url being
> loaded?
> We could always remove it later, either for all sites or specific
> sites, if some sites actually start blocking users based on that
> header.
> 
> 
> Aaron says:
> > Authors adding rich semantic meaning to markup is awesome in a void.
> > Who doesn't like the idea of  rich metadata to build applications on
> > top of? The problem is that once these semantics have an economic
> > incentive tied to them (read: search result position) they will
> > instantly become abused, and so, nearly worthless.Author-supplied
> > semantics might never happen on  a commercial web.
> 
> Never thought of it that way. I wasn't even going that far in the
> reasoning: there doesn't appear to be many consumers of that metadata
> (ex: Google probably doesn't care about rel=bookmark), so authors
> don't have any incentive to put any.
> I would still hope that if the metadata consumer/author loop was to
> successfully bootstrap, abuse of this metadata would not completely
> kill the system (ex: there is google bombing, but linking and Google
> are still useful).
> 
> 
> Aaron says:
> > At the same time, I also wish there were more robust form of
> > customization. I think the bigger barrier to that is politics and
> > laws, not technology. If all GM does is get people thinking about
> > their content being repurposed and allow a few hackers to fix some
> > annoying bugs, I'm OK with that.
> 
> In regards to raising awarness about web remixing, GreasemonkIE would
> definitely have been a good thing. Anyone looking into picking that
> project up?
> 
> Cheers,
> Julien
> _______________________________________________
> Greasemonkey mailing list
> Greasemonkey at mozdev.org
> http://mozdev.org/mailman/listinfo/greasemonkey
> 
> 





More information about the Greasemonkey mailing list