Dropping commercial sites is a good thing. Every time I search for something lately, I get tons of sites selling the thing, rather than sites with information on the thing. Then I have to use negations to get rid of those: -buy -price -order etc.
What google did was right, if it resulted in what you're describing. If I need sites that sell, then I can use terms to get them like order sell etc.
On another note, google's problems are not only with the algorithm, but with not being diligent about practices that it has been condeming for long. I still hit pages with top rankings that are using hidden text keywords - or artificiallly generated texts with repeat words.
As I said in another post - it would be a good thing if google had set up a program where users can be rewarded for reporting problem sites - maybe by paying into a paypal acccount or something like that.
Its about time that google did something about this. There has been lots of irrelevant pages appearing in top listing for many keywords, when you check the pages, you see that even old tricks such as hidden text is being used. Hell, yesterday I hit a page that said: this page is intended for search engine indexing, not for human viewing - and with all the crap text on it - and what am I supposed to do with that page, because humans are the ones following the links to such a pages after benig idnexed by the search engine - namely google.
I think google should do more, and one good idea I think: is to set up a rewards program, instead of just asking people to report bad behaving sites, they can pay people a small fee for reporting such sites, and I think that would be effective in creating a comunity effort for keeping the search space clean from such crap.
It would also be nice if they can cancel or down rank listing sites,
these are mainly sites created by affiliate suckers with
hundreds of pointers and zero content.
Its very annoying when searching for some topic to have heaps
of these listing sites popup in the search results.
In my experience, these sites are more annoying and create
more interference than blog entries.
I assume that because of their cross linking they get higher
rankings, in many cases higher than the direct search targets.
I mean for example, when searching for a brand name, you might
get some of these listing sites higher than the site of the brand
owner.
It would be nice if google people can find algorithms that can
identify and penalize such sites - for example, pages with a
great number of pointers and little content, or pages that have a
big number of keywords from a collection of reference directories
such as brand names or manufacturers etc. or maybe making a distinction
between 'clean' pointer listings and those with bogus affiliate id
URL's can offer a reasonable solution.
Regardless of the protocols being used, feedback on OSS programs should
be made more convenient to the user.
In many cases I have a comment or feedback on some OSS package or
tool I'm using, I write a note about it and just forget about it,
or sometimes I try to dig for the target URL/email for feedback,
eg by looking in man pages, or in sourceforge or freshmeat etc.
You don't want also to annoy the author because many don't like
sending bugs/comments to their personal or other wrong emails.
So the problem is getting the feedback to the author or to others
interested in it without wasting the user's time/effort.
What we need is a feature feedback mechanism that makes it easy
and convenient for people to get the feedback to the author or just
make it available to the rest of the world.
Here's an idea: if there was some kind of super blog or
repository, just like sourceforge is fore code, and every open
source project would have an entry in that repository, eg
ossfeedback.org/wget, then each GUI software can have a feedback
menu option that points a browser page to the software's feedback
page at the repository, just like there is an 'about' button,
there can be a 'feedback' button. for CLI apps it would be enough
to have the feedback page address listed with the help message,
but this should become a standard practice.
This would make it much easier to send comments or feature ideas,
another benefit of it is to free authors not interested in user
feedback from continuous user nag about feature requests,
everyone would send their comments to the repository point
because its would be the most convenient path, and for the
authors it would become a pull model rather than a push one; if
they're interested in it they can check it.
For others interested in improving a package, they would get an
idea about users feedback about features, or maybe even ideas
about new packages, without this feedback being lost in the
email/logs/backups of the original authors.
Your difficulty is not that you don't understand the GPL; it's
that you don't understand what a derivative work is. That is
defined by copyright law (statue and, mostly, precedent) and it's
not really the GPL's fault that it is unclear.
This is correct when it comes to the GPL, because nowhere does the GPL try
to offer a definition for the term 'derivative', rather
it leaves it to the law:
The "Program", below, refers to any such program or work, and a
"work based on the Program" means either the Program or any
derivative work under copyright law: that is to say, a work
containing the Program or a portion of it, either verbatim or
with modifications and/or translated into another language.
Now contrast that with the LGPL:
5 . A program that contains no derivative of any portion of the
Library, but is designed to work with the Library by being compiled or
linked with it, is called a "work that uses the Library". Such a
work, in isolation, is not a derivative work of the Library, and
therefore falls outside the scope of this License.
However, linking a "work that uses the Library" with the Library
creates an executable that is a derivative of the Library (because it
contains portions of the Library), rather than a "work that uses the
library". The executable is therefore covered by this License.
The above paragraph is actually defining the term 'derivative'
as it applies to software libraries, rather than just providing a
license term. This raises two issues:
Is there clear case law to support this definition?
The definition of the term 'derivative' uses the term 'linking',
so what is the legal definition of the term linking?
since there is no legal definition, then we use the English meaning,
but since we are in a technical context we use the technical meaning.
The question becomes: is there a clear technical definition for the term
'linking'? From the raging debate over the technicalities of linking,
there does not seem to be such a legally sound definition of this term.
Slashdot Mirror
Dropping commercial sites is a good thing. Every time I search for something lately, I get tons of sites selling the thing, rather than sites with information on the thing. Then I have to use negations to get rid of those: -buy -price -order etc.
What google did was right, if it resulted in what you're describing. If I need sites that sell, then I can use terms to get them like order sell etc.
On another note, google's problems are not only with the algorithm, but with not being diligent about practices that it has been condeming for long. I still hit pages with top rankings that are using hidden text keywords - or artificiallly generated texts with repeat words.
As I said in another post - it would be a good thing if google had set up a program where users can be rewarded for reporting problem sites - maybe by paying into a paypal acccount or something like that.
Its about time that google did something about this. There has been lots of irrelevant pages appearing in top listing for many keywords, when you check the pages, you see that even old tricks such as hidden text is being used. Hell, yesterday I hit a page that said: this page is intended for search engine indexing, not for human viewing - and with all the crap text on it - and what am I supposed to do with that page, because humans are the ones following the links to such a pages after
benig idnexed by the search engine - namely google.
I think google should do more, and one good idea I think: is to set up a rewards program, instead of just asking people to report bad behaving sites, they can pay people a small fee for reporting such sites, and I think that would be effective in creating a comunity effort for keeping the search space clean from such crap.
It would also be nice if they can cancel or down rank listing sites, these are mainly sites created by affiliate suckers with hundreds of pointers and zero content.
Its very annoying when searching for some topic to have heaps of these listing sites popup in the search results. In my experience, these sites are more annoying and create more interference than blog entries.
I assume that because of their cross linking they get higher rankings, in many cases higher than the direct search targets. I mean for example, when searching for a brand name, you might get some of these listing sites higher than the site of the brand owner.
It would be nice if google people can find algorithms that can identify and penalize such sites - for example, pages with a great number of pointers and little content, or pages that have a big number of keywords from a collection of reference directories such as brand names or manufacturers etc. or maybe making a distinction between 'clean' pointer listings and those with bogus affiliate id URL's can offer a reasonable solution.
Regardless of the protocols being used, feedback on OSS programs should be made more convenient to the user.
In many cases I have a comment or feedback on some OSS package or tool I'm using, I write a note about it and just forget about it, or sometimes I try to dig for the target URL/email for feedback, eg by looking in man pages, or in sourceforge or freshmeat etc. You don't want also to annoy the author because many don't like sending bugs/comments to their personal or other wrong emails.
So the problem is getting the feedback to the author or to others interested in it without wasting the user's time/effort. What we need is a feature feedback mechanism that makes it easy and convenient for people to get the feedback to the author or just make it available to the rest of the world.
Here's an idea: if there was some kind of super blog or repository, just like sourceforge is fore code, and every open source project would have an entry in that repository, eg ossfeedback.org/wget, then each GUI software can have a feedback menu option that points a browser page to the software's feedback page at the repository, just like there is an 'about' button, there can be a 'feedback' button. for CLI apps it would be enough to have the feedback page address listed with the help message, but this should become a standard practice.
This would make it much easier to send comments or feature ideas, another benefit of it is to free authors not interested in user feedback from continuous user nag about feature requests, everyone would send their comments to the repository point because its would be the most convenient path, and for the authors it would become a pull model rather than a push one; if they're interested in it they can check it.
For others interested in improving a package, they would get an idea about users feedback about features, or maybe even ideas about new packages, without this feedback being lost in the email/logs/backups of the original authors.
This is correct when it comes to the GPL, because nowhere does the GPL try to offer a definition for the term 'derivative', rather it leaves it to the law:
Now contrast that with the LGPL:
The above paragraph is actually defining the term 'derivative' as it applies to software libraries, rather than just providing a license term. This raises two issues: