[CLD] PCL submission to CL Directory

[Peter Seibel]

On Jan 5, 2006, at 2:06 PM, Marc Battyani wrote:

> "Peter Seibel" <peter at gigamonkeys.com> wrote:
>
>> [might as well use this mailing list]
>
> Yes.
>
>> Yeah, but even the permanent URLs look like line-noise. And I when
>> I'm just browsing a site, I want to be able to cut-n-paste the URL
>> out of my browser and email it to someone, not have to go searching
>> the page for a safe-to-email permanent URL.
>
> Well I can switch to a cookie based session tracking so that the  
> URL will be safe. Though
> people will have to accept cookies to be able to login.

That seems fine to me. Does anyone really worry about cookies any  
more? And if they do, there's a big difference between a basically  
sensible URL with some gunk on the end that would otherwise be in a  
cookie than in completely opaque urls. (Another problem with line- 
noise URLs is that I can't look at a URL and have any idea what it's  
a link to.)

>>> BTW tell me why this is less acceptable than a java encoded URL
>>> like this:
>>> cgi-bin/bv/home/Home.jsp?
>>> BV_SessionID=@@@@0941546823.1136493133@@@@&BV_EngineID=ccccaddgil
>>> mkggfcefeceefdffhdgli.
>>> 0&cacheID=f1netscape&3313945937=3313945937&BV_SessionID=@@@@09415468
>>> 23.1136493133@@@@&BV_EngineID=ccccaddgilmkggfcefeceefdffhdgli.0
>>
>> It's not. It's just as unacceptable. I can live with urls like that
>> when they're on a commerce web site because I *expect* that my
>> interactions with the site are unique to me. But when I'm looking at
>> a informational website, I expect the information to be organized by
>> a human being and for the URLs to give me some sense of that
>> organization.
>
> But the interactions are unique! I will soon add filters and  
> preferences for instance.

That may be true. But it seems that the bulk of the site should be  
fairly plain vanilla information that I can share with my friends.

> Anyway I can use URL like resource/by topic/text/regexp/cl-ppcre  
> though these URL will
> break when the tags tree is reorganized. Not to mention the fact  
> that you have to choose
> one of the possible paths so you can end up with every thing put  
> in /dev/libraries

It seems okay for there to be multiple URLs that point to the same  
resource. Or we can come up with an actual organization that we're  
okay to live with and then tags are just a meta-data layer on top of  
that. I think in the long run that's a better solution but it  
requires actual work.

>> That's fine but I think it needs to be reasonably obvious how to use
>> it without having to read a manual--most people aren't going to care
>> that much.
>
> Hey... it's obvious for me but I'm open to propositions... ;-)

For starters there's a bunch of stuff on each page and it's not clear  
how they differ. What, for instance, is the difference between a Note  
and a Comment? (You can tell me if you want but the real problem is  
that unless it's obvious it just confuses people.) Also, I don't  
understand what the distinction is between "sub tags" and the other  
top level headings on the Root page (e.g. "Documents/Web Sites",  
"Implementations", etc.

-Peter

-- 
Peter Seibel           * peter at gigamonkeys.com
Gigamonkeys Consulting * http://www.gigamonkeys.com/
Practical Common Lisp  * http://www.gigamonkeys.com/book/