[Techtalk] website overhaul/broken links

[Carla Schroder]

On Saturday 20 July 2002 01:47 am, H. Peter Anvin wrote:
> Malcolm-Rannirl wrote:
> > On Friday 19 July 2002 10:08 am, Carla Schroder wrote:
> >>How practical is it, when giving a large, complex Website a complete
> >>overhaul and re-design, to find a way to not break all the links?
> >
> > It depends on how the site is designed. It doesn't have to be that
> > difficult. For the small website I admin, I just built a redirect file
> > for all the obvious ones, and then I got through the 404 log every so
> > often and add any I missed to the file.
>
> For a big, complicated site, you may want to set up the new site so that
> the namespace is disjoint (read: it's easy to tell if a URL is new or
> old.)  Then you can use a database or transformation script to generate
> the redirects for the old names.
>
> By the way, kudos for taking the time to do this.  Far too many sites
> don't give a....
>
> 	-hpa


Well, there's a story behind this. Linuxnewbie.org had a major makeover- and 
broke all the links. My bookmarks, my referral page- all hosed. Not to 
mention the many dozens of sites that link to their Newbie-ized Help Files, 
and search engines- all toast.

So I sent a rather anguished email to Sensei. I won't say who responded- he 
was nice enough, but said that it's the responsibility of external sites to 
maintain their links. I pointed out that this meant the actions of one site 
inconvenienced many sites, not to mention goofing up search engines. I said I 
thought Web sites should think like libraries, where everything belongs in a 
certain place, instead of shelving books randomly. 

Well his second reply was rather more tart, so I dropped it. But it got me to 
thinking, and wondering if there are any technical obstacles to maintaining 
links in a very large, complex site when it is completely re-done.

What ticked him off the most was I mentioned that because so many sites 
routinely re-shuffle content and break links, I rely on cached Google pages. 
Faster and more dependable. This can cost the original site money, of course, 
they don't get those clicks. Well duh, they should have thought of that 
before they broke things.

I don't intend to beat up on Linuxnewbie.org, but I do feel an article coming 
out of this experience.  :)

Thanks,
Carla