| 1 |
maillog: 09/06/2005-20:14:33(-0700): Duncan types |
| 2 |
.. snip .. |
| 3 |
> .... |
| 4 |
> |
| 5 |
> That said, there remains the practical issue of keeping links updated. I |
| 6 |
> see a couple possibilities. |
| 7 |
> |
| 8 |
> One, there's the idea others floated -- note on the page the problem of |
| 9 |
> keeping links updated as to why links aren't provided, and suggest |
| 10 |
> searching on the titles listed, possibly with IBM DeveloperWorks listed |
| 11 |
> specifically, or otherwise with publisher or original source acknowledged, |
| 12 |
> to ease the search. |
| 13 |
> |
| 14 |
> Two, keep the links, but again, specifically acknowledge some of them move |
| 15 |
> from time to time, list them as last verified links, and display a |
| 16 |
> feedback mechanism (email, or web form based to prevent email harvesting, |
| 17 |
> or whatever) prominently asking for reader link update submissions. If a |
| 18 |
> reader says a link doesn't work and it's verified, with no new link |
| 19 |
> immediately available, simply note that it has been verified dead and ask |
| 20 |
> for updates. The idea here is to leave the links as well as acknowledging |
| 21 |
> they may be outdated, but leave the work of providing updates to the |
| 22 |
> readers. |
| 23 |
|
| 24 |
Three, keep a local copy of the articles unless you're forbidden to. |
| 25 |
|
| 26 |
Four, provide a link to a cached copy, e.g., |
| 27 |
http://web.archive.org/web/*/http://www-106.ibm.com/developerworks/linux/library/l-fs.html |
| 28 |
|
| 29 |
-- |
| 30 |
\ Georgi Georgiev \ I hate mankind, for I think myself one of \ |
| 31 |
/ chutz@×××.net / the best of them, and I know how bad I am. - / |
| 32 |
\ +81(90)2877-8845 \ - Samuel Johnson \ |
Archives