Topic on Project:Support desk/Flow

Hi.

The question itself if really short and probably asked million times already, but I'm not sure what to do with dead web links. My wiki is now full of dead web links that are reported by Pywikibot/weblinkchecker.py. What are the best practices in this case? I would like to preserve the web links somehow. Should be they just marked with text "dead link"? Or should I just remove the brackets and wrap the web links into the nowiki tag to let them be plain text? Or should I categorize the pages somehow? None of the above seems to be solid and strong.

Thanks in advance.

Weblinks themselves often are not solid. The saying that cool URLs don't change unfortunately is not taken very seriously by many website owners.

Anyway, if the information you linked to now is available at another place, you should just replace the links - which might need manual searching of the new URL and then possibly using something like Extension:ReplaceText to fix it. You might also want to check, if archive.org has the according page archived.

For links, which are not fixable, maybe you can put them into ref-tags - that way they will still be there, but just as a small footnote.

Thanks for the recommendations. Web archive is a very good option since the dead web links still refer old "web 1.0" sites. I'm not sure if I can find that info anywhere else because the referenced information is a sort of unique information. Regarding the ref tags: as far as I understand, the ref-tagged links will still be reported by the web links checker script. I would like to have something like page-baked/printed web links or so (I mean, the web links mentions should be kept somehow, not necessary to be a real link), because those links are a point of interest in the articles.

Yes, ref-tags still produce links and so I would expect that they are still recognized as broken by that script.

So, if I understand you right, you want to keep the links (kind of for historic reasons), but they should not be recognized as broken and they do not necessarily have to be clickable as they only lead to a 404 page anyway.

One possibility (of which I do not know, if it works) would be: If the checker script has a way to recognize links, which it should not check (e.g. by adding a class to the links or so), then you could change your links that way.

nowiki tags might actually be a good solution, if the above is not possible.

You guided me to the Web Archive, so I have to check the web-archived links first, if possible. That's a really very nice idea. I've just written a web-archive template:

<includeonly>

{{#vardefine:YEAR | {{#sub:{{{1}}}|0|4}}}}
{{#vardefine:MONTH | {{#sub:{{{1}}}|4|2}}}}
{{#vardefine:DAY | {{#sub:{{{1}}}|6|2}}}}
{{#vardefine:HOURS | {{#sub:{{{1}}}|8|2}}}}
{{#vardefine:MINUTES | {{#sub:{{{1}}}|10|2}}}}
{{#vardefine:SECONDS | {{#sub:{{{1}}}|12|2}}}}

[http://web.archive.org/web/{{trim|{{{1}}}}}/{{trim|{{{2}}}}} {{trim|{{{3}}}}}] <span style="font-size: xx-small;">(archived {{#var:YEAR}}/{{#var:MONTH}}/{{#var:DAY}} {{#var:HOURS}}:{{#var:MINUTES}}:{{#var:SECONDS}})</span></includeonly>

Example of use:

{{WebArchive | 20010727112808 | www.wikipedia.org | The very first Wikipedia snapshot}}

The result:

web.archive.org / web / 20010727112808 / www.wikipedia.org The very first Wikipedia snapshot (archived 2001/07/27 11:28:08)

It looks very good to me so far.

You guided me to the Web Archive, so I have to check the web-archived links first, if possible. That's a really very nice idea. I've just written a web-archive template:

<includeonly>

{{#vardefine:YEAR | {{#sub:{{{1}}}|0|4}}}}
{{#vardefine:MONTH | {{#sub:{{{1}}}|4|2}}}}
{{#vardefine:DAY | {{#sub:{{{1}}}|6|2}}}}
{{#vardefine:HOURS | {{#sub:{{{1}}}|8|2}}}}
{{#vardefine:MINUTES | {{#sub:{{{1}}}|10|2}}}}
{{#vardefine:SECONDS | {{#sub:{{{1}}}|12|2}}}}

[http://web.archive.org/web/{{trim|{{{1}}}}}/{{trim|{{{2}}}}} {{trim|{{{3}}}}}] <span style="font-size: xx-small;">(archived {{#var:YEAR}}/{{#var:MONTH}}/{{#var:DAY}} {{#var:HOURS}}:{{#var:MINUTES}}:{{#var:SECONDS}})</span></includeonly>

Example of use:

{{WebArchive | 20010727112808 | www.wikipedia.org | The very first Wikipedia snapshot}}

The result:

web.archive.org / web / 20010727112808 / www.wikipedia.org The very first Wikipedia snapshot (archived 2001/07/27 11:28:08)

It looks very good to me so far.