773659 - Stale entries on the Mozilla Site Directory

Status: RESOLVED FIXED

(www.mozilla.org :: General, defect)

Description

[Daniel Maher [:phrawzty]]

Hello,

The following is a list of sites which are currently linked from the Mozilla Site Directory, but which appear to no longer be valid (expired domains, non-existent content, etc) :

    blogs.mozilla-europe.org
    brasstacks.mozilla.com
    chilemoz.org
    mozilla-europe.org
    mozilla-hr.org
    mozilla.lk
    mozilla.mk

I would suggest that these links be removed from the community page.

Comment 1

[Corey Shields [:cshields]]

adding stephend..

Is there some kind of link checker in our webqa process?

Comment 2

[Stephen Donner [:stephend] Not actively reading bugmail]

(In reply to Corey Shields [:cshields] from comment #1)
> adding stephend..
> 
> Is there some kind of link checker in our webqa process?

There is indeed; https://github.com/mozilla/marketing-project-template/blob/master/pages/page_object.py provides a lot of functionality, and this isn't even in expensive Selenium.  We only run it for supported sites, but can run it on-demand, if needed.  There are also easy-to-run manual tools, such as Xenu, on Windows, we could use.

I believe there's a transition plan in the works for Mozilla.org (former) -> Mozilla.org (current, i.e. "One Mozilla") pages, but there are a ton of pages, so might take some time.

Comment 3

[David Boswell]

For background, those links had been manually maintained but are now generated by a script Chris has created.  He has some staleness checking in there but it may need to be tweaked to catch more examples like this.

Comment 4

[Chris More [:cmore]]

I'm re-running my scripts that generate this page amount updating the domain wiki pages. I also added a few exceptions that should clean these up. It appears that these have gone stale since the last time I ran the scripts. 

Thanks!

Comment 5

[Chris More [:cmore]]

Attachment: Updated version of directory

I've made changes to the filter and found a few exceptions. Check out the this file, which will be the new directory and see if this looks better.

Comment 6

[Chris More [:cmore]]

David: If the attachment looks fine now. Do you need me to combine it with the existing page and you commit it to SVN?

Comment 7

[David Boswell]

Yes, I can commit it to SVN if you want to send me a new version of the full list to post.

Comment 8

[Chris More [:cmore]]

Attachment: Full directory.html file

PHP code + updated directory.

Comment 9

[David Boswell]

Checked in at r108185.

Please verify that this looks right -- should be live in 10 minutes or so.

Comment 10

[Stephen Donner [:stephend] Not actively reading bugmail]

The only "broken" links (which aren't 404s), are:

https://caminobrowser.jp/
error code: 12157 (error 12157), linked from page(s):
	http://caminobrowser.jp/

https://getthunderbird.jp/
error code: 12157 (error 12157), linked from page(s):
	http://getthunderbird.jp/

And those are returning: (Error code: ssl_error_rx_record_too_long)

Comment 11

[David Boswell]

Some of the non-English text got messed up in my earlier commit -- that should be fixed with r108186.

Comment 12

[Chris More [:cmore]]

I've fixed another exception in the script and it will catch those two .jp errors and others like it. The domain would redirect to https and then the https would fail. I will upload a new copy unless David manually removed caminobrowser.jp and getthunderbird.jp.

Comment 13

[David Boswell]

I've removed those two links in r108213.

Comment 14

[Chris More [:cmore]]

Attachment: Updated version of directory

David: here is the latest version.

Comment 15

[David Boswell]

Checked in at r108214.