mbsharp

I reported one of the widget problems with random images, and that is under investigation. This report may be the same issue ie, failure to select random images properly for any purpose.
I look at the 3rd bullet point under the changelog for version 3.0.15 and wonder if that fix, including storing the selection for a time period, is causing all these problems.

I need to revise this report about the sidebar widget.

When set to random, ‘all galleries’, there are images displayed (number as in selection control), but the image(s) then don’t change as the page is changed.

Likewise, when set to ‘random’ and ‘except galleries listed’, there are images chosen on first use, but they never change after that.

Normal behaviour has been that the images are a new random selection for each page load or page refresh. I have cleared browser cache but no effect.

Not sure why you have lifted this text verbatim from my recent report on the same problem

I know I gave what I thought was an adequate explanation above to cover the possibility of migrated sites, but I am not sure the plugin does update links regardless of what I said.
I did a lot of work on my sites checking the database content and cleaning it up, but I still do not think links are updated for all content. In theory, the plugin is doing a text search and replace on all active content in the whole database. In practice I don’t know if it is not finding links, or is being crashed out while searching. There is no log of changes made or faults encountered.
Note, we are talking about links to documents here. Links for images are a completely different issue because of the filename variations for multiple sized copies of each image.

I agree. On the site which I said still had the problem, nearly all the pages are pre-migration. Choosing a very recent page which had a new media item (so the page is the media’s parent), and then creating a new test page and adding that same media item (so the test page is not the media’s parent), then replace with new name works on both those pages, but not on another pre-migration page with the same link.

To explain this for other readers.

Your site would commonly contain media links to your own media library, and these would be internal links (although the link is expressed in full with your domain name). Your site may also contain media links to other sites, and these would be external links having a different domain. It is possible that both an internal media and an external media have the same file name (say ‘xyz.pdf’). It is therefore important that media replacement with a name change only changes internal media links and the change action is therefore limited by reference to your own domain name.

A problem arises if you have migrated your site from one domain to another, or even to a different folder level within one domain. Your pre-migration content may be indexed to the former domain even if the link text within your pages/posts now points to your current domain. In this case, media replacement with a name change will bypass the pre-migration content and the link text will not be revised to the new filename.

It will be appreciated that this is an essential safety feature for a media name change. For those affected, the only solution would be a database cleanup to reassign all index values from the old domain to the new one.

Well, now I really don’t understand what is going on. I have updated to the latest version 3.1.1

I have one site where the replace with new name works just fine. On that basis I would have to say that the plugin is OK. This site was a clean install in its current location. The wordpress folder is down a level from the domain at ‘domain/wp/’. The media are stored without use of month/year folders at ‘domain’/wp/wp-content/uploads/

I have a second site where the replace with new name only updates the one parent link of the media. This site has been migrated a couple of times from its original development location, but as far as I know that should be OK. The wordpress folder is at the domain level and media are stored using month/year folders at ‘domain’/wp-content/uploads/’yyyy’/’mm’/. I can see in the plugin code where the set of update locations are selected from the database posts table. I don’t know what would happen if there is an execution failure in the subsequent database update. The echo of the number of updates is commented out.

• This reply was modified 7 years, 7 months ago by mbsharp.

Thank you for looking at this in more detail.

re link text format.
The format for the links, which appears both in the page text editor and also directly in the MySQL database are all in absolute format starting https://
However, note that there is the additional level of ‘/uploads’ following ‘/wp-content’.
Example: https://’domain’/wp-content/uploads/’yyyy’/’mm’/file.pdf

I also have a separate site which does not use the year & month directory levels ie, all media are in the uploads directory. This doesn’t seem to make a difference.

Having looked at the MySQL format descriptions, I see that a media item does only have one parent post/page declared, although it can be linked from others.

The anchor tags also have a rel=”…” term inserted by the page editor. Taking that out manually has no effect.

• This reply was modified 7 years, 7 months ago by mbsharp.

Quite sure.
This issue has been reported over a considerable time and by more than one person (see other relevant threads).

Regret, no change. To be clear with an example.

My site is mainly static pages with a few posts.
Say I have an item ‘abc.pdf’ in the media library, and I have included that via a link on ‘page1’ and also on ‘page2’. In the media library, ‘abc.pdf’ is shown as linked on ‘page1’ but I think it’s a WordPress standard not to list other pages where the media is also linked.
I use media replace to replace ‘abc.pdf’ in the library with a new file called ‘def.pdf’ and ask for the name change with change of links.
‘page1’ now brings up ‘def.pdf’ on its link as expected. ‘page2’ still has the link to ‘abc.pdf’ so it brings up a 404 report becaue ‘abc.pdf’ doesn’t exist any more.
I use media replace to replace ‘def.pdf’ with ‘abc.pdf’ changing the name and links ie, restoring the original content. Both ‘page1’ and ‘page2’ now work fine.

So, media replace with a name change replaces the link on the page cited in the media library, and does not change any link on another page.

Unfortunately, although we have a version update from 1 day ago, this particular problem has not been fixed.

I have not had this problem with WP4.7.2 using pdf and docx file types. Whether the problem is specific to doc (or some other types), it is not the case the media replace is limited to image files.

As far as I can tell, this is expected behaviour. There are several previous discussion threads on this topic.

I am using it with WP4.7 and no problems

An update to this report.

Whether I organise the media library in year/month folders, or not, makes no difference. I have two web sites with one organised each way.

As noted, WP only ever displays one page on which the media is loaded and does not display the name of any additional pages where the media is also loaded. Using this plugin to replace a media item with one having a different name, then the name link is updated for multiple places on the same page/post (which is the one displayed in the media library by WP), but is not updated at all on any other page.

I am aware of the issue with image files and different dimensions. This report only refers to pdf files.

Hi,
I am implementing EntryWizard for the Chilterns Association of Camera Clubs. I looked at the features for either prefixing the users filename or replacing it within EntryWizard. On balance I felt that it is an advantage for the user to have no particular filename format required, in which case a synthetic filename is required. But, the hassle of implementing any arbitrary filename in EntryWizard would be too much work, and there will always be incompatibilities between filenames that WordPress is happy with and filenames for some competitions display/scoring systems.
My solution is to download the images and spreadsheet, and then use an Excel macro driven system to create the required filename within the spreadsheet and rename all the image files to the new filename format. The folder/filename structure is always going to be some combination of a Category, a Sequence number, the Title, the Author and the Entrant (ie User account), with specific separators.
I have done this for the ICPro competitions system, and it could easily be modified for DiCentra or any other system.
There is no need to pile complexity into one system, when the work can be easily split and done separately.