node validate prevents programmatic node submission
| Project: | Drupal |
| Version: | x.y.z |
| Component: | forms system |
| Category: | bug report |
| Priority: | critical |
| Assigned: | walkah |
| Status: | closed |
Have noticed a few issues with XML-RPC usage since the update to 4.7.3. Some of these things (like MetaWeblog categories) are - I believe - older issues, but the main issue appears to be new.
Attempting to post a blog entry using the MovableType API and any available blogging client (I have attempted to use BlogDesk, Windows Live Writer, and ecto) results in the blog entry being posted, but not published, regardless of settings on the client or the Drupal installation. The error returned is this:
This content has been modified by another user, changes cannot be saved.
I know from personal usage that this was working properly prior to my update to 4.7.3. I cannot personally remember posting since I made that change and having an entry publish successfully. A quick scan of the relevant support forms indicates that this problem is also affecting other users, who are asking the software writers to respond.
On the flip side, categories are unable to be retrieved by several blogging clients while trying to use the metaWeblog API, claiming that the server response was "invalid."
#1
I can't say that I can concur with your expeirence with any release of Drupal since 4.6 right through to the new 4.7.X branch.
I have used several blogging clients to connect with both the MoveableType and Blogger API's and they have worked without incident. Have successfully used Windows Live Writer, w.bloggar, and a couple of others who's names I cannot recall right now. All have been able to successfully perform all relevant functions and posted content has been published without incident.
I strongly suspect the problem may reside with user permissions for the account you are using to connect for posting, or blogging client configuration problems. This issue would be better addressed in the post installation support forum.
#2
I concur with the original post. There seems to be a lot of changes in the xmlrpc code from 4.7.2 to 4.7.3 (I diffed the code). I can get a few blog clients to work with 4.7.2, but they all fail (with the exact error as above) with 4.7.3.
More details at:
http://www.cpzhao.com/drupal_blog_clients
#3
Same problem here. Fresh 4.7.3 install. Same error.
Also the same error with Aggregator2 Module, which I reported here:
http://drupal.org/node/81085
#4
Also seen here - worked fine on 4.7.2, behaviour occurs with 4.7.3 upgrade.
#5
On the concurrences from other users, I am bumping this to "critical," as the blogapi is completely useless if external clients cannot connect to it properly or use its full functionailty.
To respond to the first follow-up from this issue, no blogging client configurations were changed from 4.7.2 to 4.7.3, and the account I am attempting to post with is the *superuser* account.
MovableType API is workaround-able, but not completely useful, as posts can be created, but they are not set to "published" due to the error above. In my experience so far, the MetaWeblog implementation doesn't produce categories when a post is sent to the system, and the Blogger API is pretty much useless because it doesn't support most of the *good* functionality possible with Drupal.
#6
this was broken by the patch in #50316 -- which examines the raw $_POST variable. The problem with this is it's not set for blogapi (and other uses).
I am looking into an appropriate fix for this - though there's not an obvious one atm.
#7
http://drupal.org/node/50316
For the record, HEAD (future 5.0) will need to be updated, too. Since it is critical, it's up to you which one you want to do first.
#8
I had the same problem using Ecto.
A partial workaround is selecting "Blogger" API in Ecto.
You can post new items but you still get the same error message if you try to edit an item.
You have to delete it and send a new one.
#9
#10
http://drupal.org/node/82905 is a duplicate but has a simple code snippet to reproduce the bug , for testing purposes.
#11
#12
How embarrasing.
chx saved the day...
#13
thanks for rolling the patch, Heine.
+1 ... I'd say this is RTBC
#14
omg!
Sometimes problems that sound terribly complex to me end up having so simple and elegant solutions!
Thanks Heine.
The patch will need to be forward-ported to HEAD.
#15
tha attached patch his against HEAD and does not apply to 4.7.
#16
Against 4.7
#17
Committed to CVS HEAD. Thanks.
#18
backported
#19