Ever since I deleted an entry this morning, the following string keeps coming up on all MT screens:

MT::App::CMS=HASH(0x83799a8) Use of uninitialized value in pattern match (m//) at lib/MT/App/CMS.pm line 216.

Any suggestions on how to fix it? :-)

I'm getting this on a fresh install of 2.661, also.

The only thing I've found on this seems to be that the USER_AGENT variable (i.e., browser) isn't being set. The reasons are unclear, but it's been suggested that it could be either a browser problem or a server problem.

Vague, I know, but I can't seem to find any clear answer.

After renewing my install of MT 2.661 (full install), I'm getting a CMS error as well. The errors look like:

T::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in sprintf at /home/httpd/cgi-bin/mt-2.661/lib/MT/Util.pm line 457.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in unpack at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 58.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in substr at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) substr outside of string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 59.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in addition (+) at /home/httpd/cgi-bin/mt-2.661/lib/MT/Serialize.pm line 60.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in sprintf at /home/httpd/cgi-bin/mt-2.661/lib/MT/Template/Context.pm line 879.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in hash element at /home/httpd/cgi-bin/mt-2.661/lib/MT/Entry.pm line 110.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in hash element at /home/httpd/cgi-bin/mt-2.661/lib/MT/Entry.pm line 111.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in hash element at /home/httpd/cgi-bin/mt-2.661/lib/MT/Entry.pm line 111.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in hash element at /home/httpd/cgi-bin/mt-2.661/lib/MT/Entry.pm line 121.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in sprintf at /home/httpd/cgi-bin/mt-2.661/lib/MT/Util.pm line 457.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in concatenation (.) or string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Builder.pm line 121.
MT::App::CMS=HASH(0x806a00c) Use of uninitialized value in concatenation (.) or string at /home/httpd/cgi-bin/mt-2.661/lib/MT/Builder.pm line 121.

I'm using XML-RPC to post from ecto as well and this is a new development too. I can't track down what's amiss here and also need help fixing my database (suspected corruption now) so that I don't have to completely re-initialize once again! Any ideas?

-ruminator

Same problem here, but using 2.64 and tried to post using Kung-Log (XML RPC) with a trackback ping.

Errors in Serialize.pm are an indication that your database is corrupted.

I haven't heard of any failproof method of uncorrupting a Berkeley database. Most who encounter it end up starting over with their installations (hopefully, with a MySQL database).

I have noticed what seems to be a pattern of corruption occurring on Berkeley DBs where third party tools have been used for posting.

I was afraid that ecto (the third-party product) might be causing a problem. I was also afraid that database corruption might be the case.

I started to install mysql on my server, but am unfamiliar with it and have a lot of learning to do. (My server is my office linux workstation.) But I can see the benefit of moving to the mysql database if tools are available to fix corruption problems.

{sigh} Oh well. I really like ecto as it gives me access to all of my OS tools for spell checking and so forth. I'm sorry that it's not working.

Thanks!

I emailed ecto support and this is the first he's heard of database corruption and ecto. Is it possible that there is a bug in the XML-RPC code that is causing a problem with third-party products?

I'm not saying that ecto is specifically causing database corruption. Berkeley itself is a finicky thing, and seems more fragile than MySQL anyway, and there aren't always clear reasons why a Berkeley database becomes corrupted.

And there are probably far more people using third party tools to post without database corruption than those who do end up with corrupted databases (just as there are plenty of MT users running Berkeley who never face corruption either).

But based on corruption topics that have come up in the forum in the past, in some of those cases, another common denominator was a third party tool (I think this is the first time I've seen Ecto mentioned; in the past, I'd see NetNewsWire or KungLog). That doesn't necessarily mean any of these tools is a contributing factor, it just struck me as something to keep in mind as a possibility.

Is it possible there's a bug in the XML-RPC code? I have no idea because there again, you have the same issue where many users have not encountered a problem with posting via third party tools. It's much harder to isolate a specific cause for a problem when it isn't widespread.

Interesting. I always wind up being the oddball case, it seems, no matter what I do that is technological.

I went through this afternoon and updated my install to a mysql database, dumped all of my entries into a backup file, and now can try using ecto once again and see if the problem remains.

Will this help?

I think MySQL is always a good idea. But whether it will "help" or not remains to be seen.

Interestingly enough, I've not had another error since switching to MySQL. <crosses fingers>