This looks to be fixed. SMF got itself all confused somehow and hosed part of the database. Some db repair seemed to clean things up, and I upgraded SMF 2.0.9->2.0.10 since it included some bug fixes, though nothing that appeared to matter for us, maybe something there.

A small number of posts that came through today ended up getting deleted. That Russian thread Jared linked became the home of all new threads somehow for a period of time, and they ended up duplicated multiple times. I split out most of what ended up in there into their own threads. That thread had gotten up to 15+ pages of mostly duplicated posts so I'm sure I missed something. That thread itself also was deleted as it was otherwise weird.

All forum totals and counts were re-calculated after cleaning that up, as something to do with the weird thread messed up some people's counts.

Everything checks out fine now in SMF's db checks, and none of the errors that were flooding its error log have recurred since coming out of maintenance mode, so pretty sure this is all fine now.

Sounds reasonable to me. 
I have only used the VLAN features, personally.  Bridging and LAGGing and all that craziness screams for the need of a switch.  pfSense is not a switch.

@johnpoz:

What about users that use a larger icon set.. Shouldn't the favicon be 256x256 so they can just downsize ;) and get a nice pic..

What about the people that use 48x48 the medium size, etc..  Don't forget the blind people ;)

.ico files are containers (like zip files) that contain multiple bitmap image sizes (or they can only contain one size, 16x16 in this case), and end-use software can pick the right size it wants. That's why, as a web designer, they have to be careful on the .ico construction because , being a container, it can contain many different sizes and thus vary from small in size to huge in size (relatively), which would increase page load times and bandwidth usage. The client will download the entire .ico container regardless of which size image it intends to use inside that container

but, a more contemporary approach, you can avoid the container concept all together and use html meta data to point the devices to the right file. so the end user device only grabs the size image file it needs and not a container with every size image (uncompressed at that)

http://www.favicon-generator.org/ is a great resource that does this for you

ex:

pcap and pcapng - nice!! thanks..

It would definitely be nice if SMF handled timezones better, especially given the brokenness for those whose DST schedules differ from the US. We could run the server on UTC, but so many have set their offset accordingly for CST that it would probably break more than it fixes. That wouldn't fix the DST issue in general either, given the profile setting is only an offset from the server's time and offset from UTC changes with DST.

Not a great answer to this one unfortunately.

that would defently make the forum more usable!

Maybe here?

pfSense Forum » Profile of ________ » Notifications

or for your user id

https://forum.pfsense.org/index.php?action=profile;area=notification;u=31520

@Mr.:

For example, now I see there is a package update for freeradius. I like to check first if something is broken before I update, so her comes the cumbersome process of having to wade and wade and wade.

It makes much sense to give each package its own subsection, not only for Squid, just like the main layout of this forum is, where each 'sub system' of pfSense has its own section too.

The current system of "sub-board by function" seems to be working well for now and has reduced the amount of threads to wade through considerably. If there is enough demand and it makes sense we could add an "Authentication" packages sub-board or some other appropriate name.

My worry with the "every package gets a sub-board" system is the maintenance and usability reasons, we already have a few people that don't use the new system (although it is catching on) and we have to move their threads.

You can also report threads that are out of place, with the subcategory you think it belongs to in the report description. We try and catch them as they pop up, but this helps :)

@kejianshi:

So I don't have to periodically flush all pfsense.org related cookies and cache anymore?  ;D

if you ever hit anything that requires doing so, please let me know. I'd guess that's all fixed now though.

@GruensFroeschli:

Most spam we kill by hand is only a single post per account.
So a rate-limit won't help.

A rate limit is not intended for those accounts.  It's intended for the ones jimp mentiontions below.

@jimp:

They wise up to that and post X+Y spam posts where X=# of posts before links are allowed. It may help stop a couple, but not the bad ones… We get some that drop between 20-50 spam posts in a very short time span.

Thanks.

Thanks, I fixed that so the forum's limit is now enforced instead (~4 MB per file).

Well, your explorer.exe shell crashes… Definitely not forum or pfSense fault. See the event log.

Yeah, if we find them or they get reported as out of place, they can go to the IPsec board.

rgr that

What was the old e-mail address? Currently it doesn't show any notifications configured for your userid.