Mike at Mule Design, who's helped a lot with the new look and user experience of MT4, put it a little more politely, calling it "The Return of Real Betas", but that sounded a little bit too much like "We're bringing Beta back", and that's very Summer of 2006, not very appropriate for a brand-new app.
Here's what a real beta means:
- You will find Actual Bugs. You may even hose your blog! Trust us, seriously. Bugs.
- You will actually have a chance to contribute ideas, suggestions, and fixes. We're not pulling a "here, we're almost done, but we wanted to start promoting it now"-style beta.
- Someday (soon!) the beta testing period will end. We love our friends at Google, but their worst innovation is plastering "Beta" on stuff for a few years until they get around to actually shipping. Don't fret, after the beta, guess what -- we'll start working on the next version!
- We will honestly be listening to feedback. The beta isn't limited to a cabal of our closest friends, or some secret Illuminati-style circle of privilege. We've been on the wrong side of projects that claim to have open betas but don't actually want to hear from the world. As long as you're relatively polite (no physical violence, please!) with the requests, we'll factor them in.
- We're not going to say yes to everything everyone requests. We have to mention this here because otherwise the crazy people will try to make us do stupid things. (Not you, your ideas are great.)
- We'll be updating frequently during the beta. Did the first beta make you cry when you tried to upgrade? Don't worry, there will be another build available soon, and maybe your bug is fixed!
jerame on June 7, 2007, 4:27 p.m. Reply
I really hope the multiple authors/multiple blogs integration is as good as it sounds. I’m trying to build a scalable blogging site that will scale from 25 users to hundreds of bloggers over time. I really need something that can handle such a massive site without having to install 15 different plugins that don’t always work as promised. The MultiBlog plugin is OK, but I’d like to see tighter integration in that area.
I’d also like to see a speed bump on building pages, etc. The FastRebuild plugin really helped with this, so I’d like to see something like that added. Why on earth would you make a commenter wait for all pages to be rebuilt before sending them on? I had a rather large MT blog that I managed that took almost 45 seconds for a comment to post because of waiting for the rebuild of pages. That’s absolutely unacceptable. Accept the comment and do the rebuilds in the background. By version 4, all rebuilds should be background guys.
Can we also make creating custom feeds easier? I would like to be able to offer feeds by category archive, author archive, etc. I’ve never found an easy way to do this without having to write custom templates or something. I need to be able to offer any number of different feeds and easily integrate those into feedburner. I don’t need feedburner support to be built in or automatic, but that would certainly be nice.
I’m really happy about the built in authentication. Much better idea than TypeKey. My users refuse to setup TypeKey accounts, so comment moderation has been a pain. Thank you.
The new interface is OK, but there is a bug with the Blog level navigation. There is no direct navigation to the blog level Tools section. To get there, I have to choose the site level Tools section, then switch to the blog that I want to use Tools on.
Better spam protection would be nice. Let me tell you about a scenario I’ve been facing lately with MT 3.35 for which there is no good fix.
My ISP contacted me about excessive bandwidth usage. When they say excessive, they aren’t kidding! We were averaging 4Mbps INBOUND to the webserver and spiking as high as 28Mbs!! The culprit? Trackback spammers on my MT installation. Here’s what they would do…
They would find my trackback script by crawling the pages. Once they found it, they would hammer it with spam non-stop for hours and hours from mutiple anonymous proxies. MT was blocking the spam from ever making it live and the throttling was even set and working….The spambots just didn’t care. I captured packets and peeked at the traffic to see what was going on. It was insane, they were flooding the tb script with thousands of messages at a time…All randomized with different word combinations just to try to get them through the spam filter. I couldn’t tell if it worked for them, but they seemed to be taking advantage of keepalives to keep the connection to the TB script open while flooding it with multiple spam trackbacks.
Nonetheless, there has to be a better way of keeping this from happening. Even throttling didn’t stop the enormous use of bandwidth - one session sustained at 2Mbps for over 30 hours!! All of that traffic was INBOUND to my server and all of it flowing through the mt-tb.cgi script.
I’ve since created my own fix using the MTAutoBan plugin, a firewall, and some of my own custom scripting. Banning them at the Apache level did no good. I had to use a firewall to keep them out - and they keep trying to get through it. If I shut down my firewall, it takes less than 60 seconds on average before I’m using 2-4Mbps INBOUND bandwidth. The only way I can get them to let go of the script (since I’ve set my firewall to allow established and related connections) is to restart Apache.
I maintain that there is a programmatic error somewhere if that script will allow a connection to process that much data in a single session. If indeed you can use keepalives to keep the script from throttling and then bombard it with multiple trackback requests as it appears, I think that’s a serious problem. The only time I can fathom getting multiple trackbacks in one session is when you self-reference several posts on your same blog. Whatever the case, it should certainly have a limit. A sustained connection from any IP for 30 hours should be seen as more of a threat by the system. That IP should be banned and refused any further connections automatically, without intervention. Some kind of notification would be nice too. You know, the same IP has hit your comment or tb script 300 times in the past hour…You should probably know that since it’s eating up your resources.
MT should be smart enough to tell me in those kinds of events. I have server level controls that I can put in place, but not every user has that option. Getting caught off guard with excessive bandwidth charges or getting shut down for over usage is a very real consequence of this kind of problem. No one wants to face that. There are too many stories similar to mine on the internet for it to be a single incident or particular to my case.
The other major problem I’ve had with MT 3.35 is the search script. The malicious bots (often with the same IP as will later attack the tb or comment script) will crawl my tags with unbelievable thoroughness and multiple connections. Enough to overload my MySQL server and bring my Dual Core 2Ghz 4GB RAM server to a screeching halt. Another scenario that shouldn’t be able to happen. The search script should notice that it’s being excessively used for tags and not allow the tags to be crawled. It’s not a legitimate bot if it’s crawling tags like that.
Thanks for all your hard work. MT is really a great platform. I’ve considered leaving it several times and haven’t yet. MT 4 looks like it is heading in the right direction. Sorry to be so verbose. I hope you’ll take what I have said into consideration. I make a living using MT and I really want to help make the product better.
kwc on June 8, 2007, 5:14 p.m. Reply
I’ve been impressed by the amount of actual listening/responses to the feedback I’ve given thus far — much better than other betas, and I’m much more willing to get involved as a result. Great work!
John Newton on June 28, 2012, 10:00 p.m. Reply
to be honest, I miss the real betas! Finding bugs and later seeing them fixed makes me feel like part of the project. :)
Graham Sullivan on July 19, 2012, 11:45 a.m. Reply
Don’t worry John. They will be back soon or later :)
Emma Howson on August 2, 2012, 5:25 a.m. Reply
(Not you, your ideas are great.) This one is good. As far as I have seen the feedback many of the users are giving - 90% of them are totally stupid.
Fleur on August 30, 2012, 10:39 p.m. Reply
This is probably the best part of Moveable type, because the connection between the team and the users is really strong.
Carol Hannigan on October 2, 2012, 9:42 p.m. Reply
I agree with Jeremy about the multiply authors option. It would be great if there is just one plugin to do it. Installing multiply plugins would probably blow out my hosting. About spam protection it could be great if you implement a version which to put for moderation every post which has link into the text. This will sort it our lots of spam comments.
Amanda Robson on October 18, 2012, 2:22 a.m. Reply
I have a suggestion. Make Moveable Type useful for more people by adding new languages in the tool. The are four Scandinavian languages, many Balkan languages and many more. If you do that you will have many new costumers.
Michael Roberts on November 1, 2012, 11:18 p.m. Reply
This is what I like in beta version. When you have one you can contribute to the whole project without even knowing a bit of code. You need just a good idea and the team will make it.
Charlie Smalling on November 6, 2012, 8:24 a.m. Reply
An unwanted advice from me. When you deal with Beta versions no matter Moveable Type or something else always back up you data. Always and as often as you could. You don’t need to find I am right on the hard way as I did :)
Tarah on December 12, 2012, 9:02 p.m. Reply
There is huge difference between Movable Type and Google. I am not talking about the measure, but the way of thinking. Google team is trying to steal everything interesting on the net and making money from it. MT team is trying to make a product which to be liked from the community.
I prefer to be a MT fan then Google user.