Testing out WordPressMU.

So I’m playing around with the multi-blog version of WordPress to see how it works out. I had installed just a single copy of the stand-alone WordPress first and that worked just fine, but I’m a little worried about WordPressMU. Not because it’s not working, but because I can’t find where it put the database tables.

I’ve got one database on this hosting account, but that’s not a problem because the different CMS systems prepend their tables with unique identifiers so they stay separate. When I installed the stand-alone WP I could see the tables were marked with WP_ so when I removed it I just dropped those tables. When I installed WPMU it was supposed to do the same thing, but no new tables are showing up in the DB. Yet it’s working. Color me confused. So this post is basically to say that if you notice anything wonky happening to drop me a note and let me know. So far EE appears to be completely unaffected, but you never know.

On the plus side the MU version of WordPress works pretty well and allows for a decent amount of control over what blogs set up under it are allowed to do. One unknown at the moment is whether or not it allows you to use distinct domain names for blogs you’re hosting. It assumes you’ll be using subdomains and that’s fine for some of my blogs, but my mother’s (among others) have their own domain name. Ultimately it’s a toss-up if it’d be better to go with the MU version or just install separate instances of the stand-alone software. It would certainly make keeping track of spam and such across the various blogs, but they would be limited only to the themes and plugins I make available to them. Which may or may not be a bad thing.

At this point I’m leaning in the direction of making the switch. WordPress has come a long way since the last time I played with it and a lot of the features it was missing are now available either in the core or as plugins. I think ExpressionEngine is moving more and more away from blogging and into full-blown CMS territory and it does that very well, but as a blogger I’m feeling a bit left behind. I’d still like to see what EE 2.0 has to offer, but unless I’m blown away I’ll probably be making the switch. The hard part now is converting the 10,000+ entries across 12 blogs (some of which are admitted defunct at this point) over to WP. Not to mention the nearly 78,000 comments in SEB alone. WP has a shitload of importers for all manner of blog systems… with the exception of EE, natch. I suppose I could just start from scratch for SEB and just worry about converting the smaller blogs, but I’d hate to do that. And I still have to decide whether to do stand-alones or go with WordPressMU. I’m sure I’ll be debating it for some time to come. Hopefully EE 2.0 will release more info soon so I can factor that into my decision.

26 thoughts on “Testing out WordPressMU.

  1. Les, as you know there should be an easy way to migrate content—use MT import format. There are sample export templates in EE’s wiki, but they have a potential pitfall: Rather than faithfully exporting in BBCode/pmCode like it’s in the database, they HTML-format everything. If this is an issue, you can probably work around it by (temporarily) changing the Default Text Formatting of the relevant fields.

    EE has always been a CMF (F for framework) rather than a pure blogging platform and I don’t see this changing—which is actually a good thing. What I do see changing is that there are more and more commercial add-ons, which also seem to be getting more expensive. Rewriting EE 2.0 in CodeIgniter ultimately may be beneficial, but software engineers generally frown upon these whole-sale rewrites. Consider what they could have done with EE 1.x if they hadn’t been tied up with the rewrite for what, going on two years? I’m also not very happy about how tight-lipped they are.

    Like I said before, I myself see no urgent need to migrate my sites off EE, but I’m also not married to the platform.

  2. This is just as I’m considering a push for an EE pilot project in our college.  The alternatives, Drupal or M$-enslaved SiteCore, are not thrilling me. 

    Cost is not really an issue; if they can pay for SiteCore, they can afford about a hundred EE licenses with all the commercial plugins.

  3. DOF, EE is a great platform. Les seems to be itching to try something new, that’s all LOL

    Since you asked about disadvantages of Drupal, here are the problems I see with EE.

    The single most annoying mis-feature of EE is that a user can be a member of one group only. This is a major PITA if you want to run a site with very fine-grained and occasionally changing access permissions. It’s trivial in Drupal.

    Many of EE’s standard modules are rather, uh, bare-bones. The gallery, wiki, and forum are all functional, but there always seems to be a critical feature that’s just not supported.

    Right now, the timing of the EE 2.0 release is a critical issue. A release date of a few months ahead seems to be firming up, with it’s intrinsic and self-inflicted FUD: Is it a good idea to build a site on EE 1.x when the price and ease of upgrades to 2.0 are yet unknown? I would guess that a new 1.x site has a grace period of two years or more, but still…

    I just did a test install of WPMU and it’s like I remember: meh. One typo on the installation screen and I had to hand-edit a config file to recover—the installation screen wouldn’t reload. The control panel has gotten a lot slicker, but I doubt I could build the sites I like to build with WP(MU). And looking at the themes/templates, they appear to be as iffy as Drupal’s…

  4. The hard part now is converting the 10,000+ entries across 12 blogs (some of which are admitted defunct at this point) over to WP. Not to mention the nearly 78,000 comments in SEB alone.

    C’mon, Les, what’s the prob?  You can read, right?  You can type, can’t you?  Just type the comments up in the new format.

  5. Elwed, I think the MT Import format would work fine for most of the smaller blogs I host (my parent’s, my sister’s, etc.) but I’m not sure it would work well for the monster that is SEB. In fact I did a test export of my mother’s blog last night using the templates from the EE Wiki and was able to upload most of it into the test WPMU install. If I were to upload the file first and then import it it’d probably take all the entries just fine.

    My mother’s export file was 3.89MB, which is well within the 10MB limit of the MT Import function. I don’t think SEB would fit in under 10MB.

    I still like EE quite a bit, but there are features I’ve literally been waiting years for that either still aren’t available or are but only through a rather expensive module. Just adding tags, functionality that is now core to WP, cost $40 when I added them in. The best of the WYSIWYG editors for EE, one which actually looks like it integrates well, is called Wygwam and it costs $30 and requires an entirely different module (FieldFrame) be installed to work. Again, this is functionality that is core to WP. Not something I need so much, but it would make blogging a bit easier on my parents.

    I also admit that the fact that the ExpressionEngine devs recently said they would not be going into greater detail about what EE 2.0 will have in the way of features is a sore point with me. If I knew what was coming it would make the decision to stay or go a lot easier.

    DOF, don’t let my investigation into alternatives sway you one way or the other with regards to ExpressionEngine. It’s not so much that there’s anything wrong with EE as it is my ever increasing need to tinker with things that’s driving this. I like experimenting with new addons and features and it’s been a long time since I’ve had anything new to play with on EE either because A) it’s not available or B) it’s available, but costs money. It’s not that I’m opposed to commercial modules. It’s just that I never know whether or not I’ll use them long-term (polls anyone?) and spending money on something you’re not absolutely sure you’re going to use is silly.

    I have no doubts that EE would be great for your college’s pilot project. Honestly if it weren’t for my need to have new toys to play with there would be no reason for me to move off of EE 1.6.X. It still has one of the easiest and most powerful templating systems available. I’m not committed to change yet, I’m just leaning that way and it’s entirely a me problem, not an EE problem, that’s driving it.

  6. Les, there are workarounds to overly large export files—like exporting 1000 entries at a time or so. You can easily automate this process, too. If you’re planning to use the WYSIWYG editors, then exporting as HTML is actually a good thing and good riddance to BB/pmCode.

    And I hear you about EE’s features (freeze) and the tinker itch.

  7. I believe that for my main blog, I ended up exporting everything (in MT format, since it was MT) and then just splitting the import files with a text editor, and WP handled importing them just fine.  It was prey easy-peasy.

    The historian and diarist in me would cry if you deleted (or left stale) the old content and went for a fresh install.

    The biggest concern I faced was losing the xrefs between the posts (and anyone’s external links to the blog), since the way MT and WP make real-name posts differed in default length and in the joining character (MT, once they decided to support it, uses “_”, as, apparently, does EE; WP uses “-”.  MT also supported much shorter names).  There’s some documentation out there on how to fix that (I think I documented it during my own conversion), and it adds a bit to the complexity, but as a community effort, it’s probably worth doing here.

    Do let me know if I can assist.

  8. What ***Dave said. Most of the content can be easily moved, but links, xrefs, and character conversion issues (you know, ÖÜßàáâãäåæçèéêïö and stuff like this) can be really annoying to fix, particularly if you miss anything before the site goes live again.

  9. Not too worried about links as there was a trick I did back when I moved from MT to EE that involved a massive .htaccess file for URL redirects that I could probably do again. Google will eventually re-index the site anyway. The internal links might be an issue, but there aren’t so many that it would be impossible to overcome. Both EE and WP use xhtml so most of the character conversion stuff shouldn’t be an issue. Before heading into work I tried exporting the last 1,000 entries from SEB to import into a test blog and ran out of server memory. Will probably have to wait until early morning to try it.

    I’ve not read too much of the documentation yet, but so far it appears that the blogs I’ve made under WPMU all point to the same path, which is different from how EE handles things. Makes me wonder where it stores uploaded images and such.

  10. Both EE and WP use xhtml so most of the character conversion stuff shouldn’t be an issue.

    If you look at the page source of this thread, you will see that there is a potential character set conversion issue. The funny characters I used earlier are not HTML-encoded, but native UTF-8. Same goes for the beloved ™… It took me hours to fix this kind of thing after the fact.

  11. Hmmm. There’s probably some way around it. Just have to be clever I suppose.

    So the other thing I’m pondering now that I’m pondering things is if I do make the switch to WP should I also change webhosts? I’m using EngineHosting mainly because they’re experts at hosting EE, but I’m not real happy with their account control panel as you pretty much have to email support to do anything. There’s at least a couple of webhosts out there that advertise as catering to WP and WPMU users and are a bit cheaper to boot.

  12. As long as all the tools agree on the encoding standards, there shouldn’t be a problem; if not, it’s a solvable problem. It’s the not being aware of pitfalls that gets you wink

    WP recommends a couple of hosts, all of which are loads cheaper than EngineHosting. Whether their shop runs as smoothly as EH’s is another question.

  13. Well I just moved the install down to my root domain (jenkinsonline.net) for testing purposes. It doesn’t conflict with EE there so I’m able to run both concurrently. The only exception is the base Jenkins Online blog which hasn’t been updated by me or the wife since 2005 so it’s no big loss.

    I also exported the last 50 entries from SEB and uploaded them to this test site. One big issue I can see already is that WP eats all the embedded YouTube videos I’ve put in the entries. The standard MT import format doesn’t handle tags either. If I do switch I’ll have to see if I can dig up a schema for WP’s own import/export format and see if I can’t modify the template to output in that format.

    Feel free to play around on the test site. User registration should be possible and I’ve turned threaded comments on.

    Update: I’ve removed the install of WPMU that I was playing with so I’ve deleted the link from the comment.

  14. Used the templates from the EE Wiki. The code in the text file shows up properly, but WP eats it on import. Tried editing the post and pasting the code in using the HTML editor (as opposed to the Visual editor) and it ate it as soon as I updated it.

    Seems WP doesn’t like divs and such in an entry. Will have to check and see if they have options to relax that restriction. There’s a number of plugins for embedding YouTube in WP out there so that may be the only way around it. Doesn’t help at import though.

    Sent an email off to one of the WP experts on Twitter on whether to use WPMU or just standard WP. She’s admits to being biased towards the former, but admits it’s a bit more work to administrate than a stand-alone and not all of the themes and plugins will work with WPMU. That could be a deciding factor as while I may try to make my own custom themes, the other folks I host may want to use some of the premade ones.

  15. On a different note, here is my list of plugins I give to all who make the switch:

      Best SPAM plugin for blogs eva!
      Allows you to send a post right to Atheist Spot, works similar to Digg
    Digg It:
      Allows you to digg a post
    Paged Comments:
      Just what the title implies
      Best CAPTCHA eva!! Easy to administer, lots of features works amazing
    Register Plus:
      Must have. Gives much more functionality or the registration system of users. Allows you to have a CAPTCHA for users wanting to register.
    Subscribe To Comments:
      Just what the title implies
    Viper’s Video Quicktags:
      Gives post authors quick tags for embedding all sorts of videos
    WordPress Database Backup:
      Just what the title implies. Give a lot of functionality to this too.
      Gives extended functionality to page navigation. If you look at the bottom of my blog you can see what I’m talking about. Very customizable.
      Allows blog readers easy access to printing blog posts.
    WP Grins:
      Just what the title implies. Smilies…
    WP Super Cache:
      Must have for those worry about having their site shut down due to being dugg. Might become less of an issue as web server hosting gets increasing access to bandwidth.

  16. 1. It may be a matter of how you have embedded YouTube stuff in EE, but I can assure you that I have embedded YouTube in WP by just pasting in the embed code without any problems.

    2. The default WP character set is UTF-8.

    3. If you are looking for a RTF editor to enter stuff in with, the default WP site admin screen to add posts does quite nicely.  I use the external Linear (nee Ecto for Windows) client, myself, which is now free for download.

    4. WP doesn’t mind divs in a post.  Unless you do something really screwy with them, I imagine.  I’ve used divs for centering text and other objects, without a problem.  (It may be, in part, a matter of what theme you use.)

    5. My plug-ins are listed, with comments, here: http://hill-kleerup.org/blog/my-wp

    Ones of note are “Subscribe to Comments,” “Redirection,” “Twitter Tools,” “Digest Post” (for pulling in GReader links), and “WP-ReCaptcha”—but all the ones there are ones that I use.

  17. I find I have to remember to embed YouTube in the ‘mark up’ view, rather than the Wysiwig view, or it doesn’t work.

    The thing I like about WP is that is part of the typekey group, so non WP blogs recognise it, and so I can sign in under WP on TypeKey- PZ has just had to restart logging in on Pharyngula due to spam, and I don’t need a new account.

  18. Well, after much experimentation and dicking around, I believe I have successfully created a template to export from ExpressionEngine to WordPress’s WXR format.

    It successfully transfers categories, tags, entries, comments, and trackbacks. Category hierarchy is lost in the translation, but it’s not too hard to edit that by hand to put things as they should be. Now the only challenge remaining will be doing 65+ exports (6500+ entries at 100 entries a time) if I should finally decide to make the switch.

    I’m such a clever boy.

  19. Huzzah!

    Yeah, it’s very easy to rebuild or change category hierarchies within the WP system admin screen.

    Is the 100 entries the most your server will handle exporting?  You mentioned 1000 as an upper bound previously.

  20. The magic number is somewhere in the 100 range, but it depends on the time of day I do the export. I seem to be able to get more out in the middle of the night than in the middle of the day and the obvious reason is the lower traffic on the server.

    It’d be a snap if I could do 1000 at a time, but with some entries having upwards of 400 comments I’m probably pushing things with 100 at a time.

    Right now I’m waiting to see if I’m allowed in on the EE 2.0 beta before making a final decision one way or the other.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.