So WordPress 2.3 is out, and the release announcement is exciting to read.

First off, in 2.3, WordPress gives us real Tags. Tags are not categories. I was really disappointed when WordPress.com started confusing categories for tags, and wrote about why Tags are not Categories. WordPress has finally come clean, and seen the light, and explained what’s what over at the Tags and Categories article at WordPress.com blog. Now to take it a step further in the right direction, it would help if everyone thought about how tags add value to a blog. Tag Clouds are cool, but they’re not the be-all and end-all of tagging. Here’s some great ways to use the tags you painstaking created by tagging individual posts:

1. Improve search results on blogs by searching for articles containing the search term(s) as tags.
2. Provide an easy way to navigate the blog using Tag intersections, unions, and other common operations on sets. This would make it possible to find articles where your favorite blogger talks about two of your favorite things, or one thing and not another. The focus is on making information easier to find.

The above should replace the “site search tags” found under almost all of Lorelle’s articles, and then some.

The second step in the right direction is the removal of the links to developers’ blogs from the default blogroll. A lot of the old developers and volunteers have been inactive and non-contributing members for a long time now. Indeed, it was good to have a PageRank of 8 for my own blog, but I had the link to my blog removed about a year and a half ago when I decided to stop scaling down on my WordPress volunteer work. I did that because I thought that was a part of stepping down gracefully, and wasn’t too comfortable with going, “So who the heck is this ‘Carthik’ in my blogroll?,” long after I stopped contributing. Also, I wasn’t sure my contributions were at-par with those of some of the others who put in way more effort. It’s hard to decide where to draw the line with recognizing contributions to open source projects anyways. It’s better now since no one can point at finger at some old contributor profiting without contributing by means of selling text links on their blog now. Good job!

The MovableType importer is no longer as memory hungry as it used to be. You can also add new importers by installing importer plugins. Very cool.

A lot of the changes were from the most wanted changes/ideas proposed by users, and that is good news in itself. Now if only some of the other ideas are implemented, like searching in both “Posts” and “Pages” by default, instead of just in Posts.

So, on a personal level, these are my thoughts about the changes in 2.3. The only small thing that irritates me are the new names in the db:

Three new terms tables (schema) term_relationships, term_taxonomy, and terms support combined post categories, link categories, and the new tags. The tables, categories, link2cat, and link2post, are gone!

I am yet to understand what “terms” are, what “taxonomy” is and what kind of relationships “term_relationships” describes. If these tables deal with categories, tags, and I assume, blogroll links, wouldn’t it better to have those terms in the db table name, instead of the term “term”? I wonder. The new db table names give away nothing about their intended purpose in the database. All three terms are brand new, never before seen in the WordPress world. I am sure there must have a good reason to name them thus, which I’d love to know more about.

So, in your opinion, what’s wrong about WordPress 2.3?

BuiltWith.com is a service that tells you what runs a site, the statistics packages used, the blogging software used in case the site is a blog, whether the site uses RSS, Javascript, the works. I found BuiltWith via Mashable, one of my occasional reads.

Interestingly, BuiltWith.com also provides the percentage of profiles sites that use a particular technology, and I LOVE stats. Back in the day I had a pet project going that sought to keep track of all the WordPress powered blogs in the world, and map them on a world map. Yes, I know, I had no idea how popular WordPress would turn out to be. That application of mine soon started grounding servers, but we are getting off track here.

When I checked wordlog.com at BuiltWith I noticed that 8.27% of all profiled sites use WordPress to power the blog. So I got to thinking, “What about MT, Typepad, and Blogger?” Turns out BuiltWith cannot identify MT powered sites, like Boing Boing, so we’ll never know what percentage of the profiled sites run on MT. Maybe MT sites don’t put out that they’re powered by MT (they don’t have a “generator” meta field) so conspicuously, maybe BuiltWith never thought MT important enough – we don’t know. But only 0.52% of all profiled sites use Typepad, and only 1.55% of the sites use Blogger. In the process, I also realized how, by looking at a site, I am generally able to figure out what runs the site. If a site is WordPress or MT powered, I can usually tell just by browsing the site and looking for some signs.

I am not saying WordPress blogs outnumber Blogger blogs by a factor of 5 or more, the results might just as well indicate that WordPress users are among the more inquisitive, wanting to try out new tools, such as BuiltWith, which might explain why more profiled blogs are powered by WordPress. In any case, BuiltWith.com owes me around 15 minutes of my precious time. I suspect I will check back later, when, hopefully, a more representative cross-section of website owners have checked their sites at builtwith. I can think of some cool things those guys could add to the details – such as what resolution the site was designed for, or whether the layout for a site is fluid or fixed-width – all details that can be gleaned from the css stylesheets. Then again, it looks like only 87.6% of the sites profiled use CSS – I wonder what the rest of the 12.4% are thinking, really.

Movable Type is going open source. Yup. MovableType 4 Beta has been announced and the plans are to make it a fully GPL-ed, open source product. The announcement says:

…we wanted to explain why we’ve made this massive investment in Movable Type and its community. The truth is, Movable Type is how Six Apart got started, and one of those ideas that motivated us from the very beginning is that we’ve got a mission.

And then the rest of the post does nothing to explain clearly why they’ve made this massive investment. The rest of post explains how blogging empowers ordinary people, and that their mission is to get blogging to the masses. It also explains the role of Vox, LiveJournal and TypePad in the Six Apart world. A free version of MT was always available for personal use. So why make it open source? That’s the answer I was looking for, and did not find.

I still clearly remember the day Mena announced the new pricing structure for MovableType. A day later, I wrote up a post on how to move from Movable Type to WordPress. Then, the exodus began – with famous users, like Mark Pilgrim, Molly, and so many others shifting to WordPress. Those were busy days on the #wordpress IRC channel. The number of opinion-leading bloggers who used MT started declining, and most of the new bloggers who came after chose WordPress. While there is no reliable count of the number of WordPress Vs. Movable Type users, I dare say more folks use WordPress than MT.

At the time of the announcement of the new pricing structure, the rationale was:

In preventing web hosts from offering Movable Type for a fee, we had to put a pretty strict blanket clause on our licenses to cover all cases when compensation was earned. Buying a commercial license to perform services was a bit backwards, but at the time, it was the only way we felt we could control the situation.

So the new pricing structure intended to let personal users continue to use MT while stifling other commercial entities from “profiting” from their product. How has this changed today? Today, not many commercial entities offer MT installations, I suppose. There is no evident mass-interest, and so they probably are no longer a significant threat.

Some answers can be found at the Movable Type open source page. They say the reason for an open source release is as simple as that the users asked for it. They asked for it en-masse just about 3 years ago. So though it sounds like a valid reason, I suspect that is not all there is to it. Whatever the reasons behind the change of mind, I welcome it. WordPress needs competition. I just wish the MT 4 Beta announcement wasn’t so wishy-washy. Did they make a mistake in announcing their pricing structure earlier? Did they realize that releasing their flagship product as an open source product will revitalize their mind share? Did they do this for business reasons – if so, what is the rationale behind that move? The MT back-and-forth would make for a very interesting case study of open source products vs. products with a variable price structure. It’s a case study in the making, and I can’t wait for the results. Time will tell which worked better for them – I hope Six Apart will share the numbers and the experience in a transparent fashion.

There are at least two people in the world who definitely will appreciate the plugins I mention in this post, and that, I think is reason enough to write here. Prior to WordPress 2.1 here was a neat little “Preview” link to the Top Right of the Write Post form. Clicking on that took you straight to the bottom of the Write Post page where you could preview the post as it would appear on the site in an IFrame. 2.1 took that away from us, annoying many. So someone filed a bug, which is now marked as Fixed. The Fix in 2.2 has been to include a link in the same place as before, but which opens the post in a new window/tab. If that does not fix it for you, here’s a plugin that surely will. This adds back the IFrame.

Read about and download the Preview Frame plugin.

I apologize for not being able to try it before mentioning it, but please do let me know how well it works for you, if you choose to use it.

Update: You may need the Preview Post Navigation plugin to add the “Preview” link on top. This will work for WordPress 2.1.3. For 2.2 you will have to edit the Preview Frame plugin to add a line “<a name=”preview-post”/>” just above the IFRAME line in the plugin source to make the “Preview” link work. I would have made the change and re-released the plugin, but the original plugin does not specify the license it is released under, so I desist.

I had written in the post describing 2.2’s arrival that there was no mention, and it wasn’t clear from the WordPress bug tracker whether 2.2 included any security updates.

The recently married Mark has left us a comment saying WordPress 2.2 Getz includes some security fixes. As with all upgrades that include security fixes, you are encouraged to upgrade to this release as soon as you can. You don’t want any holes, big or small in your WordPress blog. The reason I upgraded WordLog and resumed posting is because someone used a hole in the old, un-upgraded version to upload copies of movies, including Hannibal Returns to my VPS. Yes, it can get that ugly! So upgrade now.

WordPress 2.2 codenamed “Getz” has been released. Read the official announcement for details.

This is all part of the new experiment to release often, and release regularly. Finally Matt’s 2005 resolution seems to be kicking in

Widgets in the core, that annoying bug where marking comments as spam in the admin page causes the number of displayed comments to keep decreasing has been fixed, and loads of cookies for developers, and lots of bugs including some “Highest OMG BBQ” bugs fixed.

What’s unclear is if this is just a feature-upgrade release or if it includes any security fixes? Do we all have to upgrade to be secure? It would be nice if a small note were added regarding security in the official release blog posts.

I happened upon a curious trick to find out all your pages listed as “supplemental results” in google, and some other associated supplemental result tricks. A lot of you might already know these tricks, but I think reading the rest of this article might get you thinking about these supplemental results in a new way. I spent a good part of 1.5 hours playing with this stuff and reading up on it, which I try to summarize here.

To start off, let us look at the tricks.

Finding all supplemental results for your blog

The trick is to do a search for the string “site:wordlog.com *** -spght” in google. That gives you all the pages on your wordpress blog listed as supplemental results. The search result that google returns will have a “Supplemental Result” in the text that follows the url and the short excerpt, and as you can see, all the results for the string I refer to above have that after the results. The spghy can be changed to some other random characters – it doesn’t matter.

Finding all results that are not supplemental results

The following query will show all results that are not supplemental results:
“site:wordlog.com -allinurl:wordlog.com“.

So, for wordlog.com, there are 227 non-supplemental results and 196 supplemental results. However, a search for “site:wordlog.com” returns 325 results, and 196+227 = 423. So I think some of the results returned for “site:wordlog.com” are supplemental results. At the time this article was written, page 25 of the results has two supplemental results right at the top.

What are Supplemental Results?

According to Google,

A supplemental result is just like a regular web result, except that it’s pulled from our supplemental index.

and, additionally, Google maintains that

…the index in which a site is included is completely automated; there’s no way to select or change the index in which a site appears. Please also be assured that the index in which a site is included doesn’t affect its PageRank.

So we know that there is no way to formally request supplemental index pages to be moved to the main results pages. However, one thing bothers me, sort of.

Most of the non-supplemental results for wordlog.com are the archive and category pages. I believe the individual posts should be there in stead. I have noticed, many times, that when I search for a term, I am most often led to the category or date-based archives of a blog, and then I have to manually search for the term again in Firefox, and then, since many themes display only excerpts in these pages, i have to click the article to read it to get the information I need. This is annoying, to say the least.

Fixing the supplemental results problems

There is a duplicate content cure plugin for wordpress that promises to reduce the duplicate content indexed by google by way of your archive and category pages. It does so by adding directives to google to not index archive and category pages by means of meta tags in the page headers. One would think this would cure the supplemental results problem too, and make all your blog posts preferred over the archive pages.

As a small experiment to test this theory that the duplicate content cure plugin will help alleviate the supplemental index problem, I did searches for supplemental and non-supplemental results for seologs.com, the site that published the plugin. Amazingly, seologs has 385 supplemental results and 243 non-supplemental results! So now it appears that the plugin is not the silver bullet for the problem. However, as promised by the plugin, the archive pages are missing from the pages indexed by google. Is this a good thing, though? If the number of indexed, non-supplemental pages are the metric, then it is not. Without the plugin, all of wordlog’s archives are indexed and probably will be returned as search results for some terms. The duplicate content cure plugin prevents some pages from being indexed, totally – it would be nice if it did not do that, really. It is better to have visitors find useful content via your archives if not via a direct link to the relevant article.

Ideally, I would love for the archives pages to be indexed too, with the blog posts being indexed in the main index. Heck, I would love to have all the pages in the supplemental index to be in the main index instead. There are lots of suggested tricks to avoid the supplemental index. The issue with archive pages in wordpress blogs being indexed more prominently is because all WordPress blogs have relative links to the archives pages that look like the following if you look into the source of the page:

"<link rel='archives' title='May 2007' href='http://wordlog.com/archives/2007/05/' />"

In addition to this, you also have links to the archives from the sidebar, which is probably displayed on all pages of your site. The indexing robots should think these pages are really important, since you seem to link to them from every page on your site.

So, a simple way to fix the problem, or at least try to get some pages in the main index might be to have a sitemap containing each and every post on every page in your blog. That would make the pages huge! An alternative would be to have an html sitemap and link to it from the the sidebar or footer. You could also link to posts you think are important from the sidebar. The important things to remember are that:
1) It’s better to have a page in the main index than the supplemental index.
2) It’s better to have a page in the supplemental index than to not have the page indexed at all!

I have a couple of ideas floating around in my brain that I will implement to accomplish item #1 above without violating item #2. I will try them out and let you know if the results are worth mentioning. Do you have any ideas that have worked, that can be verified in a straightforward manner? Blame it on what I do for a living, but I have come to trust verifiable results over speculation and hypothesis.

I badly wanted to display the following html code in a blog post that I am drafting:
“<link rel=’archives’ title=’May 2007′ href=’http://wordlog.com/archives/2007/05/’ />”

I tried putting the raw line of html between “code” tags and then even “pre” tags, but it just wouldn’t show up. Turns out you need to encode the html entities, like the “<” by replacing them with their html code equivalents – & lt ; is the equivalent for the left angle bracket.

There is a cute little encoder tool at wordpress.org written by The matt – however it seems to not want to work anymore. So I found a replacement Encode/Decode HTML entities tool at centricle.com. It works, as evidenced by the code displayed in this blog entry.

I sure can’t be the first to notice this.

If you go to a feed url, say http://wordlog.com/feed/ and then click on the Google Reader Subscribe Bookmraklet, the poor thing can’t find the feed!

What do you mean you can’t find it? You’re at it!

I am talking about the authentic Google “Subscribe as you surf” bookmarklet which is available in the “Goodies” store at your reader settings page. Things closest to the eye are the easiest to overlook – even for a team of very competent engineers and developers.

Oh, and by the way, when you visit my feed, feel free to subscribe to it.

WPZipper is one of those websites which I have thought of doing myself but never really got around to… Alright, well there is very little we all haven’t thought of and gone, “that would be so nifty!”.

What WPZipper does is pull your blog’s zipper to close shut.

Take 2: It is a service that provides a method for you to download one zip file with wordpress and all of your favorite plugins – plugins which you can select from a list. wpstar made it. I can think of a few small reasons I wouldn’t use it, based on the fact that unless you download it from wordpress.org, you can’t be sure it is the untainted, true, product, and that I might go over with the plugins I get. I manage my plugins, maintaining them at the bare minimum, and even for plugin-intensive blogs, it generally takes weeks to find the plugins I need and add them. But all this shouldn’t stop you from appreciating the effort behind the WPZipper service.

Did you know wordlog.com is using only the Akismet plugin at the moment? Now that’s what I mean by “bare minimum”.

Addendum: I must say I am also using the Automattic stats plugin now – so that makes it two plugins. D’oh.

A blog statistics plugin has been announced at wordpress.com that allows you to have wordpress.com style stats on your own, self-hosted WordPress blog. Before we go any further, he’s a glance at the simple, efficient stats that await you:

The Automattic stats plugin available at wordpress.org is real easy to install. You download it, upload the stats.php file to your “plugins” directory on your WordPress blog, and activate it. The plugin will then tell you that you need an API key for it to function. If you have an API key for akismet on your blog, you can use the same key for the automattic stats plugin. If you don’t have an API key, you can get one here. You just sign up as a user at wordpress.com and when you activate the account you get an email with the API key.

Once the plugin is installed, you get an extra tab on your blog’s Dashboard, and a half hour later, you can see the wonderful stats, the same way you see it on WordPress.com blogs. Besides the graph showing the number of visitors for the past month, there is also details regarding referrers, popular posts, popular exit links and incoming links – that’s about all the stats I care to know about my blog. Here’s the full screen of stats, for the Ubuntu Blog:

Now here’s the irony in all this: The day after I rekindled this blog, I decided that the first time-intensive article I will do would be one comparing different free blog/website statistics tools available out there. So I installed OneStat Free, Statisfy, Add Free Stats, HitTail, 103 bees, Extreme Stats, GoStats, ActiveMeter, StatCounter, Google Analytics, and Reinvigorate. I thought I’d let them run for a month or so and then compare them, all for your benefit. I left out mint, which seems to be popular among bloggers since I don’t have an expendable $30 to plunk down for fancy stats. I also left out sitemeter since they seem to add 3rd party tracking cookies that may violate browsers’ privacy.

I might as well go ahead and do it – in spite of the wordpress.com automattic stats plugin being available – I think there will be some merit in comparing these alternatives. I want to decide which stat tools to use and stick with them. Now I can compare these, and the wordpress.com stats, and find the best.

Please join me in wishing Mark and Sarah a happy married life. Mark has been a strong pillar of the WordPress community, with his Covered Web Services being the least of the ways he’s involved in the WordPress world.

Here’s wishing them joy and happiness, and a happy married life.

First off – sorry for the really, really old posts from this blog that are

all over the WordPress Planet. I (finally) upgraded this blog to the latest and greatest WordPress from version 1.5.x-beta. Yeah, you heard me right.

So here’s a quick recap for all interested. I am carthik, and I used to very active in the WordPress world until somehwere around January 2006, first as a support forum volunteer, back when Matt was “allusion” and Mark was “Podz”, and the other Mark was single , then as a documentor and then a little bit of development, which never really took off.

I am grad student, and at any point in the past two years, I swear I haven’t been more than 6 months from graduating Unfortunately I have had to stick around in school for a little longer than I expected. I did not like that one bit, and so I came up with a plan to graduate – reduce all other activities drastically. That has, variously, made me unhappy, frustrated, bored and cut-off so I gave up on that.

Yesterday I passed a significant landmark – completed a major portion of work, and I had promised myself that as soon as I get that done, I will try and make an effort to at least upgrade my blogs and think about their future. So here I am.

The last post on this blog was made in September 2005. Yup, almost two years ago – so many of you who read this might not know about this blog, or me. This was the first WordPress-powered blog about WordPress. There are so many of these around now that keeping track of them all is difficult. Some pander to professional bloggers (back in 2005 there were no professional bloggers), some to theme hunters, some deliver WP-related news and gossip etc. I have read all of them, and have enjoyed reading some of them. I now have to figure out what WordLog can do in the coming future. The audience I used to address has upped and left – or perhaps not. If everything goes according to plan, I will post here infrequently, maybe once or twice a week, with stuff that deserves attention. That equation has worked out quite well in the past at the other Ubuntu blog I write.

Lots of things have changed around WordPress – in fact I wanted the first new post here to be a summary of the changes around the WordPress family – but then I had to write this one now to explain the sudden planet-flooding. So see you around the planet, but no so much as some of the other regulars! If you are an old WP-friend, or someone who remembers me, drop by and say hi!

Richard Stellingwerf, author of the ClearLooks GTK+ 2.x Theme Engine is using WordPress to power his blog. Here’s what he had to say about the installation:

Well I decided it was time for a proper website. Since it would be mostly a weblog, I decided to try this thing called Wordpress.

As with anything in Gentoo, installation was a breeze. I had it up and running in a matter of minutes, and I haven’t got any wiser doing it. Automatic installations rule

Of course, on Gentoo, you can just emerge wordpress and be done with it.

Webloog.com has an interesting name, if nothing else, to it’s credit. Oh, and it is a service that provides users with a free hosted WordPress (WP-MU) blog.

So many hosted services now – you have no excuse not to try WordPress out, if you haven’t already

The University of Florida News site is powered by WordPress now. There’s also a podcast of the news items.

Via: hyku.

Update: the University of Paris and Saint-Denis also has a WordPress blog powering the Department of Sociology news page. Thanks Baptiste.

Yes, the Matt in question is our good old photomatt.

The title shocked me.

Matt seems pretty excited about his new stuff.

There was time when I knew most of the active WordPress blogs out there, by url. Then there was a time, not so long ago, when searching for WordPress in Google News would yield nothing, though I expected it to show up something, when I thought WordPress had “matured”.

The system has come of age, and the signs are showing up.

So WordPress now has a reputation of being “User Friendly”. Cool.

WordPress.com is a new venture aimed at bringing blogging to the millions. For starters, it will grow in a very linear fashion, since each invitee (yes, you need an invite) can only invite one other person. Powered by WordPress Mu, and much like Blogsome, WordPress.com promises to be interesting. May Matt’s baby steps to ensure the self-sustainability of the WordPress effort succeed. It is still not clear how WordPress.com, which, I hear, is not a non-profit, or voluntary organization, will make it’s money, but Matt has made it clear that most of the money generated will be used to fuel development and developers. Having known him for about two years, I trust him to not hurt his own baby, and what seems like a big part of his life – personal and professional – WordPress.

Alright, now to the juicy stuff – Lorelle has written about what one can and cannot do with a wordpress.com blog. For all of you who are still wondering what a wordpress.com entails, and what it’s all about, this extensive article tells the story.