Dev Team
General
The RSS feeds were slightly non-compliant in that they used a capital A in rdf:About where it should have been rdf:about. Fixed now. The code will be on the live server by the end of the day.
However, I haven't gotten a chance to implement the function to print all such archive pages.
Also, I haven't gotten a chance to properly test the recent additions so that I could upload them; they won't be available until tomorrow. I apologize; my DSL service was out for a bit and I'm too tired to go through all of the testing now.
I've just implemented a preference which allows archives to be written in chronological order, rather than the default reverse chronological order. It is more natural for people to read top-to-bottom than bottom-to-top (in most cultures), so this makes following a series of posts in the archive pages more natural.
It's not typical, which is why I'm not going to make it the default, but I can see a good argument for why a person would want it to be that way. So, now, powerblogs users have that option.
Well, I just added a feature to the update page which I hope will be more convenient. Now when you go to update a page, it optionally inserts:
<p class="update"><b class="update">Update</b>
</p>
To make it a little easier to add updates. It's configurable through the interface preferences page, of course; you can set it to prompt you (the default), automatically add this text, or do nothing (no prompt, no text added).
I hope that it's useful.
A small joke of mine was featured on The Volokh Conspiracy.
Unfortunately, the database that we use as a backend doesn't like it if you feed it improperly formed utf-8 unicode characters (i.e. it just chokes). I hadn't realized that people would generate these, but Stephen Nuño indirectly pointed out that this is certainly not the case.
The fix is quite simple, it's just a perl regular expression to replace high characters with their html entity equivalents (the numerical html entities are quite helpful for this). Now one can enter such characters into the title, name, and post, and there won't be problems. I just wish that I had thought of this earlier. Still, the fix is quite pro-active and should prevent any recurrance of this sort of problem. (I've also added it to the comments and trackbacks, as well.)
I finally updated www.powerblogs.com. There are now a few useful screenshots, and best of all I rewrote the features page so that it actually reflects the current featureset, rather than the featureset from 6 months ago.
Also, I put into practice what I learned at Kaplan about "people care about benefits, not features". Or I tried to, at least. Even if I didn't completely succeed, it's a whole lot better than it used to be. Now people will have a reasonable idea of what the powerblogs service is all about.
There was a nasty PHP bug in the signup code which added \ characters if the blog's title contained any sort of quotes. I eventually tracked this down to a "feature" controlled through the magic_quote_gpc variable in the main PHP configuration file. I suppose that this might do someone good, but turning it off did the trick. (The correct way to make your CGI secure is through using regular expressions to verify input, not by having your language automatically mangle every string that it sees.)
Tom discovered a bug in the stylesheet editing code where you couldn't clear out a user stylsheet. This is fixed now.
He also noticed that I forgot to include the code to set the variable which disables trackbacks. Fixed.
I also added a pair of features to comments:
The ability to require a name when posting without a comment account.
The ability to have every new comment emailed to the author of the post (useful for staying on top of comment discussions when you have very low comment volume or very long comment sunsets).
Finally, I changed the font size on the menu to be 12px instead of 14px. This should help it display properly at smaller resolutions.
Now it's time to rewrite the features page to reflect the features that the software actually has. If you read it now, you'd get the impression that at best it had a meagre comment system, when I think that it currently has the best blog comment system available (I'll admit that I'm biased, but even so I think that it's true).
In my last post I talked about some cool new features which I was working on. Well, they're here.
All comments and trackbacks can be moderated from a central location (so you don't need to follow every post in order to moderate every comment).
It is possible for commenters to register accounts, and such accounts can be optional or required for commenting on posts.
There is an option to require comment accounts to be approved by you before they can post.
If you allow guest comments, guests cannot use the display name of someone with a registered account, so if you get a flood of foreign commenters (because, for example, some popular blog linked to you), the invaders can't pretend to be any of your regular readers.
Also, I re-implemented the back-end storage for posts to use the same postgreSQL database that comments and trackbacks are using, for improved speed, flexibility, and scalability. It should now handle tens of thousands of posts as lightning-fast as it handles ten.