indiboi.com

Major changes have been happening over at Positive Fusion, but I’m going to take a moment to bitch about Internet Explorer.

Okay, it’s the same ol’ same ol’, really. One writes valid, semantic, xhtml… IE butchers it all to hell.

This a truly odd situation. We expect css to render differently in broken browsers (IE - on any platform, really), but Firefox is rendering with a one pixel error between Linux, Windows, and Mac. I think it might have something to do with the fonts in Linux though.

This is how the set of tabs is supposed to look, and it appears this way while using Firefox under Gentoo Linux.

This is how the set of tabs looks under Windows and Mac in Firefox. I adjusted the css to illustrate how it appears, versus taking a screenshot in Windows, so the fonts are the same, obviously, in this image.

It’s just one of those funny situations where I need to decide if I want it to look right for me, or for everyone else. ;) That is, unless I can fix it, which I’ll certainly attempt to do. I think I just need to totally rethink the implementation, bugger.

It started to bother me when I mentioned my code being valid xhtml 1.1, but the entries themselves may not be. I decided to fix that too, since I was off the deep end in terms of correcting things with the journal.

I went through every monthly archive from December 2000 through until this month, validating every page for xhtml 1.1. I then proceeded to correct any errors either through MySQL queries or by hand. It took several hours, but somehow I think it’s worth it, in a way. I realized that a couple little mistakes within some entries caused massive display errors for the entire page, whoops!

Yes, it’s determination, but it’s also procrastination. I think it’s useful procrastination for once though. I really need to start working on the company’s redesign project; I stalled it back in August. It’s supposed to be done, uhm… next month? I took the time to really work on my skills, so hopefully it’ll go much better than I had been expecting. :)

My journal has several scars from changing content management systems over the years.

The first, being the one that most annoys me, is the ridiculously high post ID numbers. This is because blogger used those weird ID numbers, and back when I ran the ‘blogger to b2′ importer those ID numbers were caught, from then on any B2 post ID was up in that insanely high range; that also includes the GM entries I also imported thereafter. I vaguely remember talking to Michel about it and I think he changed the importer script to prevent that. The damage was done for me at least, because by the time I’d talked to Michel about it I had already been using B2 for quite a while.

We did devise a method for correcting it, sort of. The problem was that the post ID was stored not only within the posts table, but also the comment table; fixing it would then disconnect the two. What seems like forever had past and after working on eMotif I learned how to fix the problem.

You might ask why I still have the insanely high post ID numbers, right? The little issues of permalinks, search engines, and all the other problems that radically changing the post ID would bring are the main reasons I never did change it. So, with that being enough of a discouragement, I still have the seemingly random post ID numbers, it’s also a little funny to me, because it appears I have 2.8 million posts.

When I ran all of my uber-complicated MySQL queries to convert my ’specialized’ B2 to WP, I never realized that WP has this post2cat table. I don’t really understand the purpose of it, other than to be annoying, why couldn’t the category field in the post table be sufficient?

I started noticing this really strange category error on the posts that used to be made private in the old b2 style. The error was very vague, but given the entries it appeared on I finally realized it had to do with my conversion. The only way I discovered that I could fix it was to open the post via the edit function and save it again. I realized that running queries on the posts table to change the categories wasn’t working, which didn’t make any sense. I finally, this morning, noticed the post2cat table, it was missing entries for all of those entries showing the error. During the import/conversion/upgrade from B2 to WP those particular entries had a category ID of zero; the conversion process never made entries in the post2cat table for them!

It was with that realization that I knew I’d be having tons and tons of problems with the next part of my little project. No longer could I easily, using MySQL queries, change the category of a post simply by searching the content and using that as a trigger. Obviously it is possible, but the query becomes more difficult because post IDs need to be cross-referenced, the category needs to be updated on a table completely separate. I made a decision that doing things the ‘quick’ way would require far too much time spent learning how to do it; I manually edited all 280 or so entries, opening and then saving them. Yes, it sucked.

Blogger and GM did not support categories, thus all entries were put into ‘Journal’, since it was my number ‘1′ category. This bothered me for quite a while, but changing the categories was such a time-consuming task that it didn’t bother me enough to want to do it. Enter mass category editing, I found the plugin, but using it is so painfully slow that it really isn’t quite worth it. That’s the problem when you have 1500 posts in the same category, that plugin wants to load every single one of them, and for some reason the script doesn’t stop at the number of posts you choose to display if you also choose to display a category; it needs a serious rewrite for usability.

Notice that I said usability. I should be even a bit more precise and say it needs to be more user-friendly. It takes the same query strings that WP itself takes, but this is not overtly clear; once I realized I could feed the month string it became considerably more functional for me. It very well may say that somewhere, I just didn’t notice it.

All of this highlights how critical I am of my writing now. Somewhere along the line, even though I’d always called this my journal, it really did become a journal. I did before that, actually, have a blog. Random and useless garbage was posted on a regular basis, it’s sad. I’m not condemning the format of a blog, contrary to how it appears, I am though recognizing that my style of writing has substantially changed over the years. I am clearly a better writer than I had been. I care much more about the quality of what I’m writing in contrast to the quantity of posts.

Why is all of this important? This journal is an accurate history of my life over the last five years, it is my most valuable document. I’m determined to correct mistakes I’ve made while maintaining it in the past by using the skills I now have.

It took most of the day, as the true upgrade procedure isn’t exactly seamless by any means. I also had a lot mysql queries (that I’d worked out previously when I made my first half-assed attempt) to run so that things actually synced between my heavily modified b2 and wp’s ‘fun’ new way of doing things.

Thanks to the ability to moderate comments, well, I think I’ll turn them back on — for various newer entries only — until I get sick of the idea again. I’ve done just fine without any comments for an extremely long time, and I’m not going to bother displaying how many there are anywhere on the index, just because I’d get sick of seeing ‘0′ all the time. ;)

I did, finally, decide to slash apart some of the more idiotic code, I just couldn’t deal with the stupidity and lack of… variability. I still think, for all the ‘good things’ that have come with wp, so many of the ‘good things’ that made b2 what it was, a simple and effective tool, have been torn out.

The decision though was ultimately if I’d write into my hax0red b2 some of the things that I wanted or I’d tear out the crap from wp… tearing out the crap is easier. So, here were are, with all the spit and vinegar required, I’m using wp.