Categories
Housekeeping

Missing (but not lost completely) comments

Late last night–as I was googling various keywords on my blog, testing the links that came up, and changing the slugs on the ones that didn’t forward properly–I discovered yet another issue: missing comments. Some posts that I knew to have comments on them now have none.

I checked the posts at cosleia.blogspot.com, and the comments weren’t there either…meaning they were not saved in Blogger’s database. I can only guess that something happened when Blogger first changed the way it handled comments on post pages. I’m pretty sure this happened before the October 2008 shift to embeddable comment forms, but I’m not sure exactly when. I wish I could nail down a date, because then I would have an idea of how much work is ahead of me.

You see, the comments aren’t lost. They existed on the static html pages on my server, which I saved. So what I can do is go through and manually copy and paste each group of comments into a new comment at the end of each post. It’s not a perfect solution, but there’s no way I have the time or patience to individually post each comment, and I’d like to retain the original timestamp information, even if it’s not listed correctly in the database.

When Blogger first enabled comments, I had to copy over existing comments from LiveJournal, Xanga, and Blogkomm, so this is nothing new…but I get the feeling the task is going to be monumental this time.

I suppose what I will end up doing is starting at the very beginning of my blog and checking every single post to fix the following errors:

  • Missing comments
  • Slugs that include the words a, an, and the (infuriatingly, sometimes these words are included in Blogger post URLs, so I will have to find a way to continue using Google to check the links)

While I’m at it, I will tag and categorize old posts that never got this treatment in Blogger.

Then I will move forward to posts starting in November of 2009. I’ll save and upload all Blogspot-hosted images to my server, then change all the affected img links.

This is what I get for having a blog with nearly 4,000 posts, I guess.

Update: So far I have found posts as recent as 2007 with no comments on them! Ugh.

Categories
Housekeeping

Some solutions, more problems

I have solved the problems of redirecting from /journal and removing the .html extension by using 301 RedirectMatch. It was really a matter of trial and error, especially getting the .html extensions to go away without affecting pixelscribbles.com/index.html, but I think I managed it. If you see any weird behaviors or post links that don’t work, let me know. (Of course, I still haven’t solved the problem of Blogger’s post links not including the articles a, an, and the.)

I became aware of two new problems while using Google’s cached search results to test my settings: archive pages and “label” pages. Blogger archive URLs were formatted like /2004_07_01_archive, while WordPress’ are the much nicer /2004/07/. Similarly, Blogger keywords/tags, called labels, had page URLs formatted like /labels/travel, while WordPress’ tag pages are /tag/travel/. I think I can write a couple more RedirectMatch statements to fix these problems.

I do wonder if all the RedirectMatch-ing is going to hose my server. Is there another way I should be doing this? I guess in time search engines will map the new addresses for all the posts, and then I can remove the redirect expressions…right?

So now my issues are:

  • Removing the articles (a, an, the) from WordPress URLs
  • Redirecting old archive page URLs [done]
  • Redirecting old label page URLs to tag URLs [done]
  • New: Replacing spaces with dashes in tag links
  • Changing image links from Blogspot to my server*

* Update: I had thought that the images were all on my server already and I just needed to change the links. It turns out that all images uploaded after November 3, 2009 are saved to Blogspot and do not exist on my server. So I will need to upload those images and then change the links…manually.

Categories
Housekeeping

The shift has been made…sort of

If you’re reading this, you can perhaps see that I’ve switched my blog to WordPress and also moved it from pixelscribbles.com/journal to plain old pixelscribbles.com. Both of these changes have been long in coming; I’ve thought about doing them both for years but just never had the motivation. Blogger’s ending of FTP support changed all that, and last night I slapped everything together in something of a rush.

This is not the template I’ll be sticking with, though I do like it. I’m hoping to design my own. This, of course, will take time; in the meantime you may see this blog cycle through various looks. My first order of business, though, is to see if I can’t do something about a few lingering problems.

My old post URLs don’t work anymore. This is sad, since many of those URLs had been around for years and years and years. I’d like to fix this problem, but to do so I will have to cope with three things.

  1. I need to redirect links including /journal/ to the root directory. [done]
  2. Blogger removed articles (a, an, the) from post-title URLs; WordPress doesn’t do that.
  3. Blogger permalinks had .html at the end; WordPress permalinks don’t.[fixed with redirect]

I’m thinking I will need to do some sort of mod_rewrite to solve these problems, which is a bit beyond my knowledge at this point.

The next issue is post images. My posts have pictures from three sources: this server, from back when I uploaded and embedded everything manually; SmugMug; and Blogger uploads that were sent via emails from my phone. The first source isn’t a problem; I simply left all the images on the server as they were in /journal/images, even though I don’t use /journal for anything else. Similarly, all the SmugMug images are still where they were before, so no need to change anything there.

But the Blogger uploads are quite the issue. When I was publishing via FTP, all the images saved into a folder on my server called uploaded_images. It would have been ideal had I left things this way, but alas, last night I was in a fit of “good grief I’m tired of looking at Blogger’s ‘We will no longer support FTP’ notice”, and I impulsively moved everything to Blogspot. It was only after that shift was done that I realized I really, really prefer to have my blog on my own server, where I can control things like .htaccess.

Unfortunately, Blogger’s migration tool was quite thorough, and it copied over all the Blogger-uploaded images and changed the URLs in the posts that referenced them. So now I have 1200 posts (apparently) that are pulling images from Blogspot.

I’ve left cosleia.blogspot.com up for now, so I shouldn’t have any broken images for the time being. Comments are disabled over there; that blog is a relic that maybe I’ll delete and maybe I’ll save for fun. Regardless of what I do with it, though, I need to somehow get rid of all the Blogspot images over here. I’m guessing this is going to involve a lot of manual searching and editing, but who knows? Maybe there’s a plugin out there that can help.

So that’s the state of pixelscribbles for now.

Despite all the pressing issues above, I’m thinking my first moves really ought to be to restore my blogroll and webcomics [done!] lists…I miss them!

Edit 3/16/2010: I’ve solved one issue with a redirect. And webcomics are back, baby.