Ahem…Is this thing on?

Sorry for this totally superfluous post – I am moving Edtechpost off of MoveableType on to WordPress (finally) and a new URL, http://www.edtechpost.ca/wordpress/. If I did this the way I hoped, you should still this post in your current subscriptions, through the wonders of mod_rewrite. Feel free to update subscriptions and blogrolls, but I’m hoping this won’t be totally necessary. The whole reason for the move – mainly to get Comments turned back on. I was getting killed by the spammers using MT, so much so that almost 2 years ago now I disabled them and never got a solution working that satisfied my ISP. I know, lame. Well here’s hoping the magic of Akismet and other WP solutions keep comments up and running. This ain’t no soliloquy, you know, at least not intentionally.

16 thoughts on “Ahem…Is this thing on?”

  1. Yeah, my first comment! Thanks D’Arcy, both for the comment and for confirming indirectly that the mod_rewrites seem to be working. You would not believe what a hassle that was to figure out, not the main stuff, like the rss feeds and index page, but all the individual archive pages. But figure it out I did, so all of the URLs to the old edtechpost should now come direct to he proper post in here. Anyone, please let me know if you have problems!

  2. Damn! I’ve been waiting expectantly for this day, my mouse-trigger-finger at the ready, hoping against hope to be the first commenter on the mighty EdTechPost. Of course, the D-Man beats me every time.

    Nice new look — congrats!

  3. Nice new look and I’m glad that comments are back on – welcome back to the whole blogosphere 😉

    BTW, since I installed Akismet in March 2006, it’s picked up 56,000 spam. That means that spam accounts for 99% of my traffic 🙁

  4. Thanks Harold. I agree, the ‘whole’ blogosphere absolutely includes comments as well, and I have been remiss for too long in not fixing the problem. Mea culpa. Thanks for not shunning me 😉 I’m hoping Akismet works as well for me as it seems to have for you and others, though the whole ‘blog spam’ thing still feels a bit tenuous, and do wish some better solutions would come along.

  5. Yup, the old RSS feed still works. Which means I can comment, finally. And so I can ask you what you meant one or two posts ago about my links.

    OK, so you don’t like downes.ca/cgi-bin/page.cgi?post=4

    I can understand that.

    I’d like to get it to: downes.ca/page/4

    But I don’t know how (I know, I could learn, if I read enough user manuals…)

    What I don’t want to do it: downes.ca/whatever/2007/01/24/some_long_post_title/

    Why not? Because they’re too long. If anybody has to type them, they’re miserable (happily that doesn’t happen a lot). They work very badly in printed documents. And they’re too wide for the text version of my newsletter (or anyone else’s text newsletter, for that matter).

  6. Stephen, that’s a good point about the shorter URLs being type-friendly, but the primary key based URLs aren’t very human-friendly. If I have a bunch of links to check out, and they differ only by some arbitrary (to me) number at the end, I have no idea about the content of the link until I open it in a browser. With the long(er) ones, I get an idea of date and context without having to open each one to see. I’ll settle for a downes.ca that’s back up and running, without much real preference to how the URLs are constructed 🙂

  7. Hiya Stephen, I saw after I made that comment that the OLDaily feed was working again which was great to see. I’m sure I’m not alone in having missed it the last while and as I mentioned before, definitely commisserate with the suffering of migrating hosts and server software.

    So while I can see some of your point about the wordy wordpress links, I personally don’t mind them, they do have some sense to them (I like the fact that I can detect the freshness of the link right in the URL).

    I think my confusion about your permalinks is that while you may have followed the letter of the law for RSS 2.0, your feed is the only one (out of the 175 I subscribe to) that works the way it does. Maybe that’s because your feed declares itself to be RSS 0.91 in the head when in fact it *has* to be RSS 2.0 (there’s no GUID in RSS 0.91, for instance). Maybe that’s why aggregators don’t treat it the same, I don’t know. I think it’s also an issue with me figuring out how you are doing permalinks (which, you have to admit, you’ve changed a few times in the last few years). Now that I realize that in your feed, what for the rest of us is known as a ‘permalink’ shows up as the link to ‘Comment’ I will know what to link to when I try to link to one of your posts.

    I am distinctly not trying to start an RSS pissing match. I appreciate the fact that by coding your own software you have been able to blaze new trails and do stuff you couldn’t find with other packages. That’s great. I guess my statement back to you would be, I don’t think I’m an unsavvy user, and every time you revise your software I am left trying to figure out where your #$@%%$# permalinks have gone, as I want to *simply* be able to link to your posts like I do to everyone else. Cheers, Scott

  8. Thanks for the tip Gabor, I actually saw on your site that you were running it. I will use that as the next line of defence if Akismet starts to fail. Let’s hope not.

  9. Darn, I thought you were going to show me how to make links like http://www.downes.ca/post/24241 out of links that look ugly. *pout*

    Anyhow, I hear you. Part of the reason for all this is that I felt (and still feel) that the blogging engines did linking wrong, and not for good reasons. In trying to drive traffic to themselves, they made the link element always point back to themselves. I had never done it that way, using the link element instead to point to what I was talking about. You did the same, and other than me, you were the last holdout that I know of.

    Anyhow, the problem is, was, and always has been that there is only one link element in RSS. So there’s no way, programatically, to identify both your own post, consistently with the way the blogging engines do it, and the link you are talking about. All that was made even more complicated when I worked on Edu_RSS, because I wanted to rack all the relationships between posts, and none of it was done in XML – it was all in the (often CDATA escaped) content (or description, or wherever).

    In my own system, therefore, I created two data tables – ‘post’, which was intended to store my links, and ‘link’, which was intended to store everyone’s links (including mine, which were harvested along with everyone else’s by Edu_RSS).

    I started out by linking to ‘post’ by identifying the table and the format: ‘db=post&format=html’ along with the ID number. Too long and clunky. So I revised that to be: post=id with html a default format. I intend to support the old style (Google still mostly lists the old style links) but I think that’s broken at the moment. The post=id is indended to be permanent, and even if I create some other format I will continue to support this format.

    I also had a link called view.cgi with simple article numbers, which was developed originally for NRC internal reporting. That is still supported and will always take you to the post=id page for the appropriate article.

    Meanwhile, for links, there are two options: to either view the listing on my site, or to be redirected directly to the link. edurss02.cgi?rd=id is the redirect link, and this is a permanent link (it’s the only thing edurss02.cgi does now; I’ve move all the functionality to other files)

    page.cgi?link=id is (or will be) working to view the listing on my site. Not much there; it used to support the ‘conversation’ feature but the Yahoo search engine made a mash of that, hitting the conversation links every few seconds (I think there was something about them that attracted Yahoo).

    When I was required to move my site from RC I thought about dumping the whole link / EduRSS database entirely (who wants to pay for the server, the bandwidth, etc?) but there’s so many links out there I couldn’t pull the plug.

    There’s also all the links for user comments over the years, but that’s a whole different story. The post=id format is permanent for them too (I can’t believe that it took me almost 10 years to realize that the only difference between a blog post and a comment is a ‘post_thread’ column).

    Anyhow, what’s the upshot of all this?

    Simply this:

    – the page.cgi?post=id format is my permanent system – I may abbreviate it with URL rewrites or whatever, but I’ll always support it

    – that said, I will always support the old formats. You don’t need to change any of your links. Some of the code is broken at the moment, but over the last few weeks I’ve been watching my 404 file and catching all the odd formulations that come in – that’s the upside of writing your own code.

    The upside of all this move, too, is that it has caused me to revisit my code again, and to really clean it up. Now that I’ve had a much closer peek at thinks like Drupal I have come to realize the value of my own code. I’m not trying to replace Drupal or anything like that. But I do think I can express some concepts more eloquently in code than in English – if I would trust myself to do it.

    So despite some of the drawbacks – such as the hesitations and false starts described above – I am going to keep working on this, and right now I am rewriting all of my scripts with a very clear eye to making them publicly available.

  10. oh I just realized you were talking about the content of the ‘link’ element in the RSS, which is currently pointing to the external site.

    That was a temporary measure. I did it when I was on CSoft because the server was staggering under the traffic.

    I will be putting the downes.ca link in the ‘link’ element again, so people will go to my site when they click on the title. I will be creating a separate link, [link], in the CDATA pointing directly to the external site, for people who can read my email or RSS but who cannot access downes.ca (like the people in Iran) or who don’t want to access downes.ca.

    This change will be soon now – maybe even tomorrow (with everything else going on, I had forgotten about the temporary measure).

  11. So glad that we can now comment – and good to see another WordPress user. I’ve also got Askimet installed, though clearly I’m not attracting the same level of spam as Harold.
    In the last couple of weeks, the spam has dropped right off, we’d had some form of attacks on our server & they tried to blame me (it wasn’t; it was someone else). However, in removing my blog & then re-enabling it; they’ve done something (I need to ask what) – which has meant that I’ve barely had a single spam since, though legitimate comments still seem possible.

Comments are closed.