From my last post, you know that I am slowly migrating this blog from Typepad over to an independent Movable Type installation. The problem is that Typepad does not want to promote moving off their service (completely understandable) and therefore has little incentive to allow users to create 301 redirects. Some research suggests that a zero-second meta refresh will be treated by search engines as a 301 redirect, so I set out to test this.
First I had to play with MT template code to see if I can generate the permalinks as they would be on the new blog site, and I have succeeded. The hardest part was importing the old posts over on this installation such that the basename could have the same 15 character cut off point as does my Typepad account. That solved, I could put a link on each Typepad page to its corresponding page here. After about a month, I see that a couple pages here are starting to get a tiny amount of search traffic. Nothing huge, but the important part is that the engines see this site and begin listing it.
After patiently waiting until I saw that result, I have now implemented the zero second meta refresh on each individual post page over at my old Typepad blog. I will be watching my search engine traffic via Hittail in order to see the cutover occur in real-time, if it works at all.
In the meantime, I will be posting here from now on rather than on Typepad, at least for this blog. Stay tuned to see if the 301 experiment works out! In any other scenario, I suppose such a test would be crazy, but in this particular case, there is no other way to do a migration off Typepad than to create a meta refresh, so even if I don't get the search results transferred over, we will have learned something useful.
First I had to play with MT template code to see if I can generate the permalinks as they would be on the new blog site, and I have succeeded. The hardest part was importing the old posts over on this installation such that the basename could have the same 15 character cut off point as does my Typepad account. That solved, I could put a link on each Typepad page to its corresponding page here. After about a month, I see that a couple pages here are starting to get a tiny amount of search traffic. Nothing huge, but the important part is that the engines see this site and begin listing it.
After patiently waiting until I saw that result, I have now implemented the zero second meta refresh on each individual post page over at my old Typepad blog. I will be watching my search engine traffic via Hittail in order to see the cutover occur in real-time, if it works at all.
In the meantime, I will be posting here from now on rather than on Typepad, at least for this blog. Stay tuned to see if the 301 experiment works out! In any other scenario, I suppose such a test would be crazy, but in this particular case, there is no other way to do a migration off Typepad than to create a meta refresh, so even if I don't get the search results transferred over, we will have learned something useful.
