Two different types of acquisitions at play here - 1) services (for a site like Blog Flux), and 2) original blogs we acquired (eg Money Crashers).

This post is about the former - I’ll have one for the latter soon.

As I wrote in a previous post on re-write or assimilate?, acquiring a site is not a fun job as most of the code is rather shabby. But Blog Flux adds an extra dilemma - scaling headaches.

Commentful was written by Kailash Nadh, a smart fellow going through university. He writes some interesting code, and I often hire him to write snippets of code. When he showed me Commentful, I was very impressed - I had no interest in the other comment trackers (who tend to be rather bloated) - this was simple and it worked like I wanted to. It took a while, but we eventually acquired the site.

So the code he writes is good - it isn’t a headache to go through it. One thing Kailash didn’t do was for the possibility of tracking 100,000 URLs (something we could hit within a week depending on uptake). The simplest example is www.techsoapbox.com/some-post - if you and I are both tracking it on commentful, it will check the URL individually. Furthermore, (when we originally acquired it), its throttle wasn’t optimal (eg after 5 days of no new comments, you need to slow down how regularly you check for new comments).

The other issues here was the setup - we have our own Blog Flux login, and internal table structure. Porting it over wasn’t a simple exercise. So we decided on a re-write.

The re-write served two useful purposes:

1. It got the programmer upto speed quickly on how things work. He had the code to compare against, but any bugs in the future are much easier to tackle - there is afterall a ‘cost’ involved with the programmer having to understand how the old code worked.

2. Ability to revamp UI and add features. While Kailash implemented many improvements when we originally acquired it (making it from a so-so product to a useful), there were still a few things we didn’t like. These have been all dealt with. Again, while there was a time-investment in the re-write, there was also the time-investment in the programmer getting acquainted with the code. Furthermore, the programmer has gained extra experience in remote-url manipulation and Firefox extensions - both a positive play for future developments. While he could have read the code, actually writing it out is the real experience.

There is one further headache - importing data from Commentful.com to Blog Flux. A pretty simple process - once Blog Flux Commentful goes live, shutdown new users and new urls. Copy database data to Blog Flux server. Ask user on Commentful.com to visit a specific page on Blog Flux (need to be logged in there, otherwise register first). Enter your Commentful.com login, and the data is instantly transferred. As painless as it can get.

If you want, you can beta test it here. It isn’t ‘released’, so play at your own risk. It also tracks much more than blogs (that list itself has to be updated to include entries like Blogspot and other forum software).