Have you looked at Conversant? It’s weblogging software built on top of Frontier that is a competitor of sorts to Manila, and it’s still in active development.

I have looked at it, but it has one major problem — it’s still built on Frontier, and as I’ve mentioned, one of my big problems is with the actual Frontier application crashing. Thus, if I’m moving, I’m going to get the hell off of the platform; I’m sick of a company that doesn’t give a crap about fixing bugs, and unfortunately, Conversant is built on the main product of that company.

This product sounds intriguing: The open source Roxen Web Server, which has a built-in scripting language called Pike and a MySQL database integrated into the server. I haven’t had a chance to try it out, but the description sounds like it offers some of the same things that made RU/Frontier attractive to me.

You can calso check out http://www.larkfarm.com/wlm/roll_your_own.htm for a couple of suggestions….

Jason, would you give a run-down as to what functions were causing you the most trouble? There might be some work-arounds the various Frontier developers could suggest. The script meridian folks might be able to help out:

http://community.scriptmeridian.org/16603

Bill, my biggest trouble is that Frontier crashes intermittently, and I have no clue what is causing the crashes.

Interestingly, I am running a version of Frontier.exe in which Doug Baron (prior to being let go) specifically included debug symbols, and I have a great Dr. Watson log of one of the crashes that contains all the debug symbols relevant to the crash. What’s the problem? I sent it to Doug, who had been let go; I then sent it to Dave, who asked me to work with Dann Sheridan; I then sent it to Dann Sheridan, who never got back to me about what it showed, and then was let go. At that point, I sent it to John Robb and Dave again, and after persisting with another request, was told that I can get no more support. (Doug Baron, after being let go, told me that he is almost positive that it crashed during a database save — which, to me, means data loss — but when I repeatedly asked Userland directly about that, I got no answer.)

I’ve also had a few other issues that were never dealt with, all of which I could fix on my own but watch get wiped out with the next release of a script via overnight updates. (Example: using Radio to edit Manila directories, which was reported in January, at which point I was told to back off because the focus at the time was on Radio as a web server. It’s probably an easy fix, but it’s within Userland’s script territory.)

The funny thing about all of this is that I used to not mind having to frequently restart my Frontier server after a crash. Last year, though, I started working on an application in the hospital that I wanted to run on top of Frontier, and after showing it to other hospitals and other departments at my own hospital, there is a lot of interest in the application. I realized that these people — mostly doctors without much computer training — wouldn’t be as tolerant of restarting the app, so I starting trying to work with Userland to track down and fix the crashing bug. Once fixed, this would have easily translated into a couple dozen more Frontier licenses a year for Userland. Now, who knows.