Flutterby™! : right... well, not there.

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics

right... well, not there.

2004-06-02 16:05:39.176534+00 by Dan Lyke 12 comments

I have thought a couple of times about my involvement with the whole "weblog" thing, how there are a few funded companies now running software that's not a whole lot more complex than what runs this site, how my lack of a reasonable install script sunk adoption by other people, how basically I have a small cool site while a few other people, some of whom started playing with this stuff later than me, have careers based on it. Rafe did some similar musings yesterday.

[ related topics: Dan's Life Weblogs Software Engineering ]

comments in ascending chronological order (reverse):

#Comment Re: made: 2004-06-02 18:17:52.777807+00 by: meuon

I gave up after spending an hour trying to get it run on a generic RedHat box..

Next time I need/want such a beast, I'll do it in PHP/MySQL, and have considered such as a combination Blog/Album system. Your stuff is too advanced and Dan-ish for me without some help.

#Comment Re: made: 2004-06-02 21:05:36.055348+00 by: Shawn

I primarily haven't had the time to experiment with the Flutterby code. I'd love to set up a blog, but I no longer have shell access at my ISP (and for the moment I'm not gonna move 'cause it's free) and I'm not sure if I can set it up under those circumstances.

I do still plan to see what I can do about setting it up on my internal network for evaluation and possible packaging. Perhaps this summer, as I'm taking a quarter off from school, but it'll have to compete with several other projects.

#Comment Re: made: 2004-06-02 21:42:24.147195+00 by: John Anderson

I'd still love a chance to play with it; I'm just seriously over-commited, timewise. One of these days...

#Comment Re: made: 2004-06-02 21:47:45.617603+00 by: Dan Lyke

That wasn't at all a plea for people to check it out, in fact at this point I'm thinking once again of trying to glom on to a more successful open source project for future development, merge my must-have features into some other package and continue development with that. Because really, to make it useful to you guys I have to do a bunch of work that isn't directly applicable to my needs; it needs to support another database or two, it needs better design and some set of schemes and skins...

I guess my point is that the Flutterby CMS code is a lot like a replay of Hyper!: Right general concept, right technologies, on the right wave, and yet... not quite there.

#Comment Re: made: 2004-06-03 00:04:27.855973+00 by: ebradway

The real problem is that the Flutterby CMS code reached a point where you were comfortable with it. You needed to have a monkey try to install it and work with it in order to make it palatable to the masses. But the reason you started the Flutterby code wasn't to make new software - it was to express yourself. That's why so many people flock to the big blog sites. They don't have to deal with the software - they can just focus on expression.

#Comment Re: blogging software made: 2004-06-03 00:28:34.317576+00 by: mvandewettering

Frankly, I think the field is still wide open for a decent consumer level blogging solution. Most blog software isn't so much a solution as a toolkit: you are expected to twiddle, tweak, prod and hack your way to a blog, instead of just designing one. If word processors worked like most blog systems, you'd be forced to write your own pagination and hyphenation routines, and you'd still have to do all your links by hand.

I recently discarded MovableType and shifted my weblog over to Wordpress, and while it isn't *bad* it isn't actually all that great either. Layout and information are combined in various PHP files in an entirely haphazard way, which leads to the unfortunate problem of requiring you to patch php files by hand when you perform a Wordpress upgrade. Nothing like this is ever going to become widely popular with the Joe Averages of the world.

I've got some ideas, but no time. When Incredibles wraps, perhaps I'll have more time...

#Comment Re: made: 2004-06-03 02:52:21.693894+00 by: Shawn

Oh... my... g...

Did I read that correctly? A Flutterby poster who is also crew on The Incredibles[Wiki]? I do believe I'm feeling light-headed. (No, I'm not really surpised, when I think about it. But it's still giddiing to be two degrees from somebody working on such cool stuff.)

#Comment Re: made: 2004-06-03 03:59:33.684262+00 by: Dan Lyke

I can think of 3 occasional participants here who I'd guess will have credits in The Incredibles[Wiki].

#Comment Re: Incredibles made: 2004-06-03 04:17:39.084407+00 by: td [edit history]

Yup. Me and tdl as well as Mark. Ooh, and doesn't Sam post sometimes?

#Comment Re: made: 2004-06-03 13:16:30.744131+00 by: Dan Lyke

I wasn't sure about Sam, he uses this server for other things though.

#Comment Re: made: 2004-06-07 16:56:57.700067+00 by: other_todd

Not sure you'll read this several days late, Dan, but I've been away. Anyhow, you've seen the Utopia with Cheese history, and you know how long it took me to do the badly needed version 2 of it. One of the things that enabled me to finally get off my rear and add the features I wanted was admitting to myself that I was NOT, never had been, building a system for eventual packaging and use by some other web site, that I was working on it just for my site (even if other people eventually post there - it has the capacity for that). One of the things that had been on the to-do list for ages was to create an installer for the thing, to make setting it up a little easier. That was also the first thing on the to-do list to get axed.

The real decision, in short, is whether you want to try to be a weblogging "service," whether you want to be a software provider (even if for a very small user base) ... or whether you are doing this primarily for yourself or your friends. I suspect you lean toward the third choice, as I do, in which case what you have on the back end doesn't matter a whit as long as it keeps you and your friends happy.

#Comment Re: made: 2004-06-07 21:18:10.096096+00 by: Dan Lyke

Part of my wanting to get other people using the Flutterby software is that I think that there are some cool features in it, and I think parts of it are modularized in good ways, but I'm not seeing where to go further with it. Unfortunately in the process of building these things the switches and ugliness have crept in to where the code isn't all that easy to read any more, but to make a conceptual leap to figuring out what I need to be doing differently I need someone else to have some needs for the code.

I don't want to be a weblogging service. With a little design skill (something I don't really have) it'd be pretty easy to turn this into a multi-blog thing; the internal structures support it, the code is structured to allow mod_perl to take good advantage of shared blocks.

It's fine as a hobby, but I think there's some features here that most other weblogging tools could use, mostly the ability to make sense of text and HTML and keep end data from causing security holes, but largely just to clean up parse trees and be able to output valid XHTML. The number of comment systems which still let HTML go through unedited scares me, but even the good ones don't do enough about enforcing tag hierarchies.

As a weblogging product, the platform requirements are too restrictive, and at least one of the mod_perl optimizations causes issues with running multiple versions on the same server. I'd have to remove some of the features I've pushed up into the database back-end, and clean up the database install, and...

So really, what's useful are some of the Perl modules, and now they're becoming dated.

Really what I should do is tell everyone "it's no longer supported", and write version 3, have a set of SOAP[Wiki] calls that run all of the functionality, a minimal web based front-end for y'all to use, but push most of the interface (including some kick-ass ideas for contextual hypertext editing that still haven't seen code) into some "web services" type apps or off into NNTP and similar clients, and see if I can enlist someone with come CPAN to figure out how best to repurpose the parsing code into something that fits reasonably into that class hierarchy system.

And use XSLT[Wiki] where appropriate.

It's just that since I've failed to get anyone using this system (and I had about 3 or 4 users of version 1 of the system) I'm reluctant to do that without doing the social engineering differently, and I really want my next hobby that's a time sink to be something with heavy commercial promise. Unfortunately, in that realm I'm reluctant to switch to Microsoft .NET as a development environment, which means that anything that's not to be deployed on the server side of the web stuff will be minimally useful, and I see weblogging tools as a lever to get into the CMS space, which is an area that'll be won based more on marketing savvy than anything.

So I guess what I'm really coming to, at least in terms of this code base, is that I should get everyone else to switch off this system (I think Todd still has something set up on my server). And then I can see if one of the open source weblogging tools has the basics of what I need and will take some of my enhancements to comment security and such the right way.

That'll let me go chase some other hobby entirely.