Opensource filePro Project

[Brian K. White]

----- Original Message ----- 
From: "Fairlight" <fairlite at fairlite.com>
To: <filepro-list at lists.celestial.com>
Sent: Monday, May 05, 2008 9:54 AM
Subject: Re: Opensource filePro Project


> Y'all catch dis heeyah?  Jose Lerebours been jivin' 'bout like:
>> I believe that they can make money in many ways:
>> 1. Donations from a potentially super large community
> 
> You're joking, right?


You are of course absolutely correct right down the line. Not merely holding a pessimistic outlook.

However I do see one aspect that works in Jose's favor that hasn't been mentioned yet.

I don't know how true this is of anyone else, but I know I, at least while in my work decision maker hat, am far more willing to invest in something that I get to have than in something I don't get to have or merely get to rent.
Ie: if fp were free, more importantly if fp were unencumbered, I'd happily contribute to the cause of keeping it unencumbered.

Even though I'm still not yet using PuTTY in main production, I am more pleased with the $2k I've spent on that than any of the several other $2k that we've spent on FacetWin licences. You get so much more bang for your buck paying for oss development. I got 3 major features implimented, one of which someone else has _finally_ implimented out in the free community (file vs registry based configuration storage). I still need 3 more smaller things done before I could use it in main production for all users, but even so, what I do have, I have as many copies as I want of it, forever, no license or permission issues installing on as many desks as I want, no slightest bother or risk of legal issues about delivery method, no consequences or impact from playing musical servers, just backup/restore/clone/sync/run/whatever anywhere, any time, no one ever has to enter a licence key and no user ever faces a non-working app. There is no calculating what that's worth. It's like trying to calculate the value of learning to read vs paying someone to read for you or paying rent on a reading machine or something. Or maybe paying for a license to use English as the language for reading and writing for one person in one context, etc..

At least when you are merely chipping in to an already existing and useful product. I'd hate to contemplate the investment it would take to start even a modest app from scratch and with no substantial pool of developers developing for free and only one entity paying developers.

Also, when an app is free for the end user, then even all the moochers are valuable, since they find bugs that you could never afford to test enough to find yourself, nor could you safely claim any particular level of buglessness except when a large number of people have actually used your thing without problems.


At the end-user level, I have donated $5 to $35 to a few things here & there that I realized I was getting especially a lot of value out of or that were especially important to me or my living. Recently I dropped $30 in the bucket for "IEs4linux" when it not only worked, not only worked at something that native windows XP failed at, but did so faster and easier than it would have been in native windows had it even worked there, even though it was running a 32 bit windows app inside wine on a 64 bit version of linux. To me that was exceptional and was certainly valuable to me considering the thing I was trying to run was work related and really needed. (the client portion of an IP console device, which is only provided in the form of an ActiveX applet served up by the devices built-in web server.) And I buy linux magazines at least partly to get the ever-updated dvd's of common distros. I figure a small piece of that goes towards general linux development via various indirect routes. Perhaps I am rare, but I can't beleive it. I may be one in fifty or a hundred, but not one in a million.


None of this necessarily means open sourcing filepro would launch fptech into a higher class of existence. You're still right that it's almost certainly fantasy land mystical thinking, about the same as what all the original dot-coms had before reality burst them all. "A million people will each give me $1" I think filepro would receive mostly a lot of ridicule from the community at large and the bad press from a few loud mouths that get off of criticizing things would cause great masses of more or less clueless end users to ridicule and reject it in turn, without ever having used it. Not saying that any of the negative points they would raise would be wrong, but that that they would only bother talking about the negatives. There are so many things that have to be apologized for when explaining filepro to a new user today. A new user, used to todays features and todays level of polish even on freebie one-man-development-team apps, sees filepro as something like the data base that's written in crayon.

Then again, if it were open sourced, perhaps many of those same annoyances would get cleaned up by users who are bothered by them, and THEN you'd be somewhere.

A lot of them seem like they should be mostly just tedious to do, not conceptually difficult. Maybe Ken doesn't have time to do them all, and none of them seem worthy of fptech assigning Ken to do them compared to other things, and so they never get done, but maybe they are merely time consuming grunt work that others wouldn't mind doing simply because they can. Even I could do things like, "find all uses of function x and make the necessary adjustments if it had 2 extra parameters and one parameter changed from an 8 bit to a 32 bit integer." or "find everywhere this hard coded assumption is made and convert it to use this variable or dynamic function."  Or at least I could go some distance into that and reduce the task to some smaller number of tricky problems that are worthy of a better coders attention. And even if I did it poorly, it's still easier for a better coder to look over a patch and spot weaknesses than to have created the patch. 
I know I often see shell script examples that I might not have wanted to bother concocting from scratch, but when I see the example someone else did and it's poor, it's easy to fix that up and I often do so.

Or maybe a lot of these wishlist things might be easy enough to do but a company that is selling a product and assuming some implied responsibility for data safety (regardless of any waivers) simply can't take chances and must do lots of testing after even minor changes just to be sure, and can't afford it, but an open source app would not have that problem. Users use whatever version they want from the oldest to the latest beta or cvs, and simply report any problems without anyone being at risk of being crucified or liable for damages or sued for negligence etc.

I think that the user base for fp has dwindled by now to the point where if were open sourced, there would probably not be enough of a user community to produce much development of the core product. Not enough users are also good c coders that out of that pool there would be very many c coder hours spent on hacking on filepro. And fptech probably couldn't survive on selling tech support and contract development.

So if we wanted the core app to continue to progress we'd probably best accomplish that by some number of interested parties chipping in to pay a few developers salaries.

That raises many questions I dont' have answers for. How many of us would chip in to that? How much would we do? How much would it take? Who would determine what gets worked on? And how? What would Ken's (or someones) life be like with 57 boss's all claming the right to tell him what to work on because they each contributed $2k this year?

I actually would contribute cash to a thing like that, but I don't know how I'd feel about steering development, or not having enough voice to steer development. I would probably be ok with contributing the cash in turn for a "seat" in votes on such matters and then probably be happy to go along with whatever the overall votes resulted in even if my pet peeves don't get worked on. Most wishlist items sound pretty agreeable to me even though I didn't come up with them or haven't encountered the problem they aim to solve. I'd just consider it paying for the app to be open, if that makes sense. I'd happily pay for the app to be free, not because I don't want to pay, but because I can't stand the artificial limitations imposed by licence managers and copy protection schemes. We do not live in a world that can tolerate those things any more. I don't have "a server". I have something more like an "environment" which is spread over an always changing collection of servers, physical locations, and net connections.

Though, I have heard a few ideas, periodically even, that I think are actually bad or not well thought out and should not be done because they would actually damage the product. And I don't trust the masses to always correctly identify and reject such broken ideas either. The recent discussion of the way fp's date math function should work re "last day of the month" is a perfect example. No way in the world should adding one month to feb28 result in anything but mar28. I think in that case the collective would have decided the same way too. That's a pretty easy one to get. But I know sometimes issues come up and it seems like only a few people fully grasp the problem or the implications. Those are times when I'm glad fptech are comprised of people who do understand the issue and are in a position to simply decree "I'm sorry you don't understand why, but it can't work any other way than that, and we are not going to change it to work any other way than that"

-- 
Brian K. White    brian at aljex.com    http://www.myspace.com/KEYofR
+++++[>+++[>+++++>+++++++<<-]<-]>>+.>.+++++.+++++++.-.[>+<---]>++.
filePro  BBx    Linux  SCO  FreeBSD    #callahans  Satriani  Filk!