> If you go with "windows only" you loose all of the developers unless > you pay them. > Very few will work on Windows for free. > > It's actually very easy to make portable multi-platform code > > One way to seriously speed up Kicad development is to leverage the > efforts of other Open Source electronic design software. Let's say > the a sophisticated design rule checker gets written for the gEDA > project. Wouldn't it be nice if it could check Kicad projects? An > expedient way to make this happen is with netlist translators. A > better way would be to come up with a common format for netlists. > This is the way to go. As of now the Open Source EDA community is > split between three or four projects. > > About paying programmers. Some hobbyists will work for $100. No > profesional would. not unless he swaps hats at night. There are > project that offer "bounties" for features but I'm but sure those > small suns help much. But there are many, many examples of open > source moving very quickly once some one hires a programer and puts > him on salary. This gives him what he needs which is TIME. If I'm > working 50+ hours a week at work writing code. Offering my $200 to > work more is not going to work because there is no time. Look at the > big Open Source projects. All of them have programmers on salary.
I am speaking as a KICAD user now. I am troubled by the state of the Kicad project. The package is within reach of being adequately usable for any board I might ever attempt. But it is not there. This situation teases my mind to find a way to cross that last bridge. Chris you make good points about raiding other free code. That could lower required man-hours, even without common data structures. But even with that approach, some of which is being done now within Kicad on the new zone support, there is not enough getting done to satisfy me. Am I the only one not satisfied with pace of the project? It seems like with 6,000 users, if each user contributed only a dollar, resources could be gathered to fund most any reasonably sized "work order". You mention bounties and salaries. The salaried individual is working for a corporate sponsor presumably. Do you see that happening here? We need to explore the "work order" or bounty idea further. Maybe a little out of the box thinking is required. Website automation would help, but tell me if this is easy or not: 1) Discussion is done on this list to create a "work order", usually an enhancement. 2) Someone posts the "work order" on a website built to support this new business model. 3) Developer(s) bid on the work order, with each bid listed seperately. Now you have one work order with multiple bids attached to it. 4) Users can tentatively commit with credit card to funding any or all of the bids. The credit cards are not billed until the trip point for any of the bids are reached. The trip point is when enough tentative commitments have been submitted. Only the first bid to be tripped fires the credit card billing. At that point the other bids are dead, and their tentative commitments are voided. On the bid that tripped, maybe a partial payment of the funds go to the developer, with balance of the funds held in escrow until the work is completed. Would not a mechanism like this turbo charge the Kicad project? Maybe within 5 years this could be the best software in the world. Within 1 year you'd have no reason to use any other software. It is capitalism, without the middle man: the software company. Yes you can find holes in the idea, but it is a starting point and demands refinement. With a handful of known developers, it is not like you are taking great risk with your $20, which is about what a reasonable "tentative commitment" might look like. You know the developers from their previous work in most cases. Great developers don't work for free. Therefore, great software ain't free. Sometimes it looks free, that is, when somebody else has already paid for it. But there's not enough of that happening on this project.
