Please create an account to participate in the Slashdot moderation system


Forgot your password?

Slashdot videos: Now with more Slashdot!

  • View

  • Discuss

  • Share

We've improved Slashdot's video section; now you can view our video interviews, product close-ups and site visits with all the usual Slashdot options to comment, share, etc. No more walled garden! It's a work in progress -- we hope you'll check it out (Learn more about the recent updates).

Linux Business

What is the Best Way to Start a Paid GPL Project? 231

Posted by ScuttleMonkey
from the will-food-for-code dept.
pooslinger writes "I know little to nothing about programming but would like to start, fund, and maintain a GPL linux POS application. I see there are a few available with the majority being closed source. I am currently starting a business and really despise the fact that I will have to spend $2-$5k on a proprietary solution. I would like to create an application where you could take a midrange PC, connect inexpensive touchscreens, barcode readers, thermal printers, credit card readers, etc; scan/input inventory; and begin selling. Something like a Debian POS distribution that boots into X and starts a POS terminal. Does something like this exist, am I just trying to reinvent the wheel?" How have other people approached starting a new GPL project, finding talent, and ensuring the code choices best benefit the community?
This discussion has been archived. No new comments can be posted.

What is the Best Way to Start a Paid GPL Project?

Comments Filter:
  • First off, you really need to check [] or [] first. There there are plenty of POS software projects listed at both. Find one that looks like what you're wanting to do and hasn't run out of steam, and give it a shot in the arm with some cash. Maybe spread your cash around two or three of them.

    That said, the question of how you start and attract talent to an open source project... I'm not professor on the history of open source, but the most successful projects I've seen are ones where a coder or small group of coders put out an alpha of their project and it was playing with the alpha and seeing the possibilities in it that got people excited enough to come on board and start pushing things forward.

    So, if you're not happy with any of the POS projects you can find on SourceForge or FreshMeat, and since you clami to know "little to nothing about programming," I'd suggest going over to eLance or RentACoder and spend a good chunk of your seed money on getting an offshore firm to build your alpha for you. While they're coding their hearts out for you (they'll want 2-3 months to work on your contract), take that time to get to know the open source community and how people launch their open source projects.

    Then, when your offshore coders come back to you with a decent alpha, pick an open source license (BSD, GPL v2, GPL v3, etc.), and use the knowledge you've picked up in the prior few months to get the word out and spread the code around. If you did your homweork well and spread the word well, that seed you planted may well sprout.

    But remember this, a strong open source project needs a strong leader who can handle the big picture outlook, keep all the volunteers in line and focused on the goal, and drive the project forward. You're going to have to approach some strong personalities one-on-one and try to recruit that project leader. Without a strong leader, failure is a definite possibility.

    Just my $0.02.

    - Greg
  • by Anonymous Coward on Friday October 05, 2007 @02:52PM (#20871639)
    i am going to go out on a limb here and say a person should
    be able to code this up in a week or two depending on the
    guidelines given to him/her.

    did you already design, draw out, and etc how this POS system
    is supposed to work? if not, you should realize that is most
    of the work. programming it is rather easy.

    it seems to me, if your business cannot afford 2000 for a basic
    kit you may want to rethink your business model. and that 2000
    includes the hardware.. so come again on how much the software costs?

    another slashdot article that is yawnnnnn
  • by hardburn (141468) <> on Friday October 05, 2007 @02:56PM (#20871699)

    In terms of opportunity cost, you'll likely spend that same $2-5k making a custom solution. Also, realize that the modern POS has over a century of lessons learned about securing cash registers from theft (particularly employee theft). You'll want to find developers who have specifically worked on POS applications before, or you won't benefit from all that knowledge.

  • by OverlordQ (264228) on Friday October 05, 2007 @02:57PM (#20871717) Journal
    I'd recommend the above except for going to RentACoder/eLance for an Alpha. I can almost guarantee you that anybody who wants to contribute to a GPL project will absolutely hate having to figure out whatever spaghetti code that the bottom-dollar code shop spat out.
  • Don't (Score:5, Insightful)

    by geekoid (135745) <> on Friday October 05, 2007 @02:58PM (#20871733) Homepage Journal
    ""I know little to nothing about programming but would like to start, fund, and maintain a GPL linux POS application."

    It will be a nightmare to maintain, and won't be fit for general consumption. It probably couldn't even be used as a framwork.

    Programming take training, and POS invlves understanging issues you haven't even thought of.

    What you could do is Fund one. It will cost more then a couple of grand.

    There are several approaches to this:
    Hire contractor, pay them for there work, open the code. This gets you something running, and once there post it and ask for contribute it.
    You might be able to get several small business to pitch in to the POS fund.

    You could get some students looking to write a thesis together to get you going.

    You could get some professionals to do it on weekends in exchange for equipment they get to keep. Or perhaps your business produce something you can use to trade.
    --I fall into the category.

    In any case, define what you want as specifically as you can. Don't do it in a language that can't be cross compiled.

    Organizing project would be a great help to. That is what slow or stops a lot of projects. No one to organize or follow-up.

    Please don't write it yourself without some training. I have worked on many system buyilt be very smart people with no training, and they all sucked.

    Use your energy to manage the project until it gets to a point where it is usable to you. That includes allowing or disallowing contributers who want to contribute any functionality you didn't originally have in mind. Let them fork it, but don't get caught in function chasing.

  • by RobertB-DC (622190) * on Friday October 05, 2007 @02:59PM (#20871747) Homepage Journal
    I am currently starting a business and really despise the fact that I will have to spend $2-$5k on a proprietary solution. I would like to create an application where you could take a midrange PC, connect inexpensive touchscreens, barcode readers, thermal printers, credit card readers, etc; scan/input inventory; and begin selling.

    You say you don't want to spend as little as $2,000 on your POS terminal? You can't buy a business-ready PC and the touchscreen for that price! Have you even priced those components? Try Froogle: $500 and up [] for an LCD, which you want unless you're operating in a cleanroom. As for the PC, sure you can get a consumer-quality box with wirez sticking out for $500. Is it designed for mission-critical 24/7 uptime? Or is it likely as not to fail under load. Do you have all the possible software installed on it to prevent hacking of your customer information? I don't think you can get those two components alone, in the application you're using them for, with less than $2k.

    Don't think you can cheap out and get everything you need at Wal-Mart and Craigslist -- you're running a business, not a hobby. You want to spend your time making money, not tweaking equipment. You don't want to spend $5,000, but what's the cost of making your customers stand in line while you try to figure out why your hacked-together hardware and software doesn't Just Work? Whether you're running a dollar store or selling overpriced speaker cables, you can't afford the downtime.

    Spend the money on a system that works out of the box. If you're too cheap to do it right, then please comb your hair into a stylish point and congratulate yourself: welcome to Management!
  • Support? (Score:5, Insightful)

    by Major Blud (789630) * on Friday October 05, 2007 @03:07PM (#20871877) Homepage
    If you want to get an open source POS package, I think the most important thing to keep in mind is support. When the package breaks, are you going to be able to contact the coders for help? (Sure, it should be written to never break, but let's be realistic.) You may know enough to fix it, but what about your employees? They won't necessarily be able to get you on the phone when they need to. I used to work for a company that produced a mediocre POS package, and the amount of support calls we used to receive was insane. Everything from hardware, software, training questions, networking....we had it all. Point is, make sure that you have someone waiting on the phone for you when stuff happens.
  • Lucky for him, he wants to develop it to use it in his company, not to sell it and make a business out of it. This is something a LOT of larger small businesses could get behind, if promoted correctly.

    That said, there isn't much difference between this and the browser-based kiosk solutions that are also available.
  • by Lumpy (12016) on Friday October 05, 2007 @03:17PM (#20872019) Homepage
    I see plenty of POS systems running Vista, just no Point of Sale systems running it.

    honestly all the good stuff nowdays is running Embedded Xp or other embedded system and does not have a PC in it wasting money but a smaller SBC doing what it needs for the Point of sale job all in a nice stand monitor with card swipe on the side and sitting on the cash drawer.

    although the article's author did not look hard at any of the systems out there. if you want the bottom feeder in cost Quickbooks has a turnkey system, just ad cheapo pc and thier box of stuff to it and you have a POS for under $2000.00 per register.
  • by mr_mischief (456295) on Friday October 05, 2007 @03:18PM (#20872041) Journal
    The OP isn't counting on selling the POS software as his business model. He's opening a business that needs POS software and doesn't want to drop $200o to $5000 on a proprietary solution. That was stated.

    His interest is apparently in using $2000 to $5000 to pay other people to do a GPL-licensed POS software system so his money won't be locked up in some unresponsive closed-source POS software vendor's accounts. He's trying to be a good business and OSS citizen by competing on the core of his business and cooperating in the portions that are ancillary and supportive. The POS software one uses is rarely a competitive advantage in retail. Pricing, customer service, marketing, location, and potentially how you tie your POS and warehouse systems together are much more important than the POS software itself.

    Of course, supporting the software might turn into a secondary revenue stream, or it might be the kernel of a start-up for someone else.
  • Live and let live (Score:0, Insightful)

    by Anonymous Coward on Friday October 05, 2007 @03:19PM (#20872063)
    "I am currently starting a business and really despise the fact that I will have to spend $2-$5k on a proprietary solution"

    These guys crack me up - he wants to start a business where he would charge other people for something useful for them, yet he despites the fact that some other people also sell something useful to him. Hey you don't have to like the price, but to say "despise" is really a short-sighted way of thinking!
  • Re:Support? (Score:5, Insightful)

    by gbulmash (688770) * <semi_famous@y[ ] ['aho' in gap]> on Friday October 05, 2007 @03:38PM (#20872311) Homepage Journal
    If I had mod points, I'd mod parent up.

    My first IT job was managing the POS system for my dad's restaurant. Given this was 1989 and everything would crash and burn if the dot matrix printer jammed during the nightly reporting run.

    Still, with all the niggling little problems, that whole thing would have bricked within a month without a support contract.

    My favorite little quirk of it was that when you logged into the system for the day (it was also the time clock), at the end of the process, you pushed the "print" button instead of the "enter" button. That was because at the end, it was supposed to print your daily ID code.

    That was very non-intuitive. So, one day I get calld down to the restaurant floor from the admin offices upstairs. One of the terminals has locked up. One of the wait staff tried to log in, but the machine keeps giving them this error code that they've filled up the screen with. Half of the wait staff and even one of the cooks is at the terminal, trying to figure it out. To show me how the error message keeps coming up, they hit the "enter" button a few times.

    I say "remember, when you're clocking in, you hit the 'print' button at the end, not 'enter'." I hit the "print" button, the screen clears, the waitperson's daily code is printed, and the terminal is back to normal.

    Remember that POS is a mission critical, live-fire production system. If it crashes or starts hiccuping, you're looking at lost money, lost productivity, etc. You're looking at half your staff gathered around it, giving unhelpful suggestions and asking dumb questions until the person who knows how it works can fix it. Thats why, even with 20+ open source alternatives, closed source flourishes. No matter how good the open source project is reputed to be, if there's no local vendor who can provide timely on-site support, there will be a lot of businesses who want nothing to do with it.

    - Greg
  • Re:Don't (Score:3, Insightful)

    by Aladrin (926209) on Friday October 05, 2007 @03:47PM (#20872419)
    One correction:

    Programming take training, and writing a POS invlves understanging issues you haven't even thought of.

    Even if he funds one, like he's saying, he'll still be the driving force for it, and all the specs will come from him. There are aspects to programming everything that seem simple from the user/admin point of view, but are anything but simple from the programming point of view.
  • by zotz (3951) on Friday October 05, 2007 @03:50PM (#20872469) Homepage Journal
    "The article poster is about to discover a harsh reality of the open source model: if you give your software away, profit-making businesses aren't going to pay for it unless there's something else in there to sweeten the deal and the software is just a means to that end."

    I would not bet on this. If a piece of software is central to a business, they will want reliable support before commiting to that software. A smart business might just hold off until professional contracted support is available.

    "If you're expecting to make money just by developing and supplying open source POS software, you've got the wrong business model..."

    Right, unless you go into the bespoke and paid up front angle.

    I also still feel that there is hugh untapped potential in Association funding.

    National Retail Merchants Association? Local Chamber of Commerce? National Locksmiths Association? How much could they benefit their members if they charged $5 extra for membership and used that to fund Free Copyleft (GPL) programs that would benefit their members / industry?

    There is money to be saved with this idea and the old saying of "a penny saved is a penny earned" still makes sense.

    all the best,

    drew []
    Underwater Fun
  • Don't be an Idiot (Score:5, Insightful)

    by Slashdot Parent (995749) on Friday October 05, 2007 @04:50PM (#20873275)
    Don't be an idiot. You're trying to start a business, yet you are creating a monstrosity of a barrier to entry for yourself. When you're in business, you do what you're good at (what makes you money) and you buy what you are not good at. You said yourself, that you have no idea how to run a software development project.

    If you do "roll your own", you have no idea how long it will take to build this thing, what the quality will be, whether or not it will interface with your accounting software, what the response time will be for breakage, what it will ultimately cost, and probably about 100 other things that neither you nor I are thinking of right now. In the meantime, you are losing money.

    On the other hand, you could buy a package and be up and running tomorrow.

    Buy a QuickBooks POS for $800 and get on with your business plan. In five years you'll be able to start a charity open source project.

    As someone who started two successful businesses, I can't believe you even asked this question.
  • by Qbertino (265505) on Friday October 05, 2007 @05:15PM (#20873593)
    I have to second the parent. I've programmed commercially funded GPL E-Learning Systems and built a few small business support systems entirely from OSS components. $5000 dollars is an absolute minimum to get an ERP or CRM system set up, configured and extended in a usable manner. It won't be any other with even the most frictionless of POS setups.

    If you have somebody competent you can trust then you can kickstart the first components of someone who wants to build a small POS system to start his own POS software business (maybe a student or so). But that's an extra load on top of your job of building up your own business so I'd be triple carefull before attempting that.

    I too strongly recommend you do some research on shrinkwrap POS systems and then ask a reputable local OSS savy freelance programmer / SMB IT consultant how he would automate your business and what it takes to implement some glue-scripts for automation and data-migration to bringe the gaps between bill-printing, the ledger and whatever other shareware you piecemeal your first IT enviroment together with. The programmer and the job(s) he needs to do shouldn't initially cost more than $2000 in total and deliver measurable speed up of your IT pipeline. A good programmer with experience and consulting skills will - in the first round - speed up a mom'n'pop shop business by up to ten to twenty weekhours and bridge the worst gaps in IT on a relativly small budget. If you go that way, you can also see if your IT specialist is for real or just a wannabe. And you won't risk to much either.

    If your business grows he'll know enough of yours to extend IT accordingly and you'll both know what problems to look out for. On the second or third iteration of your business relationship you can start thinking about funding an OSS project with your favourite programmer as a funded project-lead. All else is too early.

    Just a free advice from an E-Lancer and OSS Consultant.
  • by Archangel Michael (180766) on Friday October 05, 2007 @05:27PM (#20873737) Journal
    Most POS software is tied or at least linked to Accounting Software. To me, POS is only a subset of Accounting. Payroll, AR, AP, GL, CRM, POS, Inventory ...... all are integrated with each other, or at least should be.

    A small proprietary storefront may only want POS at the moment, but in six months of success will change that. Then you have to change POS software because it doesn't tie in with the new inventory management software. Six months later they'll want CRM, then payroll, then AP, then .....

    Which brings me back to my point, it isn't as easy as it looks. I'd like to see an open source modular accounting system that didn't suck. Only install the modules you need, where each module stood on its own and/or GL module.

    Then tie it in with a Web Interface and online shopping cart .....

    I'd love to be a project manager for such a challenging project. CRM, Accounting, POS, Online shoppiping cart .... all tied together yet independent. Good luck.

  • by ClosedSource (238333) on Friday October 05, 2007 @05:29PM (#20873751)
    Clearly if the primary business is retail than it would be much smarter to pay for a proprietary POS (or use an existing OSS version) than starting a new OSS project. The latter approach is going to take a long time to get to the level where it can be used effectively for a business and there's a significant risk that it never will.

    He should get the core business profitable ASAP and then if he has money left to burn, he can hire people to create any OSS product he wants - there's no particular reason why his choice should be limited to applications related to his "day business".

    The bottom line is that starting a software project (proprietary or OSS) to support some aspect of a new business is not a very good idea unless the software is a core part of the business and isn't available from anyone else. Unfortunately in that scenario the software is likely to hold most of the value of the business and so making it OSS will probably limit profitability in the future.
  • by pooslinger (1167631) on Friday October 05, 2007 @05:49PM (#20874019)
    I've found quite a few thermal receipt printers that offer linux drivers and they have documentation on how to get it working with CUPS.
  • by Anonymous Coward on Friday October 05, 2007 @11:45PM (#20876491)
    I can speak directly to this problem

    About 2 years ago, a friend and I were asked to quote on a project by a small businessman (he had about 8 shops) we knew but we couldn't get the numbers to work for us, while also making it cheaper for him. (BTW both of us know POS pretty well and have worked on commercial developments in the past.)

    The basic problem is volume. The proprietory vendors (there are only a few left) essentially give away the software and make money on the hardware and software support. (And the software is _very_ feature rich these days).

    So if you're being quoted at 2-5k per store, you're probably getting a reasonable deal - at least compared to paying for development yourself. Although you speak of "low cost" hardware, I think you need to look at how quickly the costs mount up.

    For example, a single station will require PC, touch screen, cash draw and receipt printer, plus optional barcode reader. You're already pushing well over $1000. If you then want to integrate your bank's POS, they'll want to test and certify your hardware and software setup to ensure that it doesn't screw up their payments system. This is not particularly onerous and they'll help out, but it is time consuming.

    For one or even a few stores its not worth doing. What the existing vendors are doing is amortizing their software development costs across many customers, getting volume discounts on the hardware and taking a margin there; and lastly making their real margin on the annual support and licence fees.

    Having said all that, it's not clear to me what you really want. Do you want someone to a.) develop a new FOSS POS solution from scratch, or b.) integrate and install it for your own business?

    You could support a.) better by donating or participating in one of the existing projects (check sourceforge and google I can't remember what they are now). Starting from scratch will take too long and cost too much.

    If b.) is what you want you might do better to partner with a local developer team - provide the hardware yourself and pay them separately to do the integration and support; allow them to take the software away but insist that it all be FOSS so that they don't tie you in to long term support contracts. If you do that, then you can reasonably expect to negotiate a low fee for the development work as they can make the money back on subsequent work and support with other people.

    b.) was the option we offered to our friend, but he wasn't interested in the hassle as he wanted more of a turnkey solution. If you're prepared to do it, then more power to your arm sir, you certainly have my support.

  • I'm glad (Score:5, Insightful)

    by Slashdot Parent (995749) on Friday October 05, 2007 @11:52PM (#20876539)
    I'm glad you're not an idiot, but even smart people do idiotic things from time to time. As for me, I'm an idiot who every once in a while does something clever. We all have to live with our limitations.

    At any rate, I'm going to have to stick with my original advice. It's great and altruistic and all that you want to create a whiz-bang open source POS solution. Your heart is in the right place, but your head is not.

    Wait until you see a lot of black ink on your financial statements. Wait until your business requires only 40 hours per week from you.

    Starting a business takes everything you've got. I know so many people who tried to start this or that type of business and just get bogged down in some irrelevant project. I know one guy who was so convinced not to hire an accountant that he got stuck researching the optimal type of corporate entity to use, then he got stuck learning basic bookkeeping, then he got stuck learning tax accounting. Meanwhile, he never made a single sale. Ever.

    I know a lady who started a jewelry business and she had two lines. Her first was wildly successful, but she decided to pour everything into a second line while neglecting the first. Oh, sure, the company is still in business. But it is bleeding red ink everywhere. It's only afloat because her husband is a partner in a law firm. I can't believe he hasn't pulled the plug on this yet. It's easily a 6 figure loss for them.

    I could go on forever. The point is, concentrate on what makes you money! This POS idea is going to take you thousands of miles outside of your comfort zone, and just as many miles off your business plan. It is going to cost many, many thousands of dollars. You will not be able to hire a few Indians to hammer this out over a weekend--cheap coders need perfect requirements and specs, which you will never have. Anyway, you do not want your business running on software that was hammered out by a few Indians over a weekend with ambiguous specs.

    Lastly and most importantly, I wish you the best of luck with your business. And since it sounds like your mind is made up to get mired in this POS POS (ha ha get it?), I'll wish you the best of luck with that as well. Just promise yourself you'll gauge its progress with ruthless objectivity and pull the plug on it quickly. ;)

1 Sagan = Billions & Billions