Slashdot Mirror
Open Source Point-of-Sale - What's Out There?
aztektum asks: "I have taken on the task of designing a building a networked Point-of-Sale system for a friends new restaurant/club. We have looked at out of the box solutions, but their upfront costs are a bit staggering, so I suggested a DIY approach. We are going to buy hardware outright, probably using Elo touch-screens and basic white boxes. It's the software that we're coming up short on. We are looking for a system that has good back-end management (running reports for end of day, from a central location and other such features), has a flexible/customizable UI, and as a bonus doesn't suffer from too much proprietary lock-in. Since Elo's screens have Linux support, I wanted to see if anyone on Slashdot has worked on similar projects and has experience with open source POS solutions. I have Google'd a bit and found some options, but I cannot spend a lot time testing out multiple setups because of a firm deadline. Quite bluntly, what's the best but least expensive option that you know of?" This is not the first time this particular issue has appeared on Ask Slashdot. The last discussion on this particular issue was way back in 2003. What has changed since then?
I know of solutions out there. They're all web solutions. You might not be looking to implement web solutions, however, isn't it possible to set it up on a local network? Maybe this is your strategy but I'm a bit confused as to what you're willing or desiring to do.
I'm also confused if it has to be "open source" or merely "free." On one hand the title says OSS but you seem to be only concerned with price. Just make a note that "free" and "OSS" are two different kinds of software. If you googled, you probably found a smattering of both. The two I suggest you investigate are purely OSS as I prefer that over free. Your last question ("Quite bluntly, what's the best but least expensive option that you know of?") seems to imply that you're willing to spend money if we can verify it's worth it. What conditions are you working under specifically?
Look into MozPos & freepos. They are web based and maybe you could install a very simple sleek version of Linux with merely Mozilla. The homepage would vary between computer and you could have a pop up keyboard or swipe strips for authentication. You would probably need a swiper that supports Linux but that's another issue.
If you're truly looking to get out of the box open source solutions, look into using websites instead of networked applications.
If you're willing to get your hands dirty, I encourage you to look at the engines some of these OSS things are based on and maybe make a GUI network application that is built on top of a layer of abstraction allowing you to use an already designed backend (database & reporting features). This is a lot of work but would be a great addition to the community if you subsequently made it available since Elo screens seem to be the leader in touch screens.
With a web based OSS application, you can download the source code and change the PHP/JSP/ASP/whatever interface code so that the HTML suites your liking considering the specifications of your Elo screens. That's why I suggest a web application and that's why I think that you'll be most satisfied if you pick one in a language you know and invest two or three solid weekends in tailoring the interface to precisely what you want.
Most importantly, change the system based on feedback from the workers.
Just a thought, I encourage you to post your choices and results here on Slashdot when you do make a selection and attempt to do it yourself.
My work here is dung.
Too bad everything rings up as "free"
"This is not the first time this particular issue has appeared on Ask Slashdot. The last discussion on this particular issue was way back in 2003. What has changed since then?"
That F/OSS still hasn't admitted that vertical apps aren't it's strong point.
It depends on how much you value your time, inventing something from scratch, building it out of disjointed components, or supporting it when things start failing.
Most importantly, think what happens to your friend's enterprise when you disappear from the scene. Will there be manuals and instructions to rebuild and restore everything? Who will do that, and how fast? Will the replacement hardware be available, and how soon? Things like that tend to stop many a DIY project, once you realize what the hidden costs and risks are.
Given that restaurants and clubs already have some serious price tags, I do not think you should dismiss COTS solutions just because of their absolute price. It may be less than 1% of your friend's other costs. There is such thing as "cost of doing business" and a POS is part of it. I'd tell the owner to pay the man and live happily ever after - unless you want to carry the burden of tech support whenever a PC goes down. Most normal owners just call the manufacturer (IBM etc.) and have the till swapped out within a few hours.
...if you give me a few hundred bucks first ;-)
This one should work for most needs. It's multi-platform, and works with any currency.
There exists no way of exchanging information without making judgments. --Bene Gesserit Axiom
But I know of osCommerce, which is a pretty popular storefront solution... and then if you do a custom solution, I have a PHP/MySQL database manager that can be used for the site's backend. Not sure how useful it would be for your site – I originally wrote it for MadTux, a Linux download site, and their needs are admittedly more than a little more specific than most sites would need – but may as well offer it if you have any use.
Creative misinterpretation is your friend.
Being a "former" employee of a credit card machine/POS company.. you might not have much luck creating this piece of software.
o ns/resellerprogram/processorlist/ ) should help for more info on that..
Mainly in part to a majority of POS software having a protocol that you pretty much have to pay for to use as code in your software. It's a standard created by MCard/Visa.. a basic way to how the transaction as carried out as a user going to a website on the net.
When you register a POS software, it needs to have hard-coded a TID/MID.. which is how the processor uses to identify a) the merchant of the account, and b) the actual device being used to do your transactions.
There might be another way around this. I am not too sure. Your best bet would be to contact a major processor/platform (Omaha, Nashville, Nova, etc..) This website ( http://www.authorize.net/solutions/resellersoluti
Otherwise than that (bullshit licensing, standards) you should be fine..
Good luck
http://www.bananapos.com/pos/home.html perhaps?
The open source POS scene sucks ass. I'd say go with BananaPOS if it's something so generic. I had to make my own in C/GTK+
The viral properties of the GPL will make any product you sell also covered under the GPL! Also Richard Stallman will visit and demand a GNU/Double Bacon Cheese Burger and then want it for free! Then he'll use the bathroom and not wash!
or RedHat, I noticed they were running RH at their POS last time I went there.
The banana POS (http://www.bananapos.com/pos/home.html) software has been around for a while, but does not seem to be developing at a rapid pace. Maybe it's already mature. Check it out.
Welcome to the land of the free...pay toll ahead...no photography...please open your bag...
http://sourceforge.net/projects/tinapos looks interesting!
We played dungeons and dragons for 3 hours.....then i was slain by an elf
I work in Point of Sale for restaurants and I am an open source / Linux advocate. Unfortunately, restaurant POS is a complicated thing and there are no open source solutions meet my needs. There is just too much money to lose when computers dont work at a restaurant. A restaurant POS doesn't just count the cash transactions. It also handles employee timecards (payroll info), processes credit cards and gift cards, displays output to a customer (pole display or drive thru verification), sends orders to kitchen (on a screen or through a printer), stores recipes for mixed drinks, keeps track of tables and guests, allows waiters to transfer checks, does all calculations for closing out each waiter and bartender, and a lot of other stuff.
I doubt your friends want to deal with the problems that you will encounter when you experiment with software that is not stable and is not fit to run their restuarant.
There is a Linux based restaurant POS, but it isn't free. I think it is called ViewTouch.
The POS that we use at work is called VersiTouch (versitouch.com). It is pretty solid. Compared to the numerous retail POS software packages I have had to support, VersiTouch is great. I prefer the DOS version, which they try not to support anymore (the DOS version is called RCS). The Windows version has hardly any additional features and it requires the constant headache of maintaining numerous windows machines. I use the mars-nwe Netware emulator on a Linux box to share files with DOS clients.
I think the biggest restuarant POS software is Aloha. Positouch is also kinda popular. Both of these are windows-based.
I have tried to configure an Elo touch on Linux, but I could never get the pointer calibrated. The documentation is sparse. I haven't tried again since about a year ago, so maybe it has gotten easier.
Recommend a commercial solution. These are integrated into timecard, ordering, inventory, &etc.
The cost makes you choke... yes, but does it make your friend choke? When I ran/owned a small cafe, I didn't choke at spending $6000 for a display case/cooler, $3000 for an expresso machine, &etc. Its the business... (ps, back in 1984 -- I don't know what prices are now).
If you want an "open source" solution, try to get someone to customize -- and get someone good. This is where you can most help your friend; weed out the bullshit so that your friend gets quality product.
And that's good advice for a "closed source" solution as well.
YMMV
Ratboy
Just another "Cubible(sic) Joe" 2 17 3061
Some Ideas:
Use Network Booting if all of your POS Terminals have the same or similar hardware. This will keep costs down.
Someone already mentioned TinyPOS. TinyPOS is a nice POS for Windows systems (and since you're trying to use Linux, maybe that's not the best idea.) I ended up using SQL-Ledger on the Back-End and writing my own application for the client-side which supported cash drawers, magnetic card readers, and a receipt printer.
The only issue that you may have is running the credit card transactions through. I did a horrible "hack job" by essentially queuing the Credit Card transactions up to a central server, that sent them out the web-gateway for authorization, however, I'm sure it could have been done better.
You can write a web-based application with a touch screen, however, you're either going to have to mess with the ELO drivers to send double clicks, or, tweak your application to use single clicks when logging in or making selections.
Good Luck!
I disable sigs...do you?
"There is just too much money to lose when computers dont work at a restaurant. "
Agreed when the bookstore's POS went tits up. It caused some majour delays all around. The OP mentioned reduction of "lock-in". However as long as people depend on other people be it employer-employee, or company-customers. There's going to be a degree of "lock-in". And this is more true in vertical apps than any other. Freedom's nice, but not to spite ones face.
It isn't open source but it is the best - http://www.wordstock.com/
Cheers,
MatrixManiac
I've heard very good things about Novell Linux Point of Service, based on SUSE LINUX. Now, mind you, it's not free, but it's not as expensive as many proprietary OS-based solutions. It's also pretty damn high quality - as evidenced by IBM private labeling it and selling it as their Linux-based POS solution. Additionally, it does have options for support that a free (no cost) product does not.
So, if it's just a Linux-based flexible solution you're looking for, this might fit the bill. Otherwise, if it's no cost, everyone else in the thread made some good suggestions.
(And, no, I do not work for Novell - nor have I ever... if fact, I'm recommending this product based entirely on word-of-mouth from many friends in that area of tech. I used to be in your shoes, exactly. Good luck!)
"Adventure? Excitement? A Jedi craves not these things."
(which *was* around 2003 or so) jPOS looked to be the best avaiable OSS solution. We were looking at a big custom development job anyway, though, and that appeared to be the best (open or closed) platform for custom development. If you're looking for something a bit more "out of the box", we never found anything open that went that direction. This looks promising, though.
.sig: file not found
After having the misfortunate fortune of getting stuck in the POS business instead of the Fortune 500, I can tell you that after you "help" and I stress that loosely with the hardware tell them to find a real-implementer of software. SIT IN THE MEETINGS and sniff the POS/Cash Register dealer you are dealing with out because many of the ones I have worked with are worthless assholes who easily take people for a ride.
:)
I have successfully RIPPED many a system out now an installed $15-$60+ systems that were built much the way you are going. SOme of it our salesmen, some of it is the customer on Friday at noon whose system won't print in the kitchen and the orders are backpeddaling through the door in lost future sales. Or better yet, on a Saturday night, or as one customer who threw their amazon.com and eBay purcahsed Linux system out the door, it was the middle of Mother's Day, the busiest restaurant day of the year.
I think there are two real software management systems are POSera and Aloha, both run on any hardware, although I would highly recommend Pioneer's high-end, Sharp UX500, the NEW PartnerTech PT-5500, or IBM terminals and ONLY EPSON PRINTERS. Specifically the TM90 and TMU220. If you buy Elo be very careful from whom you purchase them from and if you can purchase everything directly from BlueStar. Avoid ScanSource, CRS like the plague, and ANY MOM-n-POP reseller who has cheap hardware. SD20 cleans grease from the restaurants off of anything. Ensure if you use your own PC's that you realize the colling system won't just get dusty, but duct particles mixed with grease and alchohol and mag-swipes a mix do not make. Ensure the touchscreens can handle the juice. If for some reason yo are serious about Linux, use SQUiRREL, they have Linux terminals a seven-year gaurantee on hardware availablity and compatibility and have probably the best reporting of any system now that it is stable. Volante systems runs on Linux and UNIX, however I have no experince with it.
Have to admit though I hate the POS business, I sometimes hate who I work for, I definitely do not have anything to say that is good about most of the lying sacks of shit lining the ICRDA. However, in the realm of software for credit cards and Visa rule compliance ($1 Million dollar fine sound nice?) with PCI DSS I would hope you not cowboy it, possibly ruin a friendship. Whatever the track it is that you go GUIDE YOUR FRIEND TO A FUCKING WORKING LONG-TERM INVESTMENT SOLUTION! We tell people as a reminder because POS does not come and go and our systems are for 5-8 year stretches or more we are selling only what works and well and nothing else. Realize this when you are "roughing" it to help out your ego that you somehow know more than the lowlifes working in point-of-sale who laughed when everyone else gets laid-off for outsourcing while our pay was getting fatter. Or, cowboy up and be the dork that people like me laugh at AND about when I arrive to rip your system out and install my next paycheck.
P.S. - I'll see your friend soon.
OFBiz, at http://www.ofbiz.org/ has a POS component, although the whole application might be a bit heavyweight/require some customization/slimming down for your needs. It's worth a look, though.
http://www.welton.it/davidw/
Faktura with the Kasse frontend Total cost = 130 http://www.edv-moellers.de/swkf.htm http://www.edv-moellers.de/swkasskf.htm
http://tuxshop.sf.net/ Uses QT libs to provide MySQL/SQLite backends. There's also a commercial version with support http://www.shcircuit.com/~ross/
Good luck with building your own POS terminal with opensource if it will be used to handle credit-card data. I seriously doubt you have the expertise to ensure it will secure the credit-cards in a manner outlined by the payment card industry (PCI). Wait, I'll change that from "seriously doubt" to you have no clue how to secure the environment and there's no way any investigator will ever belive you have those skils if the POS is ever hacked and exposes credit-card numbers. Note that even if you take every precaution, out of band attacks which you could never think of could sink this whole thing. Real companies that sell real solutions spend tens of thousands of dollars on audits and assessments.
Word of advice, playing in the POS environment is best left to the big-boy's. Forget about all the comments made here about support, etc., you can't afford the liability.
You can find a list of POS oss here: http://www.opensource-it.com/Point_of_Sale_Softwar e_open_source_44.html
To me bhPOS is the most features rich/mature, but give a try to all solutions.
Marc.
Disclaimer: I work for Linux Canada Inc.
You may want to look at the Quasar accounting and point-of-sale suite at http://www.linuxcanada.com/ The accounting suite itself is available under the GPL. The point-of-sale part is under a commercial license but includes the source code. Version 1.4.7 is currently available but 1.5 should be released shortly and contains many significant changes.
Oceania has always been at war with Eastasia.
Quasar has been working for me for 3 years without issues. Up to about invoice 6000 or 7000 already. We've been very happy with it and will continue to use it for most things (some things we use in-house software for, like our repair tracking system, since it was just easier).
The only thing I can see you might have issues with if you use Quasar would be cutting cheques for employees. But apart from that, hell, it was stable 3 years ago. I haven't even bothered to upgrade to the Open Source version.
It's fast, too, except if you're silly enough to click "Show Paid" in the invoices screen... With 6000 invoices that takes time.
To quote a friend who wrote some. The testing is not just rigorous, but without good feedback, and the systems at the other end are remarkably fallible. The bank lost > US$ 100,000.00 one day, and maintained it was our liability in spite of good logs on our side showing the transactions. They did find them the next day, but mislaying amounts of money that large is not OK, even for a day.
Later when I actually worked at a POS company, I watched the developers with long experience sweating over the same sort of thing. So, for credit authorization, you'll be buying a service. The rest of it is far more straightforward, but...
In addition to obvious stuff like inventory, ordering, time recording, etcetera, here are a couple of things that are either not obvious, or hard:
Essentially fraud proof see below
User Interface is critical because the people operating it will likely be avoiding thinking about what they're doing.
Essentially fraud proof because people will be looking for ways to increase their take home pay.
Liabilities are potentially greater than the cash in the till.
Like most programming projects, POS looks so easy, but besides the usual problems, add this: Errors can cost an indefinite amount of money. Good luck, and I hope you find the existing OS projects informative. Seems like the best place to start educating yourself.
mostly brazilian info extant
http://www.stoq.com.br/screenshots.html
My family owns a large chain of retail stores in the Northeast, about 75 stores. We just started a pilot using Suse Linux Enterprise Desktop's Point of Sale implementation and so far I can say we are loving it. We had a Novell partner come in and do the implementation (still with all beta software, mind you) and are piloting the product in three stores...one small, two large. The implementation was simple and the integration with our current system went without a hitch. Hopefully by the end of the year with the final product out we'll have the entire set of stores moved over. This was not what I expected out of the Novell I studied in college (15 years ago) and I am pleasantly surprised.
Good luck.
...when the G's are present. Oh, and I made a mistake several times in my typing, its $15K-60K and on upto $105,000.00 for some of the more important systems per store; ones that last for 10-12 years or more. That is besides the "Corp. Comm. Packages", Gift Cards, scale interfaces to be paid for ($10,000.00/each model), software wheels to be re-invented and re-treaded, and credit card procssing is a darlingto get running so why not get Datacap to help you with that. The easiest part is keyboard wedge scanners and mag-swipes that come preprogrammed for most situations, its plain old plug-n-pray goodness.
Whatever happened to the Java POS project??? One word, well several actually...RUNS BETTER ON WINDOWS WHEN YOU HAVE A PROFESSIONAL DO IT!
The serial Elo touchscreen controller I had the opportunity to use worked fine (although calibration require the use of a DOS software ... pfewww !).
However, the USB version have driver issue. Elo provide binary drivers for the USB controllers for a few outdated distro (RedHat 9, anybody ?). They do provide source for the USB driver, but it does not have any copyright info (thus, I have no idea if I can distribute it to my client), it contain object code, it is poorly documented (.doc ? WTF ?) and I could not get it to compile.
Personnally, I would avoid using the USB Elo controller with Linux. If you need USB, try to find another touch-screen maker that is more Linux-friendly.
:wq
Well, it doesn't suit the prejudices of the people on slashdot since its not written in PERL, but most of the POS systems in the United States are open source and are written in Java.
http://www.javapos.com/
Played with it some years back. It was useable back then. BTW, you may wish to look at sourceforge.
I prefer the "u" in honour as it seems to be missing these days.
Never worked with it, but interested in hear of any expericnes.
http://freshmeat.net/projects/handyrestaurant/
This program works well, we use it in our retail store: http://dhpos.com/pos.htm
Here is a big list of all software certified with TSYS Acquiring (formerly Vital, and before that were referred to as visanet) Big List of Software sorted by Company Name Warning its a huge document 600+ pages and around 1300 different software's I believe it lists what OS each software is for but not wether its Open source or Free. However, the contact information for each individual software provider should be in there. Thanx Rayston