Live data from Hacker News

CashRegister in PyQt5 with barcode logon and 32 programmable buttons

github.com

11–20 of 29 posts

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#11

Earlier quoted context omitted.

Any specific reasons why you think it would compare poorly to other commercial applications. I have tried the ones at Walmart and they are sometimes very slow to respond .

To be fair, some commercial systems suck on UX too, but some of them are much better than this. If you compare it to eposnow[0] for instance, and see the adoption of large clear buttons. The real differences, as I see it, between this and commercial systems are 1) most commercial tills are clients that poll a server that allows central management of stock, prices, and extraction of sales data by api 2) commercial til…

It's easy to put this on a server with clients pos. For a server database system postgreSQL is used.

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#12

Programmer who also works in retail, uh no, this layout highlights all the wrong things. The list of items is half the screen, my current system has it at about 1/4 and buttons are enormous, very easy to use and rarely ever miss a button or hit the wrong one, this looks like that will happen all the time. Also we have a lot of colour in buttons that relate to use (payment buttons red, department buttons green, accoun…

You should not compare this with a commercial program, I'am a hobbyist and not a professional. I think it's a good effort for an open source product.

He is still giving you constructive feedback that you can use to improve your product. Even if it's just a hobby. I assume that is why you posted it here?

If I my add something as well. Use a local git program to commit and add files instead of using drag 'n drop from the website. Add commit massages with a short description of what you changed.

I would also remove all the logos in the program.

Keep at it :)

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#13

Earlier quoted context omitted.

Any specific reasons why you think it would compare poorly to other commercial applications. I have tried the ones at Walmart and they are sometimes very slow to respond .

To be fair, some commercial systems suck on UX too, but some of them are much better than this. If you compare it to eposnow[0] for instance, and see the adoption of large clear buttons. The real differences, as I see it, between this and commercial systems are 1) most commercial tills are clients that poll a server that allows central management of stock, prices, and extraction of sales data by api 2) commercial til…

Commercial tills also integrate with a huge bunch of other specialized periphery. Payment terminals are an important category (and an ugly one, because each country has its own, and there are many different protocols in use), but receipt printers, barcode scanners, RFID systems (to perform cashier authentication), cash drawers, checkout scales, cash management systems, RVMs (reverse vending machines; those things taking bottles back at the store) and whatever other, often custom hardware big retail chains come up with (think store signaling systems for example).

They also need to take fiscal requirements into account: in a growing number of countries, cash register systems have to satisfy fiscal requirements and must perform some sort of transaction signing and reporting to fiscal authorities, or integrate special hardware which gets all transactions and prints out the receipts instead of a "normal" printer on which you can print anything. This is an ugly topic that easily keeps several full-time employees busy, because fiscal requirements are mandated by law and thus are moving targets. The US doesn't have this at the moment, but compensates by having a rather special and complex tax system ;-)

And then there's the calculation of prices modified by promotions (think coupon codes, time-controlled rebates, such stuff) which is an entire topic on its own, especially when taking the interplay of promotions with sales tax calculation and rounding logic into account (not all countries round to .01 cents, some don't have anything smaller than 0.05, and some retailers voluntarily decide to round to multiples of 0.05 in order to eliminate handling effort with small coins).

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#15
post #12

Earlier quoted context omitted.

You should not compare this with a commercial program, I'am a hobbyist and not a professional. I think it's a good effort for an open source product.

He is still giving you constructive feedback that you can use to improve your product. Even if it's just a hobby. I assume that is why you posted it here? If I my add something as well. Use a local git program to commit and add files instead of using drag 'n drop from the website. Add commit massages with a short description of what you changed. I would also remove all the logos in the program. Keep at it :)

Thank you for your response. In the installation directory is a changelog.txt in which all the changes and additions in the program are documented.

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#17

The guy who made this also made a desktop gui ERP application to teach himself python (or programing general, I forget). He is an inspiration. Keep doing what you are doing Dirk. https://github.com/DirkJanJansen/Pandora

Thank you for your inspiring comment.

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#19

Asking out of curiosity: why did you choose Riverbank’s python binding instead of Qt’s?

I don't know the difference. I learned myself python, Pyqt5 and postgreSQL in a relative short period, mostly by examples and documentation online. Got a little knowledge from my past by learning and programming in Clipper and QBasic for my last company, long time agoo (1997-2002). Now i am retired and i like to program in python (My age is 75)

Re: CashRegister in PyQt5 with barcode logon and 32 programmable buttons

#20

Asking out of curiosity: why did you choose Riverbank’s python binding instead of Qt’s?

I don't know the difference. I learned myself python, Pyqt5 and postgreSQL in a relative short period, mostly by examples and documentation online. Got a little knowledge from my past by learning and programming in Clipper and QBasic for my last company, long time agoo (1997-2002). Now i am retired and i like to program in python (My age is 75)

Gotcha. I'm a commercial Qt developer and starting to think about Python being a first-class language in my ecosystem. But I've learned that the more robust binding, PyQt, is a GPL/Commercial offering separate from the Qt licensing.

Qt for Python is the "offical Qt" binding but it's not as robust or even supported on some platforms like embedded. It's a bit frustrating.

Still, nice work!

Post reply on HN