Linux To Run Sherwin-Williams Cash Registers

oilfieldtrash writes "According to this news article on Yahoo!, Sherwin-Williams will upgrade their point-of-sales systems to Linux ... 'Sherwin-Williams Co., the No. 1 U.S. paint maker, plans to convert its computers and cash registers in more than 2,500 stores to the upstart operating system in the next year and has hired International Business Machines Corp.'s services division to do the job.'"

Comments

Not Exactly A Win For Linux

[sputnik73]

This is part of a continuing pattern that I've noticed. The major corporate entites which are embracing Linux aren't normally leaving some variant of Windows behind but instead are dropping Unix. The stranglehold Microsoft has on Office and the problems introduced by switching from Windows to Linux (in terms of a possible inability to access old files) is really hurting Linux in the War against Windows. But what these companies need to realize is that they can convert their old files into plain text files, using the very version of Office which is trying to tie them into an ugprade cycle of doom, using some simple batch scripts. This would be quite a chore, obviously - but in the long run companies would save. I don't know why this solution isn't being offered to companies. From what I understand, many companies are hesitant to drop Windows for this very reason: loss of access to old files. But again, Bill Gates doesn't really lose on this one. Linux gains some but not in the area where I'd like to see it.

Re:Not Exactly A Win For Linux

[the_rev_matt]

I disagree with this. Ubiquity is key. People expect shit to just work in the store. The people in the store (including mgmt people) are going to be dealing with Linux every day and will depend on it to stay in business. It performs well, they have a positive impression of Linux. Managers will get tired of switching between Windows in the office and Linux on the sales floor, and some of them will try linux on their office computer.

Customers see how reliable the POS is (esp at christmas) and want to know why the store never seems to have problems with their computers (trust me, it happens. I had several conversations during Xmas of 98 about the benefits of UNIX (SCO in that case, but I generalized to all unixes) vs Win98 with businessmen-type customers who asked). People will rationalize: If Sherwin-Williams/big car company/home electronics store/blah blah blah trusts their business to linux, maybe I should look into it too...

Re:Not Exactly A Win For Linux

[Methuseus]

you *don't* need to use POT on Linux. Get an old (free) version of Staroffice and use it to convert the files. Granted you may need to make small adjustments, but in the end, the parent poster is right, in the end it will save time and money. If you don't want to ever move from your 98 machine with Word 98, so be it, but you have to at least notice that your argument is stupid and irrational.

Re:Not Exactly A Win For Linux

[autechre]

You've been using Word98, a MacOS product, on your Windows 98 machine? That's pretty cool.

(Microsoft released office 95, 97, and 2000 for Windows. I assume you're using 97.)

The interesting bit is that if you go back far enough, Open/Star Office starts to get better at old Microsoft formats that MS themselves. Word 95 did a _terrible_ job of importing MS Works files...hey! There's that whole argument about "losing old documents." Sing it with me now: "Throw it out the window!"

Also, keep in mind that Office 2000 was the first one to even attempt a backwards-compatible file format. Try to open a Word 97 document in Word 95 and see what you get. So people were forced to upgrade. Maybe 97 added some nice features for some people, but if I'm only writing plain text with some italics, why can't 95 read it anymore?

Personally, I write stuff in Abiword and OpenOffice, and people have no trouble reading it in Word (and vice-versa). So yes, while I prefer to do things in vi, I'm by no means restricted to it.

Re:Not Exactly A Win For Linux

[tomstdenis]

Your entire post is so full of shit I can't believe it myself.

I have personally worked for a few jo-jobs and I can tell you straight-up that nobody include the mgmt care what OS they use as long as it works. For instance, I worked for PharmaPlus [in Canada] and their servers run Unix [or linux, it was some *nix to be sure], the front cashes were Compaq's running win98 and the postal cashes were some other brand running NT4.

And I never had any customers come to me and say "I see you are running 98, well thats completely unreliable".

Believe it or not but 99% of the customers don't have a first clue about computers or the diff between say Win98 and Win2k other than the name.

The fact that they use Linux is neat but as another poster pointed out not entirely a big win since the goal is to crush MSFT isn't it?

Tom

Re:Not Exactly A Win For Linux

[iabervon]

I think it's great that companies are switching from proprietary UNIX to Linux, because it's an application where they have genreal familiarity with the interface and so forth. Once Linux is a proven technology in their company, they can consider Linux as a replacement for Windows. The places where they are using Windows are places it is not feasible to switch to commercial UNIX. This will give people the opportunity to run the same OS everywhere, since it is not too expensive for the desktop and stable and secure enough for the servers. Once IT is using Linux for the critical systems, they won't want to support Windows as well. Microsoft doesn't lose anything on this deal, but their competition is being replaced with a much tougher competitor.

Furthermore, POS terminals are a market that MS would like to own: the user interface matters and commercial UNIX is impractical (unless the terminals are just thin clients of a single server, which limits the UI possibilities). The fact that companies are chosing Linux as the OS with an interface that's fast, easy to use, and powerful, on cheap hardware, has got to hurt.

Re:Not Exactly A Win For Linux

[PotPieMan]

Managers will get tired of switching between Windows in the office and Linux on the sales floor, and some of them will try linux on their office computer.

That's debatable. If IBM does its job in making the POS systems usable, cashiers and managers aren't going to care what their cash registers are running. It all goes along with the "people expect shit to just work in the store" idea. Windows is good enough for the office and home, especially on the level at which these people compute. I don't see many making the connection between a POS system and their workstations (hint: one runs on hardware that usually looks little like a standard computer, and performs a limited set of functions).

In your perfect world, employees and customers would see the reliability of the Linux POS systems, and wonder how they can get that reliability at home or in the office (as you allude to). However, I look at this decision more as the orignial poster does - it's a win for the acceptance of Linux into the business services field. I don't think it will amount to much more than other companies considering Linux for their mission-critical services (it's definitely a win for IBM in that respect).

Re:Not Exactly A Win For Linux

[oni]

We run a web interface to a com object of MS Word/Excel

You got a license for that buddy?

Seriously, if the documentation for your setup is available I'd like to take a look at it.

Re:Not Exactly A Win For Linux

[teslatug]

really hurting Linux

Huh? Linux is neither a living organism nor a corporation, it cannot be hurt. Consumers on the other hand can be hurt by Microsoft's monopoly. Any conversion to Linux systems brings in more recognition. Any such conversion is a win for consumers because Linux is also a threat to Microsoft's monopoly.

Re:Not Exactly A Win For Linux

[sconeu]

had several conversations during Xmas of 98 about the benefits of UNIX (SCO in that case, but I generalized to all unixes) vs Win98 with businessmen-type customers who asked).

Before the Caldera buyout, SCO was a MAJOR presence at POS locations.

Re:Not Exactly A Win For Linux

[IAmATuringMachine!]

Customers see how reliable the POS is

I thought Windows was the P.O.S.

Re:Not Exactly A Win For Linux

[Hercynium]

Geez, I was going to mod up some people who deserve it, but I *had* to reply to your entry.

I'm a music junkie, I admit... I spend waaaay too much money at a few local music stores and when I finally get to the counter I DO notice when their POS systems are taking forever! For example, a small local chain is using a Win2K-based system running over Terminal Services to run their POS cash register systems... it royally sucks. Now, I'm not expressly criticising Win2K, but one thing they often say when it's getting sluggish... "someone needs to reboot the server..."

Other stores run without a hitch... and I often find myself feeling that I'd rather buy stuff at those places since I don't have to wait 5+ min for the cashier to complete my purchase!

Now, I'm not saying Linux will solve their problems... but a better POS system would. What's depressing is that I've developed several rigorous POS systems, one of which is now 6.5 years old and still running in a restaurant chain with 88 locations... on OS/2!

Re:Not Exactly A Win For Linux

[Copperhead]

Oh... POS is short for "Point of Sale"... I thought you meant... oh, nevermind.

Re:Not Exactly A Win For Linux

[MrCreosote]

"But again, Bill Gates doesn't really lose on this one."

Oh, yes he does. First, there is the revenue lost because the potential new customer did not buy the Windows solution (which they would have otherwise if it were not for Linux). More importantly, there is one less slave to the treadmill that is the upgrade cycle of Microsoft products, and the revenue stream attached to that.

Re:Not Exactly A Win For Linux

[cybermage]

But what these companies need to realize is that they can convert their old files into plain text files, using the very version of Office which is trying to tie them into an ugprade cycle of doom, using some simple batch scripts. ... I don't know why this solution isn't being offered to companies.

I've got karma to burn, so here goes...

Converting old files to plain text is an incredibly bad idea. Even if your only talking about .doc files, you need to understand that there is meaning in format. If you cannot take the formatting with you, you're not preserving the meaning. Simple example:

Life of Brian is excellent.
Life of Brian is excellent.

Now, beyond the issue of the simplest kind of old files, consider things like spreadsheets, PowerPoint presentations, Accounting data, etc. Converting these files to plain text will render them useless.

If you want businesses to move to Linux, look back at how we got here. There was a time when nothing was done on computers. Why start? Everyone could use a pencil. All typewriters worked more-or-less the same way. Businesses didn't start using computers because they were cool. They didn't start using them because they were cheaper. They started because computers could do things faster. Computers represented solutions to problems. What your proposing is to make a problem out of a solution. You won't get businesses to agree to problems in order to save money.

If you want to see a mass migration to Linux, here's what you need:

• standard file formats - Linux tools need to open/edit/save Microsoft formats. It is still my hope that Microsoft is forced to cough up their document formats to a standards body as a result of the anti-trust suits.
  
• cloned interfaces - Linux tools need to perform tasks in an identical fashion to their Microsoft counterparts. Corporations will be very reluctant to retrain their entire workforce in exchange for a free OS. The applications need to work the same.
  
• inter-operable OS - provide emulators to run legacy applications. There's a reason why knowledge of Cobol was needed before Y2K: Some old, proprietary applications cost way more to replace than 1000 years of OS upgrades. For many companies, their data is trapped inside computers they don't even understand. Once common applications have turned into black boxes. Make it so they can copy the application and it's inseperable data.
  
• education - Linux needs to be accessible to educators to share with students. Schools, libraries, and colleges all receive generous donations of Wintel computers. Children are taught from an early age that Windows == Computer. They learn their tools under Microsoft's guidance. There is, however, a back door. Community Colleges are chartered to teach what the community asks for. Businesses who've embraced Linux and OSS need to ask their local community college to get their future employees ready. Invite educators to tour your business and sit with your people.
  

Re:Not Exactly A Win For Linux

[Melantha_Bacchae]

Methuseus wrote:

> Get an old (free) version of Staroffice and use it to convert the files.

To what if not text?

> but you have to at least notice that your argument is stupid and
> irrational.

DavidJA wasn't arguing, but was asking perfectly valid questions:

"The question is WHY should I "upgrade" to linux????"

and

"Why would I want to change to an OS where I have to use plain old text for my word processing?"

Instead of getting insulting, why didn't you just simply answer the questions? I'm sure there is more value to Linux than a vague promise that converting all of one's files to some unknown format will somehow "in the end ... save time and money".

I'm also sure that there are a wide variety of file formats for documents to choose from in Linux, and a few word processors that can read MS Word files with a varying degree of success. I don't know about support in Linux for the more advanced features of Word used by businesses: such as mail merge, Word document templates, forms, and VBA scripting. Of course my knowledge of current Linux features is limited to my Zaurus; OS X is my forte.

These are questions that Windows users have that are going to have to be answered if Linux is going to make it on the corporate desktop. Answering honest questions with insults and ridicule is not going to help Linux get there.

For those who want some actual facts on the subject, here's some stuff I found on Google that might help:

http://www.linuxlinks.com/local/business/wordpro ce ssing.shtml
http://www.canadacomputes.com/v3/stor y/1,1017,5413 , 0.html?tag=134&sb=281
http://wwws.sun.com/softwar e/star/staroffice/6.0/

What happens when you embrace and extend Godzilla? Nuclear heartburn!
See "Godzilla 2000" (released in Japan as "Godzilla 2000 Millenium") for details.

Re:Not Exactly A Win For Linux

[tftp]

It is still my hope that Microsoft is forced to cough up their document formats to a standards body as a result of the anti-trust suits.

I believe, MS does not even have these formats documented. Probably there is some preliminary documentation that is completely obsolete, and nothing more. The real stuff is in the code. They only had to write it once, after all.

Documenting file formats is very difficult because even simpler format, like RTF, is a language on its own, with a syntax, rules, parsers etc. A binary format of Word 97 - which can contain billions of objects in other formats - is a nightmare to implement, and an impossibility to document, especially when MS has no particular reason to do that.

The worst part is not in binary encoded tags, and not in how they are attached to paragraphs of text. The most difficult part in how they affect the document! This, however, is probably beyond the responsibility of the import filter. If a lot if features depend on other modules "as they ended up being designed", then even the complete spec on the file format will be useless. For example, if the tag says "Left indent 0.1" and the rendering module snaps it to 0.25" grid, you have to know about this quirk to recreate the document in a different wordprocessor. Keep in mind that if you don't put every character exactly where Word put it, the paragraph (and the document) may flow (and the layout will be ruined). The fact that metrics of MS fonts may be not available on other systems does not help either...

If most of the documentation is in the code (comments and the code itself) then they can't release it without releasing the sources of the import filter, editor, WYSIWYG renderer... and all the rest of MS Word. IMO, chasing Word is a losing game. A new generation of wordprocessors must be created (OpenOffice is a good one, AbiWord is a decent thing too), and these wordprocessors would be open enough to interoperate well.

Re:Not Exactly A Win For Linux

[cybermage]

A binary format of Word 97 - which can contain billions of objects in other formats - is a nightmare to implement, and an impossibility to document, especially when MS has no particular reason to do that.

I can see your point that Microsoft may not even know their own file format. However, that does not change the fact that dependence on tools like Word will keep businesses enslaved by Microsoft. For businesses to consider moving to Linux, they need to be able to take documents with them.

I agree that even if the document format were a wide open standard, the exact layout will not be the same for a variety of reasons (bugs, fonts, etc.) However, I believe people will be happy as long as they can open/edit/save/print the file.

Finally, I am not suggesting that we chase Word. I'm suggesting that they be forced to stop running. The file formats for tools like Word need to be standardized and managed by an independent body like the W3C. You should be able to use any word processor that is standards compliant to create/edit/save these files and hand them to any other word processor that is compliant to do the same.

I'm all in favor of creating new tools; but if they can't open old docs, you're wasting your time.

Re:Not Exactly A Win For Linux

[tftp]

dependence on tools like Word will keep businesses enslaved by Microsoft. For businesses to consider moving to Linux, they need to be able to take documents with them.

It might be not possible, in near future. Here is a real life example, less than a week old. I insisted on using StarOffice 5.2 to prepare a user's manual. This is not too fancy document with some simple illustrations, some text and few indexes, total about 20 pages. Result? Total failure. Here is why.

SO 5.2 actually works, in 99% of cases. But it is not good enough. I thought that the user will scream and whine forever about different buttons and different menus. Well, it stopped after a while, and he got used to the new software. This appears to be only a small problem.

The real killer problem is reliability. MS Word had more than 10 years of debugging, it is fairly stable now. SO 5.2 is not. We discovered so many obscure little bugs that by now the decision to migrate StarOffice documents to MS Word is pretty much a done deal. Among those bugs:

• Broken spell checker (on Windows)
• Incorrect kerning on some zooms
• Too hungry deletion of highlighted text
• Hangup/Crash if another window gets focused
• Broken "to right margin" checkbox in some cases in index styles
• Very weird flow control
• Anchoring frames to paragraphs does not reflow
• Two indexes weld to each other, Backspace causes lockup and crash
• ... many more, I don't even remember.

The point here is that businesses need reliable, robust, bulletproof wordprocessor. A secretary or a CEO don't want to see obscure dialogs. Can SO and OO get there? Sure. But it will take time, lots of it. MS used its time very wisely, and got a very strong foothold. And anyone who says "Businesses just need bold and italics" are mistaken. Businesses need all features - because technical documents are often very complex and have a maze of page styles. Your suggestion that small formatting loss is OK is not acceptable. Well, you can sell that to me, and I can sell that to you - but neither of us will convince our coworkers that they should come to work on weekends and fix conversion bugs.

So you said dependence on tools like Word will keep businesses enslaved by Microsoft. True, but most businesses don't understand that, and they don't care either. They won't migrate to anything just because it is marginally cheaper.

IMO, the way of liberating our documents is in gradual migration from Office formats to open formats and open applications. Look how Ogg Vorbis slowly but surely enters the MP3 world. Still 99% of all songs are in MP3, but more and more appear in .ogg format. When new format achieves some critical mass - and when applications are stable and good enough not just for geeks but for normal people - then they will be embraced and accepted.

WRT your idea about W3C standards; they won't be of any use without applications, and as I said, applications are very difficult to develop. Just look at Mozilla. Just look at codebase of OO. These are BIG apps. Sure, after 4 years of planet-wide debugging now Mozilla works. We need to have a wordprocessor of exceptional quality, or else it won't be accepted. Formats are secondary, and as people suggested, businesses always can have one or two copies of old Word to open old documents; PDF is for the rest of them.

Well, this comment is not very well structured, but it is good enough for posting at 3:25am ;-)

Re:Not Exactly A Win For Linux

[cybermage]

WRT your idea about W3C standards; they won't be of any use without applications, and as I said, applications are very difficult to develop.

The application already exists. It's called Microsoft Word. I'm not suggesting that a standard get created out of the blue. I'm suggesting that Microsoft be forced to disclose the current format, that the format be adopted as the current standard, and that a seperate organization similar to the W3C be created to control it. It makes a great deal more sense to take a format that 95% of people are using and make it the standard than to create one. The fact that Microsoft has been found to be behaving criminally provides a window to do this as a remedy that will help to eventually break their monopoly.

I realize that most other tools aren't ready for prime time, but that's not the point. I won't even try using those tools until I can share my files with people who use Word and vice-versa. Because I need to share docs with others who use Word, I need to use Word. Because I need to use Word, I need to use Windows. I'm stuck using Windows because of the Word file format. You could create a word processor for Linux that typed as I thought, and I'd still be using Word under Windows until I can share my work with people who use Word (and them with me.)

Re:Not Exactly A Win For Linux

[tftp]

It makes a great deal more sense to take a format that 95% of people are using and make it the standard than to create one.

Sure, if the binary Word format can be ever documented.

But you don't need to go that far. All open-source wordprocessors read RTF, and all Word versions cheerfully write RTF. Is it of use? Not really, because wordprocessors themselves fail to make good use of the tagged data. You get the text, but lose layout (I don't even mention embedded objects, graphics, rarely used WordArt etc.)

As I said before, the only wordprocessor which can correctly open MS Word documents is MS Word itself, because every little bug and every little typo in its code affects the layout.

Of course, if I were the dictator (copyright honors go to W), I would just mandate use of HTML and PostScript everywhere. Good or bad - does not matter, it would do the job, and compatibility would be a greater good than loss of minor conveniences here and there.

Re:Not Exactly A Win For Linux

[Zathrus]

standard file formats

OpenOffice can read/write most .doc and .xls files, with good results from what I hear. MS will not have to release it's file formats - their business applications aren't even being vaguely referenced in the anti-trust suit and are not part of the suit, period.

The other points are good... although I'd somewhat question the value of the education bit. Apple has had deep discounts for decades and it hasn't gotten them anything.

Re:Not Exactly A Win For Linux

[Balinares]

Furthermore, POS terminals are a market that MS would like to own.

Actually, I discovered they /do/ have a foot in the cash register system market already. At a big gardening shop in my county, I was surprised to notice that the cash register terminal was running Windows. So I go, "Oh, nice screen!" to the cashier girl, who looks at me warily and nods. "Does it work well?" I add, with a big nice enticing smile with "Don't worry, I know what you feel" written all over it. And then the girl leans toward me and blurts, "Oh crap no! Keeps crashing!"

Still, see, MS has a foot in that market.

Re:Not Exactly A Win For Linux

[iabervon]

I have heard that Windows cash registers exist. I have no evidence that anyone likes them. That's how I know that MS wants the market.

Re:Not Exactly A Win For Linux

[iabervon]

There are multiple reasons not to switch from Windows to commercial Unix. One is that Unix workstations are an order of magnitude more expensive, and the software is similarly expensive. Commercial Unix has never been priced such that you can put a box on every desk in every office.

For that matter, you're not going to have as easy a time giving your workers machines that their consumer hardware (e.g., pilot cradle) plugs into.

People who work for radio stations still listen to regular stereos at their desks, because it doesn't make sense to give everyone the high-end equipment when they don't need it. But if the high-end equipment was free and had jacks for normal headphones and played all of the media that normal people have, they'd probably switch to using it, since it's better and there are people around who know how to manage it.

Re:Not Exactly A Win For Linux

[pmz]

The point here is that businesses need reliable, robust, bulletproof wordprocessor.


So, how does Word fit into your argument?

Word is buggy as hell. I have Word 2000 and end up cursing at it before I'm through with something. I'm generally a nice guy, but fifteen minutes with Windows 2000 and Word 2000 is enough to make me want to destroy my computer and walk to the next city.

Fact is, you are used to the bugs in Word and didn't like StarOffice, because StarOffice comes with different bugs. Live with it.

Re:Not Exactly A Win For Linux

[tftp]

StarOffice comes with different bugs. Live with it.

I do, personally. I can use anything, from TeX to Quark Express. That's the "users" who refuse.

You see, I am just tired of getting accused again and again of "sabotaging our company" and "slowing down the document preparation". Valid or not, these are exactly the arguments flying around. What would *you* do if you are daily accused of harming your company and your coworkers? They all say "Word is perfect, SO is a POS." I say "Ok, then use Word but don't complain to me when it fails on you". We shall see how soon they find out that Word is much more primitive than SO.

Re:Not Exactly A Win For Linux

[Thing 1]

We run a web interface to a com object of MS Word/Excel, and people can upload files in MS format and download in a more-open format (eg, RTF/CSV).

Someone please combine this with the latest Outlook/Windows exploit, and convert the Earth's entire repository of proprietary documents into open formats.

Yes, you'd be prosecuted, so no, I didn't tell you. ;-)

However as part of the agreement with IBM....

[b0r0din]

Sherwin-Williams can only sell the color Blue.

Re:However as part of the agreement with IBM....

[quinto2000]

That was a good first karmawhore. I thought it was relatively funny.

Re:However as part of the agreement with IBM....

[quinto2000]

Ironically, you were modded down and I wasn't, while my comment was clearly editorial, not topical. How sad.

Upstart?

[awptic]

Is that what we call something that's been in developement for 11 years?

Re:Upstart?

[theCat]

"Upstart" is what your TechDesk news editor calls it when 1) he doesn't use it, and 2) he doesn't like it. I guess Linux will be an "upstart" right up until Microsoft starts selling its own distro. And on that note, I'm going to trademark "Windows LX" just in case.

What POS software will they run?

[laserjet]

I currently admin a few stores that run a POS (Piece of Sh*t) POS (Point of Sales) program called Microbiz. It runs on Win9x, and it is an unstable piece of junk.

We need to migrate to a new software due to the fact that support will be stopping on our current software within the year. I know there is LinuxPOS, but has anyone tried it? We need a full featured POS app for a small/medium size business.

Things like this give me much hope, as I have always thought that Linux is the ideal point of sales software: it is stable, can be no frills, has good user access control, and the network and remote admin can be made easy.

Linux, while it may not be the most used for gaming and multimedia, may have a niche in the POS market. In my view, it would be the perfect OS for the retail environenment.

Re:What POS software will they run?

[the_rev_matt]

I've heard that LinuxPOS is pretty decent, but never used it myself. It would not be wildly complex to write a web based POS system in Zope/PHP/Perl against PostGRE in a weekend. Hell, if you already have an online store, you already have a POS system. You just need to add the capability for a clerk to do returns and maybe some enhanced searching for in store and you're done. There are already Zope/PHP/Perl apps that do this. Check freshmeat.net and go nuts.

Re:What POS software will they run?

[bovinewasteproduct]

If you think writing a POS is sytem that simple go for it! I've been involved with POS since 1990 (My dad co-owned one of two POS scanning companies in Houston from 1987 to 1995, I was his lead PC tech from 1990 to 1993) and I can tell you it ain't that easy. Sure, if all you want is a basic cash register it's not too hard, but when you get into scanning it gets just a little bit harder.:)

And a web based system will NOT cut it either. You'll need an X app to compeate with the windows versions out there. Plus you'll have to write drivers for the printers, scanners (those that don't have a keyboard wedge available), scales, cash drawers, customer keyboards and the other things people hook up to a POS system.

Been there, done that, DID NOT WANT THE T-SHIRT!

BWP

P.S. This is just retail, when you get into hospitality systems (ie selling food), it gets even worse!

Re:What POS software will they run?

[ender81b]

perfect OS for the retail environenment

Let's list of a few reasons for why this is so:

• 1.) Free. Really. Perhaps I should say cost-effective?
• 2.) Customizable. Completely, 100% customizable to your needs. You can modify whatever you want to make it your own.
• 3.) Can run on a old 486 if you want it to. This is a biggy for retail stores. Some of those cash register type thingies cost thousands a pop. Install linux on an old 486 which you can find in the dump for god's sake. The best part about it? You can probably go 5 years or so without ever having to worry about upgrading. I would even say 10 if you go with a pentium.
• 4.) Secure and Stable. Doesn't really need an explanation does it?
• 5.) Easy Network/Remote access admining. This is a big plus. At the restaurant I work for the software they use is a pain, and so is admining the machines. The cashiers tend to surf the internet, etc when they shoudl be working. You could eliminate these factors on a linux system

I think the really big thing for company's is the ability to completely customize the OS to fit their needs. Hire a couple dozen programmers and wham - your needs are taken care of for years to come. You don't have to worry about vendors, licensing costs, etc. Nice and Neat

Re:What POS software will they run?

[cscx]

Nah. DOS had the POS niche. It will for a long time to come.

Re:What POS software will they run?

[NineNine]

I am in the process of opening a retail store, and I'm writing my own POS. I'm not using Linux. Why? The sheer complexity. I'm writing a very simple VB app against an Oracle back end, using the COM object supplied by the scanner companies. W2K is rock solid. Linux may be solid too, but there's nothing as simple, and as easy to modify as a well-written VB app for the front end. I certainly don't want to write some cryptic C app!

Re:What POS software will they run?

[ender81b]

You are right - for a small business it probably doesn't make all that much sense to use linux. But, consider if you had 500+ (or however many) stores. It would be much more reasonable then.

Re:What POS software will they run?

[Khazunga]

You are not comparing equivalent languages. C can't be compared to VB. A more adequate comparison would be VB and Python (perhaps with Qt or GTK bindings).

I don't see how Python is any more difficult to program than VB. It's just that stuff you don't know seems more complex that what you're used to.

You, or fellow developers, I am afraid to say, are the reason I have to wait *ages* at my local hypermarket to get the groceries checked out. But I guess you'll only get it when your app is in production, with 200 terminals. By then, it'll be too late.

Re:What POS software will they run?

[Bazzargh]

You might want to read this article by Jamie Zawinski (of Lucid Emacs/Netscape fame) on his attempts to get a Linux PPOS app for his club, DNA Lounge. Includes source code for his own one - I dont expect it meets your reqs but the article is useful for a review of the state of the art.

Re:What POS software will they run?

[PD]

I agree that writing a POS is extremely difficult.

What most people don't understand is that this thing doesn't just ring up sales, it runs a BUSINESS. The business owner is going to expect that the thing is customized and tailored exactly to his specs.

That means the system has to support many different types of retailers, with full flexibility.

About 10 years ago I wrote a POS program (company out of business now) and we had immense difficulties with various things like:

-insane Pizza shops wanting the ability to invent coupons on the spot. They honored their competitors coupons, and if it was valid only on a large three topping pizza where one topping was sausage, then he had to have that same coupon in his system, with the same computer enforced restriction. Oh, and it had to be simple enough so that his 16 year old driver could program it in real time while the customer was on the phone with him

-dry cleaning has inventory, but not the way you think. They treat the clothes they are cleaning as inventory. This inventory isn't like any inventory you normally see. It moves.

-or how about this: Some dude runs a restaurant. Ok, no problem. We've got all the shit in there. But wait! He's opening a theater to make it a dinner theater. He wants his POS system to handle ticket reservations and scheduling! Ahhhhhhhh!

Ah, but you say "I'm a programming wizard. I will just whip that up in a night of programming." But it's not that simple. How robust is your code? Can it be crashed by a 16 year old hitting all the keys at once? Will that nifty routine that writes the data to the hard drive still work when the thing is sharing a 6 outlet strip with a hot dog warmer and a bun toaster? And 10 years ago, all we had was MS-DOG. No nice three-phase-commit databases. A convenience store couldn't afford the Oracle license anyway. So give up the silly notion that a POS system can be written by a 10 year old.

POS is literally the most difficult thing in the world to write. You can make everybody happy, in which case you have a program executable over 100 gigabytes in size. :-) Or, you can tell your customers that you can't help them, in which case they don't love you enough to send dollars your way.

Re:What POS software will they run?

[PD]

I just thought of something else to add to my comment:

Besides the owner requirement, there's also the issue of government requirements - TAXES! Yes, they are complicted to calculate.

I was in Michigan when I wrote that POS, and they had 4% sales tax (it's higher now). OK, no problem. Just multiply the dollar amount by .04 and add it on, right? NO! You've got to use a tax table.

The first penny is added at 12 cents. Something that costs a dollar has 4 cents tax. So, what tax does something that costs $1.12 have? 4 cents. The next penny is at $1.13. So, what might the tax be for something that costs $2.12? 9 cents. That's right, that next penny is added at the 12 cent break for something over $2.

This variation might continue in the same pattern for any amount, or the law might say that anything past $100 is just taxed with the same pattern as $1. Or, the law might specify that the pattern just starts over as if the amount was (amount - $100). Or the law might specify that the pattern above $100 is a flat percentage, with specific rounding rules. Or....

I haven't even gotten into the complex patterns that can arise when the tax is 6.125% or some other weird number. I would not be shocked to see a tax percentage of PI*2 somewhere.

There's 50 states. Probably thousands of cities across the country have their own taxes. Things get very scary, very quickly. No way can that be "whipped up" in a night of hacking.

Oh, and let me go on a moment on the joys of calculating prices with 3 figures after the decimal point. The smallest value we have is a penny, a hundreth of a dollar. So why would you need to calculate THREE places? Well, ever notice that gasoline is sold with that ubiquitous "nine tenths of a cent" tacked onto the end? $1.409 a gallon or some crap like that. And that's not bullshit - they really calculate that and you pay it, rounded up at the end.

God, I could go on for years about POS and how hard it is to write.

Good for linux, but kind of insulting no?

[Dock]

I definately think this is a good thing for linux but how is something that's over 10 years old an up-start OS? It's good PR for sure, and definately proving something we've already known, but describing something with such lax accuracy isn't going to help.

On a side note here I ran across about 10 devices like this last year that stored data and operated with cash registers. I don't remember what OS they had but it was probably some novell or DOS mix. The hardware was minimal, a single PCI slot, 12MB of memory, 800MB hard drives and all non-replacable AMB processors (probably around or under 100mhz, I can't remember).

They had floppy drives and I managed to get slackware running on one of them, but I couldn't get the internal NIC disabled (I put a NE2000 in the PCI slot), so I eventually trashed them. Let's see them try to get windows on those things :P

Re:Good for linux, but kind of insulting no?

[unitron]

What was wrong with the internal NIC and why did you feel compelled to feed the landfill instead of seeing how big a hard drive it would take and figuring out something cool to use them for?

Way with Words

[donnacha]

plans to convert its... cash registers in more than 2,500 stores to the upstart operating system

Jesus, "upstart" operating system, is there any way they could make this sound more terrifying to corporate America?

Re:Way with Words

[sharkey]

Jesus, "upstart" operating system, is there any way they could make this sound more terrifying to corporate America?

...upstart operating system that has been undergoing a pilot program at the Anderson accounting firm?

Where Linux can really shine...

[fmaxwell]

Embedded GUI systems is an area where Linux can shine. The lack of a consistent UI between general-purpose Linux software packages and the sometimes-problematic configuration and administration is simply not an issue in a dedicated machine like a point-of-sale terminal. I expect Microsoft to lose a lot of sales in that arena.

Re:Where Linux can really shine...

[sconeu]

Why on earth would anyone want office or 'net access on a POS system?

Customer: Excuse me, I've been waiting for 15 minutes!
Sales Clerk: Hang on... I've got to frag this dude!

Re:Where Linux can really shine...

[CrayDrygu]

Why on earth would anyone want office or 'net access on a POS system?

Well, at BestBuy (yeah, I work there) our POS systems (which definitely have a dual meaning there) also house the "employee toolkit" -- a collection of intranet web pages for employee use. (They run NT4, the POS software appears to be custom-made for BestBuy as I haven't seen anything like it anywhere else.)

Re:Where Linux can really shine...

[CrayDrygu]

What's on your Intranet? I mean, is it stuff that relates to your job (like looking up product information for a customer)? I guess that could be pretty useful.

It's broken up into a few different sections/sites. RetailZone has quite a bit of product information, very useful for people who want to know weird things like the bus speed of the eMachine they're buying. (Anyone who knows what "bus speed" means shouldn't be buying an eMachine.) It also has a message board of sorts that's not too active, but sometimes you can find good ideas being traded around there. You can also look up other stores' inventory here.

TagZone has information relating more to the job than the products -- current trends in sales, planograms (the diagrams that show where all the products are supposed to go), that sort of stuff.

And in the past couple months, they computerized the Raincheck system, so that's in there too.

Those are the main sites, but there's also the Employee Self-Help site, for reviewing and changing personal information (if your address changes, or you choose a different health plan, etc), the e-Learning site for the lame computer-based training sessions they make us do, and they're phasing it out, but you can also get to the "green screen" from there (telnet session to a mainframe somewhere, no idea what OS it runs though, as it's not really "telnet"... the server interface is entirely custom). The Greenscreen's mostly used for looking up information on product availability.

So yeah, it's pretty useful.

That's how it starts. Not on the desktop.

[crovira]

I hadn't even thought of cash registers and other point-of-sale systems but it figures. They need utter reliability and 100% up-time.

There is no way anybody is going to trust the collection or the handling of cash or credit card transactions to machines that are as virus prone and crashable as anything M$ puts out.

The PATH system of trains between New York and New Jersey uses some M$ box to display information to riders on iys trains and M$ is prominently displayed in all its glory when the big monitors hanging over the platforms get "Blue Screens of Death." Tens of thousands of people ride the system every day. That's GREAT advertising for M$. -NOT!

I wish somebody would replace these with some Linux servers so we riders could get systems we can use and trust.

Re:That's how it starts. Not on the desktop.

[NETHED]

Lets hope the train isn't running on Win9X, otherwise they would have to turn off the train and reboot everytime someone needed to get on or off. (There would be a change in hardware.)

Re:That's how it starts. Not on the desktop.

[sharkey]

There is no way anybody is going to trust the collection or the handling of cash or credit card transactions to machines that are as virus prone and crashable as anything M$ puts out.

Bridgestone/Firestone does just that. Goes hand-in-hand with their tires, I guess.

Re:That's how it starts. Not on the desktop.

[quinto2000]

If they run WinXP, they need to call Microsoft every time someone gets on or off.

Re:That's how it starts. Not on the desktop.

[quinto2000]

It's like getting your freak on. Without the freak.

Re:That's how it starts. Not on the desktop.

[Khazunga]

Believe it or not, once at the London Tube, I experienced a reboot of a subway. When the train stopped in the station, the doors wouldn't open. The driver said on the loudspeaker he was trying to solve the problem, and 30s later, the train completely powered off (engines, lights, a/c). A few seconds, a power up, and voliá... Freedom again :-D

Exactly A Win For Linux

[mikosullivan]

Yes, this is exactly a win for Linux, and for Open Source in general. At this point Open Source is still winning the easy converts. As it gets more popular, it will get the more difficult converts. The more big-business, mission-critical apps there are out there that run on open source the better for the movement as a whole. It's getting harder and harder for MS to insist that open source is a fringe movement and that it can't be trusted. That's a Good Thing for Open Source.

Re:Exactly A Win For Linux

[fireweaver]

So what? It is just this kind of boring infrastructural stuff that holds the world together. If Linux can run a cash register well enough that the company can almost forget that it even -has- cash registers (i.e. high reliability / availability coupled with low upkeep); then this is a big win in the long term.
Other people with cash registers (particularly the Windows-based ones) are going to look at Sherwin-Williams and go "hey, I think those guys are on to something" and at least think about converting when the time comes to do so.

Personally, I would like to see this one succeed.

Re:Exactly A Win For Linux

[nomadic]

Win? What do we win? Money? Prizes? How does Sherwyn-Williams using Linux gain me anything?

Microsoft's Rebuttal

[Surak_Prime]

So, any wagers on how long it will be before Microsoft tries to counter this with a lawsuit against Sherwin Williams based on their trademark on the terms "Paint" and "Paintbrush"?

Invisible Linux.

[vkg]

Really, at the end of the day, nobody gives a shit what OS the POS is running: as long as it's doing it's job, who cares?

Very much like the server market - as long as it works, nobody gives a damn what OS is running.

I'd say this is an ideal niche, and there's no reason to use anything *else* on a POS, is there?

Back-Handed Compliment

[donnacha]

From the article:

But, he said, Linux isn't being asked to do too much high-stress computing here. "It's just a nice, low-cost platform for doing kind of everyday computing."

Sooo... if they actually needed it to do anything other than the computational equivalent of a nice picnic, they would gone for a "serious" OS?

Like Windows?

Re:Back-Handed Compliment

[Bostik]

I noticed the same quote, but consider the wording used...

[...] nice, low-cost platform for doing kind of everyday computing. So in effect they are saying that they consider Linux ready to be used where-ever you need a good, stable and reliable platform to run their applications. Now, isn't this kind of everyday computing just the thing that most users do at home?

I think that quote is indeed very nicely put. It may even prove valuable.

Re:Back-Handed Compliment

[donnacha]

Now, isn't this kind of everyday computing just the thing that most users do at home?

Sure, but the most important focus for Linux right now is, I believe, industrial and embedded use; an open OS used at every level of our society and throughout the supply chain will be of tremendous benefit to everyone on the planet, far into the future.

Home use is pretty small potatoes when you consider the stultifying effect that proprietry, closed software has as a whole.

Re:I love linux as much as next guy, but...

[sharkey]

Just off the top of my head...

multi-user
Cashier login, "su" to manager for overrides?

preemptive multitasking
Sorry sir, you'll have to wait 5-10 minutes for your change. The activity logging routine is running, and we can't interrupt it.

protected memory

Whoops, make that 20-30 minutes. My register just crashed.

Just my, admittedly not expert, two cents.

Re:For our young geeks...

[autechre]

According to some employees, it also stands for (or used to, don't know if it's still this way) "I've Been Moved", after the company's tendency to transfer employees all over creation.

Remember when Silicon Graphics Incorporated actually changed their name to "sgi"?

More details

[acordes]

In addition, the International Business Machines equipment won't use Advanced Micro Designs processors, but will be compatible with Transmission Control Protocol/Internet Protocol. No comment was made by the National Aeronautics and Space Administration or Minnesota Mining and Manufacturing.

Re:More details

[Guy Harris]

In addition, the International Business Machines equipment won't use Advanced Micro Designs processors

Well, yeah, I wouldn't expect Sherman-Williams to be too happy about Duron in their stores.

Re:More details

[adolf]

Who is this Advanced Micro Designs company?

All of the Durons I've seen were made by Advanced Micro Devices.

Sorry to cut this post short, but my 1993-vintage post-American Telegraph and Telephone buyout National Cash Register-manufactured laptop is in need of a nap.

Lowes Uses Linux

[RageMachine]

The hardware store Lowes has used Linux for sometime now, on the machine they use to take job applications. The machine is an IBM running an X system with some modified version of Netscape. The GUI is a Motif type environment (you can tell just by the buttons). I took my ex-gf down there one day so she could apply, and I watched as she went through the screen. The cursor is just an X, indicating its just a bare X system... Basicly.

RadioShack

[NETHED]

I worked at RadioShack for a brief stint and this is what I saw at the POS system.

a Tandy machine (no idea on speed) running Win95. It crashed REGULARLY. It was fun watching Scandisk do its thing while a customer is waiting.

The interface is a custom app that pulls its inventory data off of the 'server' sitting in the back room. To do this, each POS was networked to the 'server' in the back room. For some reason, each POS also had Serv-U FTP server running on boot. There was no cashdrawer interface as the cashdrawer was a SINGLE wooden drawer behind the desk with a 'fingercode' access inhibitor. All you needed to do was pull with your middle finger pulling the most.

IF anyone else worked at RS, tell us about it, i'm curious about the current RS situation.

Re:RadioShack

[tweakt]

"Server" in the back is a *NIX box. I remember one time the manager "upgraded the system". This consisted of software updates from HQ off tape.

Re:RadioShack

[ddent]

buahahaha. Are you serious about the cashdrawer security method? I've never heard of that... its quite funny.

Is so!

[The Pim]

It's a "sher" win.

Linux migration in the financial industry...

[rakeswell]

...is perhaps more interesting and is becoming more widespread.

Reuters recently announced that it's market news aggregation system (RMDS) is being ported to run on Linux servers. The system currently is running on Solaris and was ported to Windows (but the Windows port is no longer support/persued)

This is just the latest example of the financial industry turning to Linux. Morgan Stanely, Credit Suisse, E-Trade, the NYSE have all started to move to Linux.

It's true that the migrations are generally coming out of the hide of Solaris and AIX. IBM is coming to terms with Linux, and recent signs look hopeful that Sun will follow suite as well.

I suspect that the economy has had a hand to play in the receptiveness of the big players in the financial industry to start looking to Linux-based solutions: everyone is looking to save money right now, and I think it's no accident that the financial industry seems to be taking the lead in terms of being early adopters of Linux in the enterprise.

I can only hope that with the trend towards moving systems over to Linux, these business will be exposed to open source ideals, which -- who knows -- might one day lead to MSFTs fall from dominance.

Isn't it plausible that while Linux may be eating Unix's lunch, this gives it a better chance to spread open source/free software ideals in a new environment, which -- in the long run -- might be what takes the *big* chunk out of MSFT's hide...

IBM, Linux and business

[Mr. Firewall]

This is a direct result of IBM's billion-dollar commitment to Linux last year, and Sherwin-Williams isn't the only one.

Just last week I participted in a rollout for Sears Optical (the little department inside of Sears stores that does eye exams and sells glasses, etc).

The hardware was IBM. The OS was Linux.

According to a friend-of-a-friend who is an IBM rep, IBM has already gotten their billion dollars back in increased sales, and is now ready to pump ANOTHER billion into Linux!

Politics surely does make strange bedfellows. Seems it was only a few years ago that we were calling IBM the 'evil empire' and now all of a sudden they're on our side.

Re:For our young geeks...

[aebrain]

Around when IBM sold Business Machines?. Yes

Even managed to kludge some hardware together to drive an IBM Golfball typewriter from my Exidy Sorcerer , which at 2.1 Mhz clockrate was the fastest gun in the west. In 1978 that is. Pre-IBM-PC. Pre-Mac. Contemporary with the TRS-80 Model 1 , the Commodore PET and the Apple II. Just have a look at the Old Computer Museum reference.

So just remember that one day, arguments about RedHat vs Debian will be considered "quaint", as the newest alphageek-wannabes argue shrilly about direct-neural-induction vs alphawave-heterodyning on the new Petaflop quantum-Beowulf-cluster-wearables.

While old codgers like me will still be trying to stop said wearables from having the usual code bloat and buffer overflows caused by AOL-Time-Warner-CNN-MicroSoft-General Motors-Unilever-Bell-Boeing-PepsiCo 31337 hackers rather than Software Engineers.

Re:Success stories?

[Anonymous Coward]

Dunno if it's a success story but here is a link to a similar roll-out of linux as a POS.

Burlington Coat Factory

sushi places in Vancouver

[ddent]

On a related note, there is not a single sushi place with a computerized billing/ordering system in the Vancouver area that is not running linux. It seems that one vendor had a multi-lingual program which is well suited to sushi places, because they all seem to use the same program. As well, there is a screensaver with tux and the name of some consulting company which seems to set them up.

Re:Oh Really?

[CrayDrygu]

Best Buy uses Windows NT 4 for ALL there POS systems in the company (over 500 US stores). Rarely do they experience problems.

Heh, you obviously don't work there. I once managed to crash a POS by swiping a customer's credit card through the reader on the keyboard when it should have been swiped through the one facing the customer (same pad you sign on). And it wasn't just a fluke, because I could recreate it.

So admittedly they don't crash often, but when they do, it's especially annoying because the P166's take about two minutes to boot, and then the POS application takes about five minutes to load. "Oops... sorry sir, the computer crashed. Let me take you over to the next register..."

This is how it starts

[ctid]

I've often felt that Linux is ideal for jobs where only one application is run. Then you don't need to worry about training people to use Linux, because all they see is the application. If it's simple enough (eg barcode scanning prices), the user doesn't know what OS is running anyway. I know it's stating the obvious, but this is true of servers of all kinds too; they do one thing (OK each service running on the server does one thing) and the user has no idea how that is happening.

What will be interesting is how the support structure pans out. Everyone knows that you need staff to support your servers. So if they're running Linux, you need someone who is competent in that. But since you've got to have this person anyway, surely they could do some support of the desktop machines. For example, those which are only used for word processing. So long as the user isn't going to be installing new software, or switching between multiple applications, who cares what OS your WP program runs on?

The odd thing about this is that people say that Linux is OK for sophisticated users and not for the newbies. I'm inclined to think that it's the other way around. So long as your user has to use the same application every day, and doesn't get the opportunity to change things, Linux has to be better (file formats permitting of course), because it's cheaper. So it's the unsophisticated users who can be switched to Linux first (as proved by the POS successes; you don't get any more unsophisticated than swiping past a barcode reader), because they see less of the OS than the sophisticated users.

Re:Way with Words: 'upstart' linux

[peddrenth]

Simple way to address that... just start using "a convicted monopolist" to describe microsoft in everyday conversation. It'll soon catch on!

Comment removed

[account_deleted]

Comment removed based on user account deletion

Doing embedded apps not as easy as you'd think...

[Svartalf]

You need a comm/network stack. If you don't use an OS, you end up rolling that yourself. That's NOT as easy as it sounds.

You need a GUI stack unless you're doing a simple register. If you don't use an OS with an app framework, you end up rolling that yourself. That's NOT as easy as it sounds.

Most people will choose to go with an embedded OS and GUI setup of some kind. The "thin" ones won't give you what you need (you don't want the daily reports locking up the machine- you might get that if you don't use a more advanced machine...) and the ones that DO give you what you need are pricey (Read QNX, for example...).

Linux is a good fit in this sort of role.

So what?

[Svartalf]

The desktop market is the same way, believe it or not. Nobody truly gives a damn so long as it works- it's just that MS has convinced people that Windows is "easier" to use, which it really isn't.

Oh boy!

[Junks Jerzey]

This is going to majorly affect my life as a geek! Now I can convert all my friends to Linux, because they all love Sherwin-Williams!

Seriously, I don't know how any of this matters to anyone. UNIX has always been a behind-the-scenes OS, and Linux is certainly not the underground geeks-only OS that many geeks so badly want it to be, so why does this matter at all?

I feel your pain, brother.

[Ilan Volow]

From 1998-1999, I worked at a Radio Shack in NC. We were actually one of the first stores in the country to go from SCO unix to Win95 POS.

The win95 machines were somewhere in the region of pentium 75-133's. Yes, Radio Shack management was dumb enough to buy their hardware and then wait two years before shipping out the system with the software. One of the silliest things about the "upgrade" is that it was really nothing more than making almost an exact carbon copy of the curses-type SCO interface in a Windows 95 GUI, essentially replacing CLI text fields with identical GUI text boxes that really don't take advantage of the GUI paradigm. Just because you make a bad interface pointable and clickable doesn't mean you've made the interface that much better (in his book GUI Bloopers, author Jeff Johnson refers to this as a "TTY problem"). Our manager was discouraged by Tandy technical support from calling a bug a bug. He was told to call it an "issue".

One of the silliest things I remember about the radio shack machines is that none of them had a cd-rom drive. Guess what we had to do if we needed to look at the Tandy catalog CD-ROM? We needed to get the key to open up one of the cabinets for the display computers (the ones sitting above the fake computer shells they use to demo the latest models), type in the password to stop the demo, stick in the CD-ROM to get what we wanted, restart the demo when we were done, relock the cabinet, and then finally put the keys way. There was not a whole lot of incentive to get out the CD-ROM when it wasted time we could have used to earn the commision necessary to put ourselves over minimum wage.

Whenever I hear of Radio Shack being called "America's Technology store" I laugh heartily.

Re:For our young geeks...

[sketchkid]

i believe this is where i am supposed to show repect for my elders

(there's no double meaning here, just face value)