Slashdot Mirror
Kenwood Chooses Linux Over NT for ERP
Structured Audio writes "Kenwood America (the speaker maker) has moved from Pick to Linux for its enterprise resource planning
apps. This Techweb article explains why Kenwood chose Linux over NT or a commercial Unix." ERP and financials are among the most important areas for Linux and open source to shine if they are to be accepted by the corporate world.
And no, a traceroute to my ip does not end in Redmond ;-)
The whole thing about 6-8 hours for w aindows box and vitrualy instant for Linux...
That was comparing their legacy mini to a Linux server, both doing data import/export to Windows boxen.
The 2-four processor thing is actually QUITE believable.
Kenwood is hardly a loser company. Its product line includes a complete line of communications for industrial applications, which is a very profitable market. Considering their quality over the latest Motorola products that I have to use, Kenwood looks very promising. I work in an industrial environment which requires that can deliver a clear voice. Its not much to ask, but the $450 radios from their competitor are a pain to hear.
Some links to ERP vendors and what they have to say about Linux support:
(Check out the cool penguin
So, answering my previous question, it seems like Kenwood is limited to choose between SAP and SAP.
Hi!
Exactly the right attitude - can World Domination (TM) be far behind? :-)
"I will take the Ring," he said, "though I do not know the way."
Exactly. Maybe having "losing" companies embrace Linux is the best thing. If companies are using NT and sinking, and then they convert to Linux and become successful, then maybe other companies won't wait for NT to sink them :)
-Brent--
Probably not. No one is *forcing* you to read it either. I wouldn't be posting this if I wasn't interested. If it bores you, when you see it on the front page, sigh, and move on to the next story.
Linux has passed the novelty phase.Stories like this are like awards. Linux is a real player now. Some people want to hear about companies using it.
The folks at Microsoft must laugh when they read stuff like this.I am afraid that you'll have to explain this one. Microsoft laughing because they lost business? Because Linux is overtaking NT? I think that Microsoft would be desperate to cover up all stories about companies deploying Linux.
Oh, you seem to want to cover up all stories about companies deploying Linux. Maybe that explains it. Are you Microsoft?
-Brent--
Back in January, a /. story lead me to this article containing a prize anecdote:
;)
Those well worn issues of how one make money with open source technology were batted back and forth and Richard always won the argument. Gerry, IBI's CEO, said at one point that SAP, the second largest software company in the world, does not give away its software for free, and it never will. SAP customers pay lots of money to buy their software and don't want it to be free. Richard responds by saying that he is going to write a GPL'ed version of the software SAP sells. It will take time, but there will be a freely, source code distributeable version available sometime in the future. How can you argue with that.
When I read that, I wanted to kiss his pate and laugh. At the time, I was one of three in our company charged with finding an ERP package to replace all our RS/400 OLTP and PC business intelligence applications. It would then fall to me to implement these changes company wide in 90 days. We were left with choosing between the lesser of five evils. Why? Our budget required that the software licenses be within US$200,000 per 30 users. Final cost for the project worked out to US$417,000 per 30 users.
My response in the thread was essentially "FreeSAP: Now that's the work of a patriot! Is this a declared project? Where do I sign up and what can I contribute (equipment, money, industrial eng research, code, APICS pulp, blood)?"
In reply, AC summed-up the problem:
I found his comments about writing GPL'd SAP clone software to be hilarious. He can't be serious. Just who does he think is going to write all that code? I mean, the free software community considers the Linux kernel and Mozilla to be "large" projects, both of which are trivial compared to writing software to compete with SAP from scratch. He would have to mobilize an army of free software developers, most of which would have no personal interest in or use for the kind of software SAP sells.
A clarification was submitted by RMS.
My point in this is to reaffirm that no Open Source software venue, aside from the NOS, could be of greater benefit than OS ERP applications. In the original 'Request for Comment and Quote' to ERP vendors, I listed 300+ functions that were essential to our business. The winning package checked-off all but two. In fact, only about seventy of those functions are handled meaning I had to write separate applications and find little better than compromises. In so many cases the goal could be met with a SQL trigger but all the application logic is in the client app!
Frankly, I'm awfully tired. I know that closed-source ERP applications have cost many others as dearly; I believe that hurts our economy broadly.
The ISPs I know all use some BSD flavor for their mail and DNS services. I just can't imagine some Linux flavor ever standing head and shoulders above all other contenders in this area.
(Maybe someday I'll figure out what this funny "IT" thing means. I think it's what we used to call what the mindless data-processing drones did, and whom we all despised. It's certainly no more related to computer science than my cat is. Sounds like some kind of business degree, or sentence.)
IBM to role out Win2k Company Wide From the we-are-bigots department
Did you even read the article? Here are some quotes:
Although IBM plans to deploy several copies of Win2K Pro as its standard internal desktop, the company won't place Win2K Pro on every desktop.
IBM will not perform an extensive internal deployment of Win2K Server. According to Sullivan, IBM is primarily using its servers, not PC servers. IBM relies on large, clustered servers, like RISC System 3000s and doesn't plan to do much domain reorganization.
Like a lot of customers, we decided to wait and not get [NT] 4.0,
When IBM was deciding on a desktop OS, the company didn't use total cost of ownership (TCO) extensively as a determining factor.
TCO was not necessarily a major factor for us. To achieve a lot of Win2K Professional's benefits, you need Win2K Server," said Sullivan. Because IBM is not using Win2K Server, it wouldn't see those benefits.
This is actually quite interesting, and not particularly good for Microsoft. IBM is NOT using Win2K server, and many of the benefits(sic) of Win2K Pro are tied to the use of Microsoft's server. So if you decide that the srever is not your cup of tea, it suggests that TCO may not justify upgrading to Win2KPro on your desktop.
If this is the best the Win NT zealots can come up with, well, I am not impressed.
First of all, Perl is not a `product'. That's MBA talk. Perl is a program. Sometimes it's a toy. Sometimes it's a joy. Sometimes it's a way of life. :-)
What is seeping through around the edges is the same discomfort you would see if you stuck a quiet math major into a vat full of business majors or jocks.
As a kid, I always threw out the sports section and the business section of the paper, sight unseen. It was the science and technology sections, and magazine, which appealed to me. Never in my wildest dreams did I imagine that I'd now find more the stuff about computers in the business section than the science section. And yes, I do keep the business section now -- it's got Dilbert on the front page. :-)
You'll find that amongst hard-core programmers--those for whom programming is a game, computers a toy, and who see computer science as having no more to do with computers than has astronomy to do with telescopes--the whole business thing is often an annoyance. Experience constantly shows that programmers as a group mistrust business people in general, especially marketing and sales types, whose job programmers perceive as lying. If you expect to be taken seriously by a programmer, you don't show up in a suit and tie. This strong mutual dislike makes it very hard to run certain kinds of business, but keeping engineering in a different building than marketing certainly helps. :-)
There are some folks who do computers because they're trying to make a buck. You know the types. "MIS majors", people who are on their way to MBAs. They seem computers as a means to an end: money. These are folks whose idea of a career in computers leads to some sales job or managerial position.
Real programmers also see computers as a means to an end. It's a complete different end. And most of us think it's just nifty that we get paid to do what we'd just be doing anyway for fun. Yes, the different end is to have fun programming. It's not something I expect business people to empathize with very much.
Remember the old joke:
If you haven't known "IT pro-fess-ion-als" who were mindless drones, then you are a lucky, lucky man. We used to call them secretaries, librarians, or data entry clerks. Now they're "IT professionals". Give me a break.ERP is one of the prefect applications for open source since, as you said, so much of the logic is in the implementation.. and more money then I care to think about has been lost due to the lack of an open source ERP package.
Ironically, the soultion to one of the biggest problems with ERP packages could be exactly what would make an open source ERP package's development possible.. modularisation.
My limited experence with ERP packages is that they are these monserous programs which do a few things you need, so you must write your own system for the critical part and take advantage of the ERP program for some steps. It seems to me that an orginized effort to develop the necissary modules could provide us with a library that would get anyone who needed ERP a large part of the way to a working system.
I suspect that the cost of writing more code vs. implementing more SAP work-arrounds would eventually even out.. and the resulting product would be MUCH better suited to the needs of it's users.
The big question here is what should the modules. It is pretty straight forward to seperate some parts of the process like international tax handling, but some other parts may require a little more though as to how they sould be seperated from the rest of the system. It might be a reasonable vaporware open-source to just try and figure out what needs to be done and make a list of jobs?
I suspect that it would be extreamly profitable for a company to donate employies to wrok on bits of an open source ERP package because those developers will eventually be the best imaginable implementors of your companies ERP system.
Jeff
The Christian religion has been and still is the principal enemy of moral progress in the world. -- Bertrand Russell
Sounds like good news for Microsoft to me. Nobody would expect a company that produces its own servers like IBM to move to Microsoft's flavor. On the other hand, the fact that an anti-Microsoft company like IBM is already committed to Win2K Pro on the desktop says a lot about how big Win2K Pro is going to be on corporate desktops. I use it on my ThinkPad (with Win2K, NT4, Solaris/SPARC 7, and RH5.2/6.1 on my servers), and I've got to say that no other OS comes close to matching the quality of the "user experience," for lack of a better term.
Just as slapping the Win95 GUI onto NT resulted in a growth explosion for NT4 Server in business, the migration to Win2K on desktops and laptops will ensure a strong role for Win2K on the servers, because Microsoft really has done some impressive things for TCO and administration when the two are used in combination.
Frankly, whether an IT head decides to use Win2K on the server or not, if he sticks with Win98 clients instead of going to Win2k Pro, he really ought to have his head examined. It really is that good, and I haven't even tried anything beyond RC2 yet.
Cheers,
ZicoKnows@hotmail.com
I'm not sure jBASE qualifies as ERP, a term referring to software covering a whole range of business processes and usually requiring some business process re-engineering in order to implement it. ERP vendors include SAP, Baan, J.D. Edwards, etc.
Having said that, it's still an impressive win for Linux.
It seems to me that this is a very brave move by Kenwood. Quoting from the article:
In other words, they first decided on Linux and only then they try to find a software package. Very brave, given the uneven support for Linux among the vendors.
What ERP packages support Linux?
Is this the start of a new trend? Companies goes Linux for the OS and the software applications will have to follow? That would be a Good Thing.
Hi!
Bruce
Bruce Perens.
Oracle is the real enemy and is causing most ERP vendors to suffer. It's about time the open source community does more to fight the real enemy--Oracle.
PostgreSQL is a very serious open source SQL database project that is making a lot of good progress. Over time it is becoming a very capable product.
If the ERP vendors are suffering due to Oracle, they should put some money into PostgresSQL development.
On the other hand, the fact that an anti-Microsoft company like IBM is already committed to Win2K Pro on the desktop says a lot about how big Win2K Pro is going to be on corporate desktops.
I have no doubt that Win2K is going to be huge on desktops. What other choice is there but Windows? Windows really is a monopoly as far as desktop client OS's is concerned. Exactly what alternatives does IBM have?
if he sticks with Win98 clients instead of going to Win2k Pro, he really ought to have his head examined
If you read the article, you will see that IBM, like many businesses never deployed W98. They are doing their first OS update in 5 or so years.
The interesting comments were that IBM is not deploying W2K servers, that they are not justifying the changeover on a TCO basis (Microsoft keeps blowing that horn in their sales literature), they are not switching all of their desktops over to W2K, and so on. This is very different from the impression that the original poster was trying to make with his "W2K all around" at IBM posting.
In reply to your comments about the various ERP vendors supporting Linux, let me add the following about Peoplesoft (yes, they're my employer):
...
...
PeopleSoft currently supports the use of Linux as a DB server as long as you're running a certified version of your RDBMS.
From what I know (which may not be much as I work in a completely different area) we are planning to fully support Linux once we get a proper version of Tuxedo on Linux (the current one still seems to have some technical issues)
No, this is not an official Peoplesoft statement; this is my view of Peoplesoft's Linux position
It's important because of familiarity. As you so rightly say about BSD, more familiar people are with something, the more likely they are to specify it. And businesspeople have the power of the purse, so their first idea is to use a system similar to what they have on the desktop: Windows.
That's the reason for the rise of Windows in the server space. A businessperson probably knows a drone-style-programmer who can do VB, and ASP is basically moving VB to the server.
The only way to counteract this is to encourage Unix in the desktop, in my view. Make people more familiar with it, and they're more likely to specify it when the time comes to do a server-based system.
D
----
One of these misunderstandings is about "BSD's AT&T roots". I hope you understand that BSD doesn't have any AT&T code. You're thinking of the "bogging down" of the AT&T-fuelled FUD during the lawsuit days. What BSD does have is decades worth of highyl refined software engineering. You just don't have that in Linux, or most of it. Sometimes it shows. But time should smooth even those rough edges, and faster this time around.
The next misunderstanding is about the Linuxes having "all their own code". What they have, by and large, is rather new code. Sometimes this is good, and sometimes it isn't. And the part of their code which isn't new tends to derive from the BSD project, anyway. :-) And the Linux kernel, at least, certainly has Bell Labs roots: the API is still Unix, just like BSD and so many other operating systems variously loved and hated.
You're right in that it doesn't matter which flavor of Unix wins. Unix is Unix. Granted, I'd personally be a lot less happy stuck on AIX, Irix, or HP/UX than I would on BSD, Linux, or SunOS (in that order), but it would be infinitely better than any of the populist non-Unix systems that are out there. I'm sure we'd all agree on that.
But wait a second. Wins what? I'm trying to figure out what is meant by this whole "Battle of the Desktop" notion. Since I've always run nothing but Unix on my desktop, and so have my friends, and so do nearly everybody I nkow from the government, research, and educational labs in my immediate vicinity, this has a very odd sound.
What appears to be meant here by "desktop" is the desktop of people who don't program, people who by and large don't really understand or like computers. Certainly Unix provides a more programmer-friendly desktop environment, but a lot of folks aren't friendly with programmers. But it also provides a flexible substructure with which to apply idiot-friendly veneers. Of course, programmers have never seen much point in doing this. But idiots, users, and other non-(computer-)professionals have money, and so people who are trying to sell systems to idiots have a vested interest in doing this.
I'm not entirely sure what I think of this idea. I don't know about the real viability of this whole "make the masses use Unix" idea. I could argue most any direction on it. Yes, it's a better base. But Unix is a professional system, made by programmers for programmers. I'm not sure that non-programmers are ever going to be enamoured of our idiosyncratic charms.
But you know what? I don't really care all that much about what non-programmers are doing. Like most people, my initial concerns are for my own world. Is it something I'm familiar with? Can I get what I want to do done? Does it make me happy? Is there a risk of that going away? I'm sure the non-programmers feel the same way.
So when you say "the battle of the desktop", you're not talking about my desktop. I'm running Unix in various flavors there. I always have, and I can't imagine that changing. No, You're actually talking about the attempt to imitate a {non,anti}-Unix system enough to make non-programmers happy. To do that, you need to make sure that they can get done what they want to get done; you need to make something that they're familiar with, which means you have to copy Microsoft. I'm not very interested in that, but I can see why some people are, whether for economic or geek-ego reasons -- which, mind you, I'm not saying are bad.
Finally, as for "Linux having more for the desktop than BSD", I've no idea what you mean there. My only guess, which is probably wrong, is that you might mean "Linux is better at making non-programmers feel like their still running Microsoft stuff?". If so, I don't know why programmers care, so that must not be what you mean.
I've got both systems various BSDs and various Linuxes. Both classes of machine run exactly the same things. I don't know what you've got that runs on Linuxes that doesn't run on BSD, but it must not be something that I care about, because I've never noticed such a thing. :-)
For me, the BSD systems end up being a lot easier to manage (sidenote: nonprogrammers often can't even manage their own general-purpose computers, and issue that can't be underestimated) than the Linuxes are, and are in general a lot more comfortable for me personally. To my critical eye, they're more refined, robust, integrated, complete, and friendly. For me. I'm sure the Microsoft people say the same things about their systems. Familiarity is important, and I've been playing with BSD for longer than some of you have been alive. I've seen it develop and grow, evolve through decades of hard work by smart people constantly making it better. Because I've watched this happen, it's all very familiar to me. That may be the most dominating factor here.
The last point is, I think, interesting. Quoting from the article:
So, I wonder, is DNS and e-mail the ultimate "killer apps" for Linux that will finally propel it into the IT departments across United States and the rest of the world?
Hi!
See also the article The "Lintel" Value Proposition on the same site.
It makes an interesting read and basically argues the same point: you get more "bang for your hardware buck" with Linux.
Good stuff; should be required reading by all IT decision makers.
Hi!
My theory is that we can continue using the command line stuff as long as there is a Unix, and the people who hate command lines can use the GUI.
I didn't imply that the GUI would blend together seamlessly with the command line, or that you could easily automate GUI stuff. But you can use the existing interfaces to automate tasks.
The big problem I see in the present state of computing is that we're making simple things complex in a quest to make complex things simple. In the end, simple winds up not even being an option, and that's an enormous regression from the Unix way.
For instance, C programmers under Unix can do a three line plus include files hello, world. It will be command line only, but it gets people started. C++ programmers in Windows use a Wizard to generate a 500 line mess that nobody understands. I find it mind-boggling, but my sense is that most people learning programming on Windows don't understand what the hello, world windows program works or what it does. This is bound to result in a brick wall and far less of an appetite for programming in the future.
I suspect that's a major reason for your desire to not even learn Windows as we see it. And I must agree with you, even though I've been doing all too much in the way of Windows stuff over the past few years. I don't particularly want to understand it either, and that inevitably limits my skills and chances for advancement. I'm hoping the recent rise of Unix/Linux will fix this for me - I'm getting an increasing number of Linux jobs.
D
----
Wow. Seems like, regardless whether there was any truth behind the Mindcraft "vindication" (and IMNSHO it is valid, much as I personally wish it wasn't), Linux still wins in more typical situations than Mindcraft used in their tests.
I remember a prof from my undergraduate years who took a serious amount of salt to any "benchmark" result or any single form of CPU performance measurement. His argument is simple: benchmarks, CPU measurements (like MIPS, MHz, etc.) are basically sticking a single number to complex system/device. There are so many parameters involved that sticking a number to something in this way is like collapsing a high-dimensional object into 1D and using that as a measuring stick.
This is the case with the Mindcraft benchmark (and any other benchmark for that matter) -- their results simply show that in a certain environment, under a certain configuration, Linux loses to NT. OT1H this simply means Linux has room for improvement, OTOH it says (close to) nothing about how Linux performs in other environments. And then you've got to take into account all sorts of other factors, like the cost of maintenance, the minimal required hardware, etc., that benchmark results hardly begin to take into consideration.
Well, my point is, while we're working on improving Linux so that it won't lose to NT in environments like Mindcraft's benchmark, we can rest assured that in general, Linux is better than NT, in terms of performance under typical environments, cost of ownership/maintenance, and giving desired performance on minimal hardware, etc.. This case simply proves this point. We've had many reports of this sort in the past, and I'm sure we'll be getting a lot more as people begin to realize the advantages of Linux.
As for FUD... slowly, as more and more people step out of the M$ realm and discovers that the "outside world" is not exactly like M$ would like to have them believe, I think FUD will eventually just become CCA (confidence, certainty, assurance).
mikre he sophia he tou Mikrosophou.
Interesting how this works.
Someone wants to quickly set up a DNS server, so they grab a rusty PC and install Linux, instead of buying a pricy UNIX machine or buying a copy of Windows NT.
This seems lets people get aquainted to Linux and consider it for larger applications.
I've seen this happens elsewhere too.
Pick was originally developed as a business information tool that ran on bare hardware. Ports took months and were only available for a very few platforms. About 10 years ago, Pick and a number of clones (Ardent UniVerse, etc) began porting to Unix. In the past five years they ported to NT. Recently they ported to Linux.
The Pick environment runs as a shell on these Ports.
The problem with legacy Pick (like McDonnell Douglas) is that there is no connectivity to the outside world. The only way to get data into and out of it is via 9-track tape and serial port, typically. Any modern Pick shell running on Unix or NT would beat this easily. Can you say samba?
One of the features of Pick is it non-first-normal form database structure, using strings with delimiters to store all data. This is extremely flexible, and something that Kenwood is not going to leave. jBase incorporates this structure. They will probably be able to leave their application software largely unaltered.
Pick does have an interesting similarity to Linux. The early ports to Intel hardware had 8 serial lines running off a 286, blowing away DOS and Windows for speed.
Another interesting bit of trivia - this ancient McDonnell Douglas system incorporates the date windowing that was recently the subject of a patent. This isn't a recent development - they've been using this windowing for 25 years!
More info at the Pick FAQ
That's why I searched around their site looking at their radios. I found lots of neat products that connect to almost anything, such as portable radios with integrated digital cameras. Fun stuff!
The reason a large percentage of software today is so bloated is because the companies behind them are basically there for the $$$ rather than technical excellence. Yes, some might have technical excellence as a secondary goal, but basically, the bottom-line is $$$. I've always believed that:
When your primary goal is $$$, chances are that you wouldn't mind compromising a little quality to get the product out the door fast, to beat your competitors. A little bloat here, a little unnecessary feature there, that won't hurt, we'll just ship a bigger (and therefore "better") product, with a few extra features that our marketing folks can boast of to our customers (but which they really don't need). Hence, bloatware. OTOH, the beauty of Open Source is that the people who contribute do so because they are interested in making it better, not because they have to get it done so that they can feed their families. Hence, they will not compromise with bloat and unnecessary features.
mikre he sophia he tou Mikrosophou.