Ok, lets have a poll! Maybe the result will help convince Baan to release a Linux porting set. The one that is floating around was 6.1c.03, kindly provided by some gurus in Baan Tools Development, and worked perfectly, even under the newer Linuxes (yes, I got it working :D ), like SuSe 7. So there is a good base to start with!

Let's hear it!

There's been much discussion already about the Linux portingset. It's a definite must for Baan to do this as Linux will grow to become a dominating server OS.

Yes Pat, this is absolute right!

Many Baan customers want to use Intel hardware and at the same time use their Unix know-how. Linux is growing with incredible speed and so the the linux porting set is a must for Baan!

Linux becomes more and more popular so it is no surprise Baan should consider it :)

I heard that a Linux Porting set was going to be released with Baan 6.. (and only Baan 6)

Here's my question though, how should it be released? should they make it only for RedHat? What about SuSE, Debian, Slackware, etc? How about the *bsd's? Once you get into BSD's doesn't that open the door for Mac OS X Server? (based on BSD unix).

Or should they provide source for the porting set, that you have to compile upon reciept? I doubt that would happen. Or maybe they should write the porting set in java? That seems like a bad idea...

Its been a while since i've played with linux, maybe its more "standard" now.. But It seems like the best they're going to be able to do is say "RedHat - GCC 2.6" and that's it.. I don't know if that's better or worse than not doing Linux at all...

Dave

And when do they plan Baan 6 issue? I've heard only 3-rd quarter of 2003? Although we ourselves still using IVc4 :)

So, by giving Linux only in Baan6, do they want to force us to upgrade? Typical Baan! The main reason for a Linux porting set would be the small and midrange market, which is for the biggest part running (quite stable) on Baan4. And they would want Linux to cut costs, not increase by another conversion/migration with possible partly reimplementation!

imho the main problems are:
1) Which platform to support (linux is too scattered)
2) On BaanIV: How can we make a portingset stable enough to be compatible with the mature Unix portingsets
3) Who pays for the portingset development, may it be sold as a separate product or should it be payed for by the innovation fee
4) Is there enough "body in the market": customers who demand it and knowledge to support it.

The point of view of M$ might make a difference, but on the other hand, they bought Navision too...

I agree (i)Baan6 & Linux does not make sense.

I think a lot of current MS boxes would run Linux if there was a portingset available. Because of the cheap replacement parts (hardware) and OS costs.

Hi OmeLuuk,

I think the costs for the linux porting set can not be the problem for a company like baan. The problem is that they don`t will it.

Maybe the questions in the poll should be more specific and changed to:

Linux-Portingset for Baan IV (yes/no)
Linux-Portingset for Baan V (yes/no)
Linux-Portingset for Baan VI (yes/no)

We are interested in the Portingset for Baan IV

regards

Dieter

OmeLuuk : I agree (i)Baan6 & Linux does not make sense.... at least not for the current installed base. But from what I learn from Gemini is that is is supposed to be an attractive alternative for the current BaanIV users... in that case Linux may be one of the minor extra arguments to migrate.

As I learned, Baan indeed plans to release a linux porting set only for gemini.
I hope that some people from Baan germany read this forum and consider the following:
Many companies in germany are very keen on a linux porting set to install (perhaps additional) testsystem(s).
You (Baan) could create immediate sales by selling this porting set.
Releasing it only for Baan 6 will be like not releasing it at all.
How many customers are the with Baan4 or even older releases?
I would guess that more than 60% of all Baan installations in germany are < Baan 5.
If you want to keep on living in a dream world, simply go on ignoring your customers and their needs. For sure they will all come then and buy gemini.
As for the linux flavour: there is this movement call united linux, where nearly all distributors of linux packages have agreed on a common standard. So there shouldn't be a problem to release a porting set for this.

regards

Norbert

It's not only Germany that has the biggest slice of Baan customers on BaanIV. It's worldwide!

By the way, Norbert, I love your signature! That's another reason for pushing for a Linux porting set! :D

Pat, could you tally the different releases that the users have mentioned in their profiles? Meaning a count of users per release? I just did a quick scan on the Dutch Baan World Users pages, and there the following numbers emerged:

Triton 2.x 22
Triton 3.x 30
Baan IV 203
Baan V 15
This does have some members double counted, because they use more than one release, but it gives a good idea...

That's pretty impossible given the fact the the version field in the user profiles are free text. Well not impossible but any selection would be pretty unreliable unless you hand count them.

Look at what I've found in portingset 7.1d.02 notes !!

Hi squaresh,

This is Baan 5c. Most of us out there are with Baan 4c

- Bader

I hear a lot of "Yes we need a Linux portingset". As far as I see most of you only want to have it as play ground. Are you really interesed to trust your business on a Linux server?

The current tests are done on the IBM supported Linux version (forgot the name). If this would become the only Linux supported version would you still vote Yes?

Currently it is already possible to put Oracle on a Linux machine so for people who want to see if Linux is really an optioon they can test there database on a Linux machine and face that Linux may sound as heaven, but you will face sufficient problems when putting your business on it.

Does this sound sufficient to get a flame war?

Kind regards,
Dick

Originally posted by Dikkie Dik
I hear a lot of "Yes we need a Linux portingset". As far as I see most of you only want to have it as play ground. Are you really interesed to trust your business on a Linux server?

If companies trust a backward product like M$ Windows to run their software on then surely yes, they would consider Linux!

How's that for a flame? :)

Hey Dick

Stating that Linux is not mature enough for serious business is surely not what you mean! There are lots of companies already trusting their complete business to Linux. And if SAP deemed it necessary to port to Linux, where does that leave Baan? Do they really want to lose yet another race? :(

If you want to, I can give you the name of a Dutch Baan customer who have already placed their operational Oracle database on Linux, and it works perfectly. The bshell is running on their old Sun SparcStation, and using Oracle SQL*Net to access the database. They for one will surely go for a Linux portingset immediately! And for production as well! (By the way, they run Linux on all their workstations as well, using BX for Baan, and StarOffice for normal office automation.)

And then all the other small companies. Like the poor lad in one of the other thread (http://www.baanboard.com/baanboard/showthread.php?s=&threadid=10599)s, having to decide whether to buy two M$ machines for their new Baan platform. Give them Linux, and they need only one machine!

I've been running BaanIV on that one and only Linux portingset for two years now. And another two installations, one on a SparcStation 20 and one on a M$2k machine. Guess what? The Linux machine is an old PentiumII-266 with 256MB RAM. The M$ machine is a PentiumIII-667 with 512MB RAM. And the Linux box runs circles around the M$ box. For instance rebuilding a complete package combination, without any table changes, thus no reconfig: Linux: 5 minutes, M$: 10 minutes at least, depending on how soon after a reboot I run it. After running for a day (as a single user) it already takes 15 minutes!

So how can you say we don't need Linux? Just think of the enormous savings on hardware! No more expensive ungrades of those "outdated" M$ machines, but wipe them, install Linux, and double your performance!

Talk about flaming!!!! :D

To Pat:
Indeed I also believe that Linux is more stable than M$ so you win here.

To Nico:
I am afraid that you compare apple with pears when comparing the 2 machines on your Linux box you probably run on tbase and this is not the case on your M$ box. And as we both know: you can never, never compare these 2.

I believe that there are customers who realy want to use Linux, but I still have the feeling that most of the reactions are based on the idea for a cheap play ground. There is nothing wrong with that, but don't expect that Baan will do a lot of effort for play grounds. It is not only building a porting set, it need to be tested, benchmarked, certified for different DB's and supported.

Did SAP backport their Linux version to a more than 5 year old Release? No, you had to ake the latest release (and pay for it). So, also the comparison with SAP is not realy honest.

BTW: I love Linux, but want to give some feedback.


Kind regards,
Dick

Hello Dikkie,

Originally posted by Dikkie Dik
I hear a lot of "Yes we need a Linux portingset". As far as I see most of you only want to have it as play ground. That´s a misunderstanding. At least we don´t want to play with it, we want to test it and then roll it out.

Are you really interesed to trust your business on a Linux server? Yes, we do already in other areas.


The current tests are done on the IBM supported Linux version (forgot the name). If this would become the only Linux supported version would you still vote Yes?
Yes. (as long if it´s for our current system - Baan IVc4)


Currently it is already possible to put Oracle on a Linux machine so for people who want to see if Linux is really an optioon they can test there database on a Linux machine and face that Linux may sound as heaven, but you will face sufficient problems when putting your business on it.

What kind of problems will occur ? Are they already known ?


regards

Dieter

Dick,

It really is a self-fulfilling prophecy isn't it? As long as some people/companies keep believing that Linux is an OS only for a bunch of nerds and geeks then nothing will ever happen. But the truth is, people/companies are running business critical applications on Linux OS. IBM and HP are shipping thousands of Linux servers every year and the number keeps increasing rapidly. Throughout the IT sector analysts agree that by 2007/2008 Linux business will have increased multiple fold. Can Baan really afford to miss that boat? No, I don't think so because if they do they are in for a rude and cold awakening in some years to come.

I don't see why customers shouldn't compare with SAP? At least they *are* doing it. Albeit only for the latest application release. Think of the huge confidence signal this would be to existing Baan customers. Something they are desperately looking for after many years of living wtih FUD and a tarnished software supplier. I believe this is a huge chance for Baan to give a clear and strong signal to their customers and to make them stick out amongst competitors. And yes, that does mean providing a portingset for BaanIV. Imagine walking into a customer's office who is contemplating between 3 different ERP vendors and you can say 'hey, I can do this for you on Linux!'

Why does Baan keep pushing customers on the - in my humble opinion - far inferior M$ platform. Why force customers to spend a good chunk of their investment budget on the ridiculous M$ licensing scheme. Why not push these customers onto Linux and rake in that piece of the cake themselves? Hey, you could even throw in Linux Support services. Companies may not pay for the software itself but they will pay for added value in services and they will still be cheaper off than going for a M$ solution. Maybe all those millions of Linux nerds around the globe may come in very handy then indeed!

Obviously choices will have to be made. The fragmented landscape in the Linux distro world is a problem but putting that forward as a showstopper is a red herring. Pick the 2,3 most viable Linux platforms and customers will be more than happy to stick with those. After all, it's not the distro that counts but the engine underneath it.

Linux is on snowballing speed. Whether Baan wants and/or can believe it or not. I hope they better realize it sooner than later.

Dikkie Dik :Did SAP backport their Linux version to a more than 5 year old Release? No, you had to ake the latest release (and pay for it). So, also the comparison with SAP is not realy honest.One of the signals Profuse is issuing in their circle of customers is that they support BaanIV for at least another 5 years. In these days this is a strong signal in the market: we believe the old application can cope even the next 5 years.

Also: what did newer versions of software bring in terms of better suitability for the buisiness software? In general, win95 has a lot that still can be used nowadays, except for the stability, security etc. But at that time it was deliberate to leave some of that out... In Baan terms: new features in BaanERP: some of them were good for some customers, but most of them could do with the old version.

Customers (certainly in the Baan software sector) do want stable software that can fulfil their needs. If it is a choice between M$ or Linux then the proven stability of the latter is certainly an advance.

Even in newer versions of software platforms there are components that can do whatever is needed without changed code. This is true for each platform (although edlin is not available anymore, edit still can edit your textfiles, just as good as notepad).

Finally some real arguments. Let's now execute: How can Baan be conviced to deliver a Linux porting set? Is BB an organisation with influence? Should all customers send letters? Any strategy?

Kind regards,
Dick

Originally posted by OmeLuuk

Even in newer versions of software platforms there are components that can do whatever is needed without changed code. This is true for each platform (although edlin is not available anymore, edit still can edit your textfiles, just as good as notepad).

I know it´s off-topic, but at least in W2K edlin is still there :-)

Dieter

Originally posted by Dikkie Dik
Finally some real arguments. Let's now execute: How can Baan be conviced to deliver a Linux porting set? Is BB an organisation with influence? Should all customers send letters? Any strategy?

Kind regards,
Dick

As to the BB influence bit: very little to none. Baan management has done a great job in completely ignoring Baanboard over the last 2 years. And it's not like we haven't tried. Therefore any power that would come from Baanboard would only be possible through its large group of individual members.

If Management was paying attention to BaanBoard, they would have known Linux is HOT:

I have made a list of top 15 in number of views (What are BaanBoard members interested in) and in number of posts (What is really stirring BaanBoard members and urging them to post)Top 15 views
4950 107 lock errors
4630 Editor for Baan (SCITE)
4189 Background info for bshell options (BaanIV)
3569 Baan on Linux
3364 What is bad about BaaN?
3199 DEV: Extend Baan code without sources
3085 Launching Excel from Baan
3034 File Browser
3000 Baan Consultants Required
2767 Baan Company Up for Grabs
2751 DEV: Create vanilla PDF files from Baan with Open Source components
2732 CODE: BaanXL - Yet another Baan to Excel utility.
2416 Poll: Baan Salaries
2310 Why Baan ERP on Linux?
2203 Color in BW

Top 15 posts
62 Baan on Linux patvdv Operating Systems & Databases
62 Is it a mans world?
59 The big question.. are we Planet of the Geeks?
51 File Browser
48 Question on integrating Excel with BaaN.
47 Editor for Baan (SCITE)
46 So who's going to win the cricket world cup?
41 World cup
39 Error 113 on adjusted tables
38 Baanfans - R.I.P?
35 DEV: Extend Baan code without sources
35 Launching Excel from Baan
35 Installation CD for Windows NT/2000
34 Background info for bshell options (BaanIV)
33 Why Baan ERP on Linux?So if you take both Linux related threads together: 5879 views and 95 posts for the top 15 Linux threads alone:
http://www.baanboard.com/baanboard/showthread.php?s=&threadid=23&highlight=why+baan+linux+portingset
http://www.baanboard.com/baanboard/showthread.php?s=&threadid=5758&highlight=why+baan+linux+portingset

Dieter: ...at least in W2K edlin is still there :-)XP *professional* still has it...

Pat:
As to the BB influence bit: very little to none. Baan management has done a great job in completely ignoring Baanboard over the last 2 years. And it's not like we haven't tried.
I agree with Pat. BaaN Board is not listed at the BaaN site for the User Forums.

http://www.baan.com/customers/baanusersgroups/

I wonder why an active forum as BB is ignored :confused:
How can Baan be conviced to deliver a Linux porting set? Is BB an organisation with influence? Should all customers send letters? Any strategy?
I think the only other forum which can knock BaaN's door is BWU-
http://www.bwu.org//Content_Page.aspx?name=howto

User Request Process/Responsibility Flow
ER > User Request Process/Responsibility Flow
Revised: 05/19/99
ER Submittal/Logging
Sub-committee Initial Review
Enhancement Request Procedure
Determination of Appropriate ER Submittal Flow
Company Submit
Committee Submit
Chapter Submit
Baan Status
Sub-committee Test

Originally posted by Dikkie Dik
To Nico:
I am afraid that you compare apple with pears when comparing the 2 machines on your Linux box you probably run on tbase and this is not the case on your M$ box. And as we both know: you can never, never compare these 2.
Okay, Dick, I accept the point of not allowing a comparison between the databases. TBase is by far the best regarding performance :D. So, I did a new test: without Baan, purely O/S, copying a dd directory. The reason why I selected this, is that basically that is what the O/S has to do during a Create Package Combination. Agree? Here are the results:

Windows2000

time /T
xcopy ddOPER_RHC ddTest /I /Q /O /T /E
time /T

19:39:30.71
19:39:48.58


Linux

date;time cp -r -p ddOPER_RHC ddTest;date

Mon Jul 14 19:39:43 CEST 2003
real 0m0.963s
user 0m0.040s
sys 0m0.460s
Mon Jul 14 19:39:44 CEST 2003


Thus, just on this part, a ratio of 17.87 vs 0.96 seconds! :) Does this qualify as comparing apples with apples? :p

Nico,

I did a dd copy on a very fast SUN system with high speed disks. It took me more than a minute to copy the whole dd. Ok it was B50c but even when having half the dd would take more than 30 seconds so it seems you have an amizing fast Linux machine.

Indeed the file handling of any M$ machine is not comparable with Linux, but did you know that Oracle benchmarks on M$ are comparable to Linux? So, indeed Linux is superior on price, but on performance they are sometimes close.

kind regards,
Dick

NvanBeest

I am keen on Linux and have previously voted. However, looking the latest result it is clear that this poll is not representative of the Baanboard population.

Only 61 people have voted. All of these want Linux. No one else has taken the time to vote, for either "don't care" vote or "don't want it" vote.

What can Baan conclude from this ? Could they conclude that there is a maximum 61 user sites out of the 1000's represented on Baanboard who would be interested in a Linux port and the rest don't care about it?

Terry

Dear all,

I agree with Terry. When only 61 people are interested in a Linux porting set, then it is not enough... But I think it is not the fault of the other members, we just took the discussion in a too technical direction. I am sure there are more users interested...

We are not focused. NPRo suggested to talk to BWU, and we are looking on ways to influence Baan to do the move. Why do we need the benchmarking discussion here? (Even though it is very important, but it takes away people that are less technical...)

Even though I highly respect the BB members with a technical background, what will make Baan move is IT managers and CIOs.

I go back to what NPro said, maybe we need to get our local BWU groups involved, BWU, DBug, BWU-nl, Baan Users-UK and Bug-anz and all the rest...

- Bader

OmeLuuk: ... Linux is HOT: ... top 15 in number of views (What are BaanBoard members interested in) and in number of posts (What is really stirring BaanBoard members and urging them to post)...Because I am aware the poll actually does not give usable information. Threadviews and ditto posts does not lie.

I agree with Lucas that a poll is not providing useful info. All BaanBoard users without a technical background, or without interest in this O/S & db won't even see this poll. Seems to me that the 61 users who did vote, are all "techies" :) Thus, to answer the question by Terry: No, I don't think the rest doesn't care about Linux specifically, but they probably don't care about O/S'es in general. Functional users will tend to do this, since for them, as long as the application runs, they're happy.

As for the benchmarking, I actually added it to prove that Linux has more pro's than low-cost only. Even if the database on different systems performs more-or-less the same, with the amount of O/S work done by Baan (reading menus, sessions, forms, etc...) it becomes important that the O/S is fast as well. Taking this back to the functional side, I believe that all users will be happy if they can work on a faster system. ;)

The suggestion of NP is actually the best. The only way we will ever get this really of the ground, is by having the user-groups bombard Baan with requests. But, if the user-groups are not convinced by the pro's and con's of an O/S, why would they ever start the bombardment? That's what the info in this thread can help with: convincing the user-groups...

Here's another prime example why Baan on Linux is a hardcore necessity:

Hello!
We have BaanIV, NT and Oracle 8.1.7.3 on our system, and sometimes baan is working very slow so we have to reboot server then is OK. When I stop baan services and oracle services all 4 procesors are stil on 80% becouse oracle.exe is stil on.
Can someone can tell me what could be wrong or what to look for opimise the system to avoid rebooting server every day

The quote was taken from this thread: http://www.baanboard.com/baanboard/showthread.php?s=&postid=45520#post45520

Buy Windows and get a daily reboot with it for free :rolleyes:

I never understand why companies accept a daily reboot of their business application server. Even customers who run Unix sometimes reboot more often than once a week: because their NT network servers need this, they assume it would be good for their Unix servers too.

Absolutely agree with Omeluuk. Unless you are having problems with processes leaking memory, abnormal kernel table fragmentation etc, there is no reason why you should bounce your box. If you have one of those problems, fix it and stop rebooting once they are fixed.

We used to have a customer as well who insisted in rebooting each saturday from cron...

I think that baan should be ported on linux at least for this reasons:

1. linux is an operating system that is easy to manage if you have unix skills. A lot of customers with production system on Unix will like Linux for test environment;

2. when you have a problem with the porting set, you may use some more debugging instruments compared to Windows (that is the only possible option on intel machine)

I do think that Linux is better Windows. The performance may be compared on disks, file system, memory usage ... but the main advantage of linux is the stability.

I would love to see baan for linux at least for i386 and powerpc and I don't care about the hardware/distribution supported by baan: I think it will work on my Linux box without a problem.

Why don't baanboard propose to baan to create a user forum (plus a few baan developer) for supporting linux on many distributions? This would let some more freedom to customers and will not tie them to a specific distributor.

There's one aspect of this discussion that keeps coming back (in the other Linux threads as well) about the flavour/distribution that should be supported. I checked a couple of different distributions, namely SuSe, Redhat and Mandrake, and can not see any differences that could pose a problem. Double checking on the Apache distribution list, I see that it is only split between processor type (i686, ppc, s390 and x86), but not specific distributions. So why does this question of distributions to be supported keep coming back? Isn't it true to say that once you have a product running stable on a certain kernel version, that it will run stable on all machines with that same kernel version?

Just asking...

Originally posted by eppesuiG
Why don't baanboard propose to baan to create a user forum (plus a few baan developer) for supporting linux on many distributions? This would let some more freedom to customers and will not tie them to a specific distributor.

It's very easy to propose and we would be very willing to do that. However, the ball really is in the camp of Baan. We are getting a bit fed up talking to a deaf wall for nearly 2 years now.

Originally posted by NvanBeest
... So why does this question of distributions to be supported keep coming back? Isn't it true to say that once you have a product running stable on a certain kernel version, that it will run stable on all machines with that same kernel version?

Just asking...

It is not as simple as you say: different distributions use different kernel versions and different libc versions. This is really a mess: did you tried to install oracle on something different that suse or redhat?

Moreover, different distributions use different package management tools: RPM, TAR, DEB,...

Bye,
Giuseppe

It is not as simple as you say: different distributions use different kernel versions and different libc versions.
So, if Baan ties their porting set to the lowest common kernel and libc versions, it should run on all (higher) versions as well. Isn't that correct? I have some old software, written for kernel 2.1.10, which runs perfectly on 2.2.18, thus backward compatibility doesn't look like to be an issue.

Moreover, different distributions use different package management tools: RPM, TAR, DEB,...
But all distributions have TAR, or at least, you can install it! And a porting set does not need an installer, thus distribution as a tar-file is sufficient.

When looking to distributions, Baan probably only has to care about distributions that support supported databases.

Dick

Originally posted by NvanBeest
So, if Baan ties their porting set to the lowest common kernel and libc versions, it should run on all (higher) versions as well. Isn't that correct?

It is, but new versions not only fix bugs, they also add features: so probably baan should look for a GOOD kernel and libc6. Usually good means latest :-)

Bye,
Giuseppe

Originally posted by Dikkie Dik
When looking to distributions, Baan probably only has to care about distributions that support supported databases.

Dick

In my opinion Baan do not need to run on the same machine as the database does, so probably the database supported are not so important.

But it is also true that baan need to compile its drivers (ora8_srv, ...) using libraries like the oracle OCI, so you are right.

Bye,
Giuseppe

What about this?

http://mindcraft.com/whitepapers/openbench1.html

-Suhas

That report and its results are more than 4 years old. Things have moved on since then!

Here is more recent stuff: http://www.kegel.com/nt-linux-benchmarks.html

By the way, any benchmark 'commissioned' by Microsoft, I would take with great suspicion.

Hi,

I've found these bug fixes for linux in the latest porting set release notes (6.1c.07.01 yet to be released) - and it looks like it is for Baan IVc!!!

=====================================================

MaintBaanIVc: # 17948 (BDUX11389): Fixed compilation error on Linux (and may be
others)

Date: Thu, 10 Apr 2003 15:04:32 +0200

Created on: MaintCorelli
Type: porting specific change

Problem Description (Technical terms)
Fixed build problem on Linux

Test Procedure
run Util/Boot on Linux. Check that no error occurs


=====================================================

DevBaanIVc: # 17810 (BDUX11275): Minor bug in cd_read, occurring on Linux

Date: Thu, 3 Apr 2003 09:30:38 +0200

Created on: MaintCorelli
Type: porting specific change

Problem Description (Technical terms)
cd_read on Linux used less program to show content of files. Less caused confusion, because 'q' need to be pressed, when shown file reached EOF

Workaround
type 'q' when less has shown the installation info (readme.mas)

Test Procedure
Run cd_read6.x and verify that no 'q' key press is needed anymore

=====================================================

Phil

I have seen them too... but do not forget that (supposed to be) good changes are ported over versions. Avoiding problems on one platform may also benefit the other platform. Does not mean that the portingset will be ported to Linux by default.

The source of the ported fix is "Created on: MaintCorelli".