This article is made in hopes of providing some clues about the strategy to migrate MS Windows to linux, especially in the case of Open Office, Samba

This article is made in hopes of providing some clues about the strategy to migrate MS Windows to linux, especially in the case of Open Office, Samba (file server).

All the company's management realized it was time in the middle of the economic conditions and competition is getting tighter business, all forms of saving as much as possible should be done. And also we know, during the last few months Microsoft Indonesia and his colleague BSA (Business Software Alliance) chase incentive company in Indonesia to pay a license for the use products.

This is not something we can blame and it is their rights as copyright holders are program. But as we know, the most expensive license from MS Office basic which includes only Word, Excel, and Power Point is about $ 150 dollars (or Rp. 1.5 million). If suppose a company has 100 PCs, then he has to pay around Rp. 150 million. Not to mention OS like Win XP license which is about $ 140 dollars. Means the total price is around Rp. 3 million per PC, multiplied by 100 means USD. 300 million. Not to mention the license server, CAL (client access license), etc.. For medium-sized companies that have about 100-300 PCs, the cost could reach billions of rupiah!

But of course, migration is not as easy as we imagine. Companies still need to weigh costs and benefits that exist. Therefore, the strategy is:
Analyzing the use of PC and its application in detail on all existing PCs. Here we have to do inventory on what program been used. Where the PC is only used for administrative function like typing, spreadsheets, etc.. Where the PC is used to run a certain program. Whatever it Program, platform and programming language is being used, etc..
Based data we get from the above analysis, then we do deployment coverage analysis. Within this analysis, we do research on all PCs that have: - a PC which can do 100% migration, both from the OS, or office applications.
- PC which can do 50% of migration, if only just office?
- PCs which are due to demand, migration can not be done at all, aka 0%.

Based on the above 3 scenarios, we also make our analysis of the costs for the management persentasikan. Of course it would be nice if all we can migrate your PC 100%, but from experience, it is difficult to achieve because usually there are some functions of the company that has / is already attached as a Microsoft product or any other proprietary. Pathetic indeed.


Case study:
For clarity, we discuss the following case study.

An advertising company has about 300 secondary PC that spread over several branches in and outside the city.

management is concerned because it had a letter from Microsoft Indonesia that it "appeals" to get the license status of inventori OS and software used in their PC2. After looking for info from rekan, obtained estimated price that it cost is about Rp. 1.5 million for the license Win XP, and Rp. 1.5 million to license basic MS Office (Word, Excel, Powerpoint), and a few dozen million to license a W2k3 server. Rough count cost to buy a license of about 300 PC is Rp. 900 million s / d 1.1 M.

By using the above analysis stage,  following results:
1. Approximately 30% of PCs are there, only used for office applications, ie word processor and spreadsheet.
2. Approximately 50% of PCs are running accounting programs, and sales are written using VB shared on the server.
3. Approximately 20% of PCs are running finance programs written using the shared Clipper Novell server.
4. Existing PC2 specifications bervariatif, starting from the Pentium Classic (mostly to run the program Clipper), Pentium 2, 3, and Pentium 4.

Well, here we have been able to begin to focus on what we can move. Because of the pressing time, so we concentrateg god to PC2 which can migrate.

PC2 using VB program before we leave for a while.

To PC2 which is only used for office applications, then we can immediately move to Linux, but because there are varied specifications, then we must specifik as follows:
1. For the Pentium Classic, we use the distribution Mandrake 9.0 or RH 8.
2. For the Pentium 2 and 3, we can use Mandrake 9.2, 10, or RH 9.
3. For the Pentium 4, can we use FC4 or Mandriva distributions 10.2.

For PC2 using shared applications using Clipper and Novell, we can "simulate" using Samba and your lecturer.

Here are trick:
1. We copy file2 applications into a Linux server, and then we share using samba.
2. We make it a samba share as usual, but do we add this configuration:
- Create a special user to access the share on the server and the workstation.
- We sharenya chmod 777 directory on the server.
- We add this line in smb.conf:
create mask = 777
default case = upper
preserve case = no
3. We chmod 4555 / usr / sbin / smbmnt on workstations
4. We created this line in / etc / fstab masing2 workstations:
/ / ipserver / namashare / home / user / mountpoint smbfs rw, user, noauto
5. Dosemu installed on workstations masing2

So we can connect to a samba server and run applications in a way that Clipper:
1. Open the console.
2. Connect the samba share and make point mount on workstations:
- Mount / home / user / mountpoint
- Enter the password if necessary
3. Turn your lecturer
4. Enter the mount point, and run a Clipper program as usual.

From the above explanation it occurred to hope that we can make savings up to several hundred million dollars from licensing fees. This is certainly not a small amount and the company can breathe a little.

However, based on experience, there are still some analysis to be done for the success of this migration process, namely:
1. Test needs to be more profound extent program Clipper compatibility is to walk in the environment Dosemu and samba. For a very different course in fungsi2 in a Clipper program, and maybe things are not going well.
2. Needs to be checked again to what extent the use of specific feature2 MS Office is used by the user in the document officen. Because even though OpenOffice has been able to accommodate most of the functions and features of MS Office, still not 100% compatible. If we find it is not running fungsi in OpenOffice, then we must think of a solution, if it can not be done at all in OpenOffice, OpenOffice or been able to do it but must of the pure native files in OpenOffice format.
3. And that certainly no less important is the first backup of all files that will be used. So when any time found problems, then the first version is still there.
4. The last thing I experienced is also important, to give understanding to the user, why the migration is done. Give an acceptable explanation of the user, and also Lead and accompanied by user in using the new program. Give comparison which can be seen directly by the user, for example, to print in MS Office we're open this menu and this ..., well in our OpenOffice open this menu and it ...


Thus tink about strategy or rather the experience that I can share in migrating from MS Windows to a solution OpenSource or GNU / Linux.

Summary:
Why waste hundreds / billions of dollars if available solution that is much more logical? Gitu lho ... :)

Grateful though is low profile and is not widely known, it appeared that some large companies are / have been migrating to Linux in a large scale. So, shut up but washed away, the Open Source Movement in Indonesia, keep flying.

Hopefully this paper can provide inspiration or idea for us who want to make licensing cost savings. It is time for management of migration to Linux see this as a logical, feasible, and have properly carried out by the care management of the business.