Switch from proprietary to open source software

Josey Wales

Hi,

I'm trying to prepare a presentation on whether or not Open Source Software (OSS) is a viable option for an IT section in a workplace. It is only a theoretical presentation, it won't affect the decision whether we do or do not switch to OSS. It is however quite important for myself (part of a promotion process). Unfortunately, I don't know a huge amount about OSS.

I am wondering what people would think would be the logical first steps you would take when looking into something like this. I am thinking that replacing all our proprietary software is not the way to go. We develop apps on Windows XP machines for use on Windows machines so I don't think I'd suggest replacing our operating system straight away. I was thinking of suggesting replacing MS Office. It is not critical to the work we do and something like OpenOffice.org has ready made replacements.

Is that a sensible suggestion? Are there some important questions I should be considering in the switch to an open source office suite? Would it be an easy switch over?

Any suggestions would be appreciated.

Thanks.

Capt'n Midnight

Is it to save money or for other reasons ?

In addition to OpenOffice have a look at OpenProj as a replacement for microsoft project.

Other big costs are email, but the open source solutions are complex too. If all you need is email then I'd recomend mdaemon from AltN - not free - not opensource, but excellent value for money, robust AND the configuration is based on text files so managent is very easy. If you need an scheduling/workflow package then you are looking at different animals.

If you have many servers it's cheaper to use client CAL's in which case there aren't many savings on having a linux fileserver.

What do you develop in ?
How much of the microsoft API do you use as that is the key lock in. Also do your apps link into microsoft office, in which case you would need to keep it too.

One BIG thing to look at is what do clients need to run your software ? It's easy to design stuff that uses microsoft Access to generate reports, but this means the client has to spend quite a bit too. eg: Windows / Office Pro / Exchange / SQL / windows CAL's / Exchange CAL's / sql CAL's and it all adds up. Terminal servers don't save you any money on microsoft licenses since the client has to have a license for the software used on the terminal server. ( yes I've had to buy Access licenses for Mac's :rolleyes: )

Having something like a java based solution could possibly save the client a grand per user.

First work out what you have, license count based on what purchased. next work out what SW you absolutely need, for example one copy of each OS supported - how much of this could be done with MSDN developers kits instead ? Then look at functions that aren't version/brand dependent this is where OSS comes in

no doubt there are better methodologies

Josey Wales

Capt'n Midnight wrote: »

Is it to save money or for other reasons ?

The brief was just to answer the following, "Does open source have a place in Revenue IT?"

Capt'n Midnight wrote: »

In addition to OpenOffice have a look at OpenProj as a replacement for microsoft project.

Thanks for the heads up on OpenProj.

Also we develop standalone applications that don't access any part of MS Office.

croo

I am heavily involved with an OSS ERP called Adempiere, and I can tell you one of the reasons I chose to work with OSS is because of reading Eric E. Raymond's “The Cathedral & the Bazaar” article [http://en.wikipedia.org/wiki/The_Cathedral_and_the_Bazaar ].
As a developer I believed Raymond that OSS was best way to achieving quality software... and as software and the environment it must run in becomes more and more complex, quality is becoming harder and harder to achieve. OSS that allows the world to use and potentially debug your code [if only you can encourage the world to get involved and use the application or join your community] should result in higher quality code. For me, in a business environment, quality is king.. hence I believe OSS will deliver IF it can achieve the needed community size.

Capt'n midnight gives some good examples of what applications you could use and I agree that for a company which already has spent the money on desktop solutions there is probably more that can be done in the back room servers such that the front office windows users would not even be aware of a change. Firewalls, File & Printer servers with samba, groupware applications many of which will integrate with outlook or thunderbird. And this is a typical advantage of OSS, Flexibility. Flexibility to use your OS of choice, or email client or choice. I use both windows & linux machines and MS Windows orientated applications tend to be windows only whereas many OSS applications are technology neutral supporting windows, linux or mac. This flexibility and of course the many options for any given task, to me represents less vendor tie in... and less vendor tie in represents less risk to me!

But ultimately the decision to use an OSS application or not should, I think, be based on the same process you would use to decide any other selection of software. Clearly set out your objectives, research your alternatives, understand the consequences, trade-offs and risks of each alternative... then your are in a position to decide.

Gandalf23

I dont think this decision has anything to do with the technology tbh. FOSS can do everything proprietary software can do, and the arguments about less bugs, more eyes, massive developer network, etc, etc, etc are well won at this stage. OpenOffice and StarOffice are almost totally interoperable with MS Office these days too. Not to mention "wine".

This decision is not a technical one ... its a social and political one. I did a lot of research in a big organisation in Dublin who migrated from MS Office to StarOffice. The migration ultimately failed because of a variety of complex reasons ... chief among which was the staff view that StarOffice was an inferior product to MS Office (the whole "free" = "worthless" debate). I'm now working on a very large multi-national FOSS project where sililar themes are emerging.

OP I have a lot of material and presentations on this ... PM me and I'll help all I can. You have today and tomorrow ... I'm snowboarding for a week on saturday and traveling fairly extensively till end of May after that.

Gandalf23

PM received and reply sent!

SouperComputer

Interesting article......

http://perens.com/Articles/Economic.html

anfirrua

My thoughts on this are why limit yourself to one solution

A switch to OSS in any pre-deployed organisation will be a staged process.

The desktop could be a start, with gnome and kde being excellent and user friendly alternatives
There are many Windows application delivery server products that will allow you to deliver all windows applications (MS office + other commercial apps) to the KDE/Gnome desktop in a seamless manor. My recommendation here would be a few high end Citrix servers.
Start off on the desktop save on the XP and avoid the Vista licenses, centralise the management of your windows app's and delivery of them to a few high end clustered windows servers

R3al

Josey Wales wrote: »

The brief was just to answer the following, "Does open source have a place in Revenue IT?"
.

If this is for the Revenue AO IT role a friend of mine had the same interview a couple of weeks ago.

He chose one of the other criteria for his presentation but they did question him about the other presentation topics so it might be an idea to do a little research on the other areas also

One of the questions they asked in reference to open source was what issues might be associated with using it. Most civil service interviews are fairly structured so you can expect similar questions for each candidate.

Also expect to be challanged on any of the points you make in your presentation.

With regard to open source examples you might want to use something deeper than just open office. Bear in mind that revenue are very much focussed on their revenue on-line system and data mining techniques to identify and target higher risk companies/individuals.

Good Luck and don't fret too much about the presentation

R3al (Disclosure) ICCMHosting.com rep

Josey Wales

Thanks for the response R3al.

You are correct about the interview. The introduction of an open source replacement for Office was only my initial idea. I have since being doing a lot of research and have much more (hopefully) insightful proposals. Funnily enough some of them are about data mining techniques as you suggested.

pushpop

Hey all

I saw this post and had to reply! I've recently performed a fairly substantial switch from proprietary to GNU software on Windows XP.
The list of switches I made are as follows:

Word 2003 -> AbiWord
Excel -> Gnumeric
WinRAR -> 7Zip
MSN -> Pidgin
Thunderbird -> Sylpheed
Photoshop -> GIMP
Visio -> Dia
Notepad/Editplus-> Notepad++

I also use MinGW for C/C++ compilation, MikTeX for LaTeX document preparation along with TexnicCenter, Wireshark for network analysis and Cygwin for Unix-like emulation

I still use some proprietary software like XP as Trinity's wireless network requires you to jump through hoops and do some horrible messing with scripts to get on to it with Linux. I use ESET Smart Security as my antivirus/firewall as it's the best available and there are no self-managing OSS antivirus tools.

The reasons I did this were, again, largely political but partly functional as well. With every new release of Microsoft/Adobe software it just gets slower, more complex to use and uglier. They have a penchant for randomly dialling home for updates and god knows what else and I don't trust them one bit.
With all of the tools above, they are lightweight, easy to use, fully functional replacements for their proprietary equivalents without the bloat or backdoor mentality.

mick.fr

Well there are hundreds of equivalent apps on Linux, many of them with a *NIX background, not really a problem to technically switch to Linux.

The biggest issue would probably be the equivalent of infrastructure products, asset management, software inventory, enterprise portal, rights/identity management, documents workflow, general support, Kernel issues.

If it is decided to switch to Linux for the enterprise, considering the number of "enterprise type" products available with Novell, I would probably think to them first.

Just my 2 cents

MacAonghusa

pushpop wrote: »

I've recently performed a fairly substantial switch from proprietary to GNU software on Windows XP.
The list of switches I made are as follows:

Thunderbird -> Sylpheed

Thunderbird proprietary?

Vixen Sexlectra

Hello All,

I am doing my Client Research Project (Thesis) on "The differences between Open Source Solutions and Commercial IT Solutions"

As part of the project, I am conducting an online survey/questionnaire to assess:

1. Peoples knowledge of the IT Solutions available.
2. Usage pattern of both types of users.
3. Opinions & attitudes towards both systems.
4. Idea's & comments on how to improve both systems
etc

I would be very grateful, if you would take part in the survey/questionnaire, as it would be a great help to me in the research process.

I am looking for Irish people & Businesses to give me their insight and knowledge into this subject.

Can you please send me a message if you are interested..
Looking for sample size of approx 40 people.

Thank You.
Mia