Build Your Business With Open Source

PCM2 writes "InfoWorld this week is running a ten-page guide to building your business entirely with OSS. The guide highlights OSS alternatives for many enterprise applications categories such as CRM, ERP, content management, and so on. It's not exhaustive, but where it skips the obvious categories like databases and Web servers it includes some others that you might not expect."

Comments

woo

[grub]

"Build Your Business With Open Source"
By Darl McBride & Chris Sontag

An interesting demographic

[rob_squared]

I've dealt with management at different companies I've worked for and the biggest issue they seem to have is that it will upset "the order of things." It seems that this is the perfect market for F/OSS. If you're already using it, its not as big of a headache to start. Now you just have to worry about the technical level of those that are starting their own business.

Who is listening?

[bogaboga]

While I appreciate Infoworld's piece, I wonder whether anyone relevant is listening. My boss for example will not even take a look. He says, M$ products have been doing fine for him for more than a decade and can still do more for another few years.

Question is: Are the people who matter reading these kinds of reports?

Re:Who is listening?

[TripMaster Monkey]

Actually, I've recommended a few open-source alternatives to my management, and so far we've saved a few grand. My boss will do just about anything to save on the bottom line, and when I tell him that I can fill a particular need with OSS and get out cheaper, he's beside himself wth joy.

Re:Who is listening?

[Chibi]

My boss for example will not even take a look. He says, M$ products have been doing fine for him for more than a decade and can still do more for another few years.

Well, you have to consider that there will be costs associated with switching over. There's manpower to actually install and configure the software, and then there's training and learning curve. All for what? To be doing the same stuff you were before. So, it might not be appealing to your boss from that perspective (this is assuming that you guys aren't constantly upgrading MS apps).

Your best opportunity with your boss might be when contracts/licenses are being renewed, or when you guys need a new application, and an open source solution might work out better.

Note: Of course, it's entirely possible your boss is just an ass, although the two are probably not mutually exclusive. :)

Re:Who is listening?

[hikerhat]

Your boss could be right. Transitioning from one platform to another can be incredibly expensive. It doesn't matter what kind of license the software has. Invariably, the new software doesn't do something the old software did, so you have to re-implement existing functionality. Data gets lost in the transition. The customer might see a few delays as you work the kinks out of the new system, costing very valuable customer confidence.

Replacing existing working software is a huge risk. If the transition doesn't go perfectly you've racked up more costs fixing the problems than two or three years of licensing the old product (compare the cost of a few IT people working on a problem full time over a few days to a one year MSDN subscription, for example).

Re:Huh?

[ucahg]

The solutions are as obvious as the categories.

Does anybody with the required knowledge of databases not know about Apache and Postgres/My/whatever SQL?

"build or buy"

[ir0b0t]

I've wondered for a *long* time why coders do not prefer a build-it model to servicing mass-produced proprietary code.

The profession of coding would be stronger as a profession if coders kept the source open and sold time to build individuals what they needed. There is little danger that non-coders will suddenly wean themselves from the need to hire coders just because the source is available.

Doctors generally don't keep their medical knowledge secret to make money. They share knowledge and concentrate on practicing.

It's more than just choosing applications

[Anonymous Coward]

Keep this in mind: a big-ass list of "open-source replacement alternatives" sort of implies that the closed-source path is the "normal way to do it" in the first place. If you're starting from there you've already lost. Every alternative choice will need to be justified to death and most will lose.

Better to bring the philosophy in this way: "We will use the best tool for the job. We strongly prefer open source for reliability and flexibility reasons; we will consider commercial products where appropriate." And then do the best job you can do with the tools you've chosen. A record of excellent results, even a very short one, is the best way to give open source a toehold.

Re:Bookkeeping software

[UnderScan]

The only quickbooks that I have used was a version for DOS back in 1995, so take my advice with a grain of salt. I haven't used this product myself, but Linux Canada makes Quasar a GPL'd accounting program. If you need it, you can buy tech support from them & if necessary you can buy the close source edition too.

Home Office

[doombob]

The company I work for always provides me with Non-OSS supplies like Fireworks, Dreamweaver, Frontpage, MS Office, and Windows XP. But my work at home involves various types of media projects including audio, video, and web. Right now I use Nvu for development, Audacity for my audio editing, and I'm trying out Jahshaka for video editing. And of course Open Office for everything else.

Human Resources Management

[Linker3000]

Almost as confirmation of an 'ask Slashdot' question of mine a while back, there still seems to be a big hole in the area of Employee/Human Resources Management.

If a company was smart...

[MarkEst1973]

... they wouldn't need this kind of thing. It's kind of like Paul Graham when he mused about his competitors and how it didn't really matter whether they knew he was using Lisp or not because, in the words of Robert Morris: "If they were that smart they'd already be programming in Lisp."

If a business was smart, they'd already be using open source as a competitive advantage. Google knows about servers and handling load. Your local PHB does not. Your PHB wants to buy Windows Server 2003. Google customized their own Linux distro.

I know enough to follow the really really smart people, like the ones at Google.

Re:If a company was smart...

[YrWrstNtmr]

If my company was in the 'insanely huge search engine' business, I'd likely do it pretty much like Google does it.

As we're not in that business, what works for Google (customized Linux distros running 10's of thousands of servers) may not work for me.

Re:Huh?

[TheRaven64]

Actually, yes. I was recently (peripherally) involved with a project that is going to use MS SQL as the backend and IIS as the front-end. The reason? They didn't know there were alternatives other than Oracle (which they couldn't afford). Even pointing out the lower TCO and lack of vendor lock-in, they still went with the MS solution because they'd heard of MS, and not of the other projects.

Not all businesses have competent IT people.

Missing items

[Lumpy]

and many times it's specific to what industry you are in but overall Sales and marketing tools are always missing from OSS. Where are tools for customer prospecting? how about tools for industry research off of aggregate databases available for purchase? Let alone a decent 4gl accounting package that exists as OSS.. dont get me wrong, I can buy a closed source real accounting,Inventory,and POS system for linux (no not that newbie crap like quickbooks or peachtree, a REAL accounting system) but there is no OSS stuff available that has a nice set of modules and Open scripting programming language set like 4gl so I can whip up a nice custom shipping module.

hen we get into the specalized apps, where can I get an OSS program to mine my Scaroborough or Nielsen databases I get sent monthly? How about a Traffic and Billing system for commercial sales in broadcast?

It's a neat idea, and with crossover office I can run those "special apps" but you can not realistically run your entire business on OSS. your accounting system at a minimum still needs to be a closed source app.. No commercial quality Accounting system exists in a useable state yet.

Some other factors

[plopez]

I read the article but if it is mentioned, I missed it; but there are 2 factors which should be considered in the 'build or buy' equation:

1) Lower risk of orphaned applications. If your vendor goes casters up or is bought out you may find your most useful application(s) unsupported.

2) I have a real problem with the 'one size fits all' ERP model. Suppose you have a business process which gives you a real advantage over your competitors. If you go with an ERP package which requires you change to the same business processes your competitors use, you just lost an important advantage. There is nothing to differentiate you from the competition (not to mention the fact that all real software should model the business process, not vice versa).

1) seems to be poorly understood by most PHB's, the thought never seems to come up.

2) I think this is due to PHB's being trained in an industrial paradigm. A paradigm which says it does not matter, all 'widgets' are the same and so the process should also be the same. Which may be true when building dishwahers and refigerators, but since most of the US economy is now a services economy this does not work in a services based industry. Services should be unique, otherwise you are *only* competing on price, which is insane.

Excellent Fit...

[wgray8231]

My department at a research hospital/university was recently (almost 2 years ago) formed. (Formerlly a division in another department.) The new chair moved everyone to GNU\Linux (Debian) because he hates everythin M$ stands for. It works out great with limited funding b/c the department spends less on software and many of the tools used in the field are available as OSS anyway.

What doess XPPro and Office cost for 20 or so computers, anyway?

Re:Alternative option.

[TheRaven64]

You are not paying extra for ease of use, you are paying more for vendor lock-in. The choice is:

1. Buy a single-vendor solution and hope that vendor keeps supporting it. It won't really fit your needs, but you can pay someone else to customise it and then be locked into using two vendors. Next year, support will run out on the solution you paid for, and you will be required to pay more for the upgrade.
2. Start with a Free solution and pay someone to customise it. Require that they release the customisations to you as Free Software (usually by assigning copyright to you). Next time you need to migrate systems, you have all the rights you need to employ a different contractor to do the work. You might stick with the old one, since they are more familiar with the code, but you are not forced to.

No system lasts for ever. Eventually you will need to migrate to something new. The cost of migrating away from a platform should always be factored into the initial purchase decision.

Re:All well and good...

[99BottlesOfBeerInMyF]

Until something doesn't work, then who do you call?

Umm, your vendor or whomever you contracted for support.

...in a corporate(business) environment, if something goes down, it has to be back up fast and without support, how does one accomplish that if it isn't withing that admin's realm of expertise?

If your admin can't manage a recovery plan and/or can't figure out how to run and install the software you need then you need a new admin. This has nothing to do with open vs. closed source or commercial vs. free software. Do you work for the government or something? That is the only place I've heard of where decisions are made that way. "We wanted to build a concrete building but the contractor we hired only knows how to build log cabins, so the building will be made out of logs." You choose your employees and your software based upon their strengths and weaknesses. If you can save 100K a year by using Apache instead of IIS across your whole enterprise, but your systems administrator can't figure out Apache, fire his ass pronto. He's got to be incompetent. It's as bad as those correspondence school programmers who want a job at a real development shop but can only program in visual basic and are completely unable to learn any other languages. It's just sad.

Actually...

[bobalu]

if they're smart they do what works for them given their employees, time requirements and other resources, regardless of the prevailing fashion.

think like a boss to convince your boss

[Brigadier]

My boss knows nothing about computers and doesn't care to. Once they allow him to meet his bottom line he will never change a thing. I've learned that whenever I speak to him instead of dicussing details and technical mumbo jumbo I break it down into profit and loss. I explain to him that by moving to a linux based OS server we can reduce our number of servers and downtime, and that the productivity incurred will = profit. Our P200 firewall/vpn/ftp/www servers have been running straight for over a year. While our windows boxen have brought the company to a hault on more than one occasion.

OpenConnector.Org

[Kunta Kinte]

We're working on it at OpenConnector.Org

But help is always needed. The code is still in alpha, though I'd like to release the first public Beta in November, in time for the projects 3 year aniversary.

Problem is that this is not a simple piece of software; there's a reason it hasn't been done. Very few people understand MAPI, and those who do, understablely want to get paid for doing it.

We need people experienced in MAPI, funds to offset coding time, etc.

Re:Huh?

[TheRaven64]

That's a myth. It depends on the vendor. Microsoft is simply not going to close it's doors in our lifetime. I'd bet my life on it.

Vendor lock-in is not about the company going bust. How long will the product you are using be maintained? When it is not, what are your migration options? How much will it cost?

With a Free package, you can pay someone indefinitely to back-port security fixes from the latest branch (which need only be a day or so a month, and can be cheap).

Plus, with databases, vendor-lock in is a moot point. Why? A. No competent business will change it's main database willy-nilly.

Exactly. Which is why you do not want to have to depend on a single source for support. If MS decides to EOL your DB five years after you deploy it, then this is a problem - it means you need to test the latest version, buy copies of it, and migrate. If it still works then all you want is to keep receiving security updates. Sounds like you want a Free DB...

If your company is doing anything with the database more complicated than a recipie list, any competent database developer is going to use stored procedures heavily, which are ALL database-specific.

So there is database lock-in, but not vendor lock-in. These are different things. I can easily use a different consultant to maintain my ultra-stable (only bug fixes allowed) branch of PostgreSQL - I can't do the same with MS SQL.

I started from scratch

[guacamolefoo]

Transitioning from one platform to another can be incredibly expensive. It doesn't matter what kind of license the software has.

Bingo. I had the option of starting my business from scratch. Nothing was in place, and I picked and chose from OSS and traditional software. I run a law office, and I ended up with the following:

1. Windows XP (needed it for my accounting package, plus training my minions on LInux was not an appealing idea)
2. Server OS: GNU/Linux
3. Website/Content management: PHP-Nuke
4. File server: Samba
5. Search tool for office network: Swish-e
6. Mail: Thunderbird
7. Office suite: Open Office
8. Browser: Firefox
9. Accounting: Quickbooks
10. AV: AVG

That's really all I need. I have a few pieces here and there: Paperport (which came OEM with my MFC machine) and Palm Desktop (came OEM with my Treo) for instance. I tried to mix and match based on my needs, budget, and consideration of implementation costs (that killed the idea of Linux on the desktop, though that's not out the window (so to speak) just yet -- I may ultimately make that move).

My standard rule in-house is to look OSS first, commercial second. I am clearly the exception in my community right now, by I am spreading the word. I'm not taking a ideological standpoint, simply a cost/beneift approach when spreading the word. I know OSS wins on initial cost (which is important to me now) and my staff has transitioned to Open Office pretty easily since there isn't a huge installed base of MS Office forms in place. In other areas, if an OSS app scratches an itch, I go that route if the software works inthe manner I need it to. If there is no OSS option, or there is a bad one, I do not hesitate to go commercial, and I don't feel badly about it.

FWIW, I know people who still run their offices on DOS Wordperfect versions, and these folks are giving serious consideration to OOO right now as a way to upgrade to a GUI office suite. They don't want to shell out hundreds per seat for MS Office.

YMMV, but ultimately, I think OSS will win/lose on the merits of the software rather than any ideological notion about how software should be created/licensed/distributed, etc. Upfront costs are a significant issue for me as well, but if the OSS software was not good, I wouldn't use it, even if it were free.