Open-source needs a game plan...

As I look at the direction that open-source software has taken in the last few years, I feel very enthusiastic about its future. However, to compete with companies like Microsoft, I think the open-source movement needs to be better at focusing the effort of the individuals who are willing to contribute their time to the development of new projects. I'm not suggesting that we have a group of benevolent dictators who are passing out programming assignments. What I'm suggesting is that it would be beneficial to us if we could create an environment that improves the chances of projects succeeding and makes individuals feel like their efforts won't be wasted.

When I go to Sourceforge I'm always struck by several things:

• How many projects have failed or never got started.
• The amount of duplication of effort.
• The number of projects that are server-side or web-based.
• The number of ideas that are missing.

The open-source community has had some great success stories like Apache and PHP. I think it would help if we could somehow create some turn-key packages that would help push *nix into more widespread usage. Here's just a few ideas of some of the things I would like to see developed:

1. A school administration system. Attendance, library administration, grades, email, collection of fees, posting of homework assignments to the web (so I can check if my son is lying to me so he can play Warcraft III instead), etc.

2. A church, club, social & charity organization system. Member info, dues/tithes/contributions, expenses (an accounting system).

3. Inventory control. This could be used for warehouses, libraries, rental stores, etc. Stuff gets brought in. Stuff gets taken out. Stuff might get brought back. There is a fee or cost with removing an item from inventory. There could be late penalties if it's an item that is being rented instead of bought. Items could be reserved (books, videos, infrastructure - like meeting rooms or projectors).

4. Vertical markets - Doctor, dentist, veterinarian offices. Banks. Department & grocery stores.

5. Government institutions - police, fire, courts, parks, public works, etc. Thousands of municipalities around the world are tracking the same information in proprietary applications/databases.

6. A replacement for Microsoft Access. A database that packages all the tables, queries, forms, reports, macros and code into a single file. People addicted to the Microsoft world complain about this missing feature a lot. As easy as MySQL is to use, they're still intimidated by it. They can get a working database with wizards using access.

Of course there are plenty more. This is just a sampling of some things that I think would push open-source into wider usage. We don't need ANOTHER Linux distribution or bulletin board or text editor.

One of the reasons that I think Linux hasn't succeeded on the desktop is because not as many programmers are proficient at doing Linux desktop development. Windows has tons of VB, Visual C++, Delphi, etc. programmers. These programmers have the tools and the expertise to develop rich, fat clients. Most of the Linux programmers I know either do web programming or server programming where no GUI interface is involved. This is one of the reasons I got a copy of Kylix 3 and have started learning it. The development environment isn't free but that doesn't mean I can't give away the source code. Before anyone blasts me for bringing a commercial product into the mix, I still have to write code for the Windows world and these programs have to have a certain level of functionality. I can use Kylix, C++ Builder and Delphi to write programs that will run on both platforms. If I didn't have to feed my kids, I could get by on a lot less.

Anyway, this was just a brain dump of some of the things I had been thinking about as a prompt to see what the rest of you guys thought.

Good stuff dcaillouet. I will add one thing pertaining to this ->

As we discussed last night, that just don't cut it and won't really help in the long run to better the open-source community. There is a wealth of opprotunity out there for the taking (the stuff you touched on in your thread), but not many want to step up to the plate.

Honestly, any firm that is profiting off of open source web technologies should feel obligated in some way to help push open source in the corporate sectors.

(This is exactly why I would like to have that discussion once a week on irc)

Again, good stuff dcaillouet.....

I know that's something you want to focus on in your discussions of small to mid-sized markets, I would just like to warn that I don't see that happening at already established companies. A large number of corporates probably don't want to spend the extra time/money/effort pushing free, open source software they use if they don't have to. It's not their business to advertise their resources after all (just think of how many companies are undoubtedly using Apache/PHP/Perl that could be helping to push open source by letting the community know, but don't).

That said, it's not something we should rely on, but rather, we should rely on a more focused marketing group to push to those corporates that are still latched to leaky, closed-source apps.

Of course, don't forget the importance of speaking up on behalf of open source the second you get a chance - every little bit still helps. I won a small victory at my business by pointing out, with solid, factual data, that IIS/PWS was no match for Apache's stability and security. We're still using IIS to serve intranet content, but I now have Apache on my machine, and I hope to use that as leverage in the near future to push IIS off the server and replace it with Apache.. one small step...

To get massive changes however (and this drum has been beat many times before, but I still think it's the best option) open source needs an organized, well-targeted marketing campaign. That's the one thing it seriously lacks. Everyone out there from Apache to PostgreSql to PHP is trying to push their own product, but not enough people are trying to push all of open source as a (nearly) all-encompassing alternative to Oracle/Microsoft/etc. The other problem with the people that are doing it is that it seems to focus strictly on Microsoft. While Microsoft is the biggest and the baddest, they are not the only grinch out there. I think rather than focusing on taking potshots a Microsoft and responding to their utterly ridiculous diatribes against OSI, we need to let their comments be heard and evaluated by the decision-makers. Past experience tells us that the people who are listening to Microsoft's drivel about OSI are hearing it for what it is - drivel.

We are dealing with (mostly) intelligent people who (usually) want the best products at the best prices. In many fields, OSI has the best products at the best prices. We need to focus on that and get out the word that OSI can deliver MORE at much better ROI than most systems. Someone needs to centralize the products that we have available and make them known. We need centralized resources - not just effort, but also $$$ - you simply cannot be a serious open market competitor without the all-knowing god of cash.

Unfortunately, while all this is well-meaning, I am at the same loss as most other people who feel this way - I don't even know where to begin organizing all of this, because there is so much out there that it's difficult sometimes to weed the useful, innovative products from the redesigned wheels... I swear... if I ever write a BB system... someone shoot me - but that's the state of OSI as dcaillouet said... duplicated effort and lack of focus. Microsoft knows this is our biggest weakness. We have better products, but the scope is so limited - hackers don't like to admit it, but you must adapt or die in the business world. OSI is NOT going to redefine the way business works on an administrative level, so we need to redefine the way OSI works instead (at least - in our face to the outside world )

BTW.. I was inspired by a program one of my profs wrote in Delphi for a dental management system. It does billing, scheduling, accounts, etc. I thought of doing something like that using perl/Tk then trying to market it to my dentists as a free alternative to the VB application they're using. I would get paid to implement a Linux server on the backend, then I would install the main system there. The interface to the secretary's / doctors / etc. would be written in perl/Tk for windows to make them comfortable with the design. I'm still kicking the idea around and researching what needs to go into it. Any tips would be appreciated, and, when I'm ready, I'll probably post in the Project Help Wanted forum.

All companies outgrow their systems at certain points in time. So, lets take this as an example ->

------------------------------

Mr Eds Automobile Insurance starts operation in July 2000. They start operation with limited cash flow and in return only purchase 10 workstations - bundled with win 98, office, etc.. Its now July 2002, their client base is now enormous, things can't go no better - its time to hire 5 more employees to expand their operations. Mr Ed realizes that with more employees, he needs to add more workstations. The new workstations are much more advanced than the old workstations, thus causing a overhaul of the other 10 workstations - BUT WAIT - Mr Ed also realizes that his company can also market automobile insurance from the web.
This senario ='s a golden opprotunity

-------------------------------

Now, 10 old workstations + 5 new workstations and an online marketing presence. This puts us in the prime spot - supply 5 new workstations running open source, systems integration with the other 10 that will now be ran by open source and not M$, a new content repository server that your sales person has talked Mr Ed into and an integrated content management package to help control and coordinate workflow between the office enviroment and web enviroment.

The above is the open source solution. Weigh that against an equivelent proprietary solution and look at the enormous difference in costs.

Our open source solution saved Mr Ed a large chunk of money. The best part of it all is that Mr Ed is now extremely pleased with what he has purchased. You just made a freind out of a key business owner within your community.

----Oh, forgot to mention----

Five years later Mr Ed goes national with offices in 20 of the 50 states. Who is he gonna call for his purchase of 200 new workstations?

That's the PERFECT sort of scenario we can jump on right now, but I'm concerned about large, established companies that take the "We use Microsoft, always have, always will" attitude even as they and their employees complain about the poorly implemented systems they are using. There are a lot of companies out there that, unlike Mr. Ed, still have money to burn and will continue to burn it. Which is easier for some clueless management sod who needs to make a tech decision:

1. Go with Microsoft like they always have. If something goes wrong, blame it on Microsoft, OR, blame it on the people who originally decided to use them years ago.
2. Go with an open source solution for the first time. If something goes wrong, take the blame for not going with old, "reliable" Microsoft systems (you know - the ones everyone in the company was griping about right up until they yanked them).

And let's face it, that's the attitude that's really keeping companies like MS afloat. Even while they complain that MS is unreliable and trouble-ridden, they won't switch to OS because they're comfortable with the MS problems they always have. We need a strong, integrated marketing drive to push OSes best products into places where they can benefit. Let's face this fact too awhile:

OSI does NOT have a product that can fully compete with Oracle or Sql Server. However, a lot of companies that are using those products don't need anywhere near the power that these two systems offer and would function just as well now and for the next 40 years with PostgreSql and MySql. It's that burning money syndrome again. If it's expensive it must be good. OSI, IMHO opinion, needs to focus on:

1. Killing the idiotic idea that price = quality.
2. Pushing through new innovation, as we've already determined that, largely, OS has somehow stagnated already.
3. Letting companies that we're competing with run their own ships. Who cares if Microsoft want to tie their own noose? We spend so much time lambasting them for the stupid things they do that we aren't as productive as we could be. Licensing 6.0? Don't waste time picking it to death (which I myself have been guilty of). Run out and show the big boss that you can save $25,000 a year by replacing Sql Server with PostgreSql and still function at 100%. The numbers speak for themselves, let's organize them and push them!

One other note:

Be patient. There's still a lot of stale decision makers sitting near the top these days. They learned Microsoft or Unix and they don't want to or can't learn a new Linux / BSD system. The younger generation, however, is taking a much greater interest in these types of systems because of the basic fact that Windows 9x - XP Home does nothing useful, while Linux/BSD with perl, apache, php, etc. can do all sorts of things. As this generation starts moving up through the ranks I think you'll start seeing a lot of massive moves towards Open Source and, generally, better products. Microsoft's time has come, they just can't see it yet. They no longer have the best products, they are not innovative (well.. never were really), and they simply can't compete anymore. Wait and see. It's happening.

This is a viable option until a competitor or smaller company starts hurting them with lower prices or better service because they are using open-source software and less expensive hardware. I've seen some open-source advocates hurt themselves by bad mouthing Microsoft with an undue religious fervor. There are many managers and decision makers who have gotten a lot of work done using Microsoft products. Over-the-top criticism of MS can sometimes backfire and discredit the speaker. I've had more success quietly introducing open-source products as prototypes or small projects without management's knowledge. Its easier to ask for forgiveness than permission. They're always pleasantly surprised when I demo something for them and tell them it won't cost them a dime in licensing.

I was reading an article about open-source databases recently and they mentioned Firebird and SAP in addition to the usual MySQL and Postgresql. I had seen Firebird mentioned a couple of times but never really paid much attention to it. Turns out that it is Interbase 6.0 and has some pretty impressive statistics:

• Maximum database 32TB. Largest current database 200 GB.
• Maximum number of tables 64k.
• Maximum number of rows ~2 billion.
• Maximum row size 64K (excluding BLOBs).
• Windows, Linux, HP-UX, Solaris and other Unix
• Triggers, Stored Procedures, User Defined Functions
• Referential Integrity, Check constraints, Domains
• Automatic two-phase commit, distributed two-phase commit.
• Replication
• Online backups
• Plus a lot more...

This was interesting mainly from the viewpoint of trying to convince skeptics to use open-source. If a manager says they're afraid they won't be able to get support, I could just tell them that they could always buy Interbase and let Borland support it for them if they wanted to. Same for SAP. This suggestion was not meant to disparage either Postgresql or MySQL. I'm simply talking about allaying the fears of people who equate commercial = better. I just loaded it on a Linux box and my laptop and it looks promising.

OSI should focus on the lowest hanging fruit, the stuff that's easy to grab. We're not going to get Delta Airlines and Chase Manhattan anytime soon. This is why I suggested schools, local governments and small offices. But anything we design should have salability, security and portability designed in from the start. Over-engineer even though it is initially for mid-size businesses with our target being larger corporations in the future. This will take years, but I think it is an achievable goal if we're patient.

Well, this is opening up a topic that could go anywhere... but I think there are some excellent ideas here. I like dcaillouet's list.

I will say that if you show you can do your stuff, most small-to-medium companies don't care if the software is non-MS. Of course, my field is web/intranet applications, but I think this can apply in other areas as well, perhaps to a lesser degree. In the past 3 years, working in a heavily Microsoft-infested area (South Florida), I have only done one project with non-OS tools. Even the one corporate site where the owners insisted on IIS/NT, I still managed to convince them that I could do the job more effectively with PHP than with ASP.

Anyway, this is a broad topic, as I said. Hard to get a good handle on it. I have been thinking about these problems for the past 3 years. Let me throw in my 10 cents:

1. I agree that one of the many problems with open source software on the desktop is the lack of standardization and easy IDE's. It's all about the GUI. Most people don't know or care what goes on underneath. And by standardization I don't mean whether KDE or Gnome should rule. I mean a standardized, easy way to build desktop applications.

This is why I am plugging for my favorite new technology: XUL. It provides a level of front-end standardization accross platforms that the computing world has never seen. AND, it is a perfect language for GUI IDE kits, because the front-end is handled in XML. You only use programming code for the back-end areas that are actually handling real programming logic (what a concept). (By "back-end", I mean wherever the logic occurs, whether on the client computer, or server-side programming). And the code can really be written in any language. It can run from the desktop, or from a server, or any crazy combination of both. And, it bridges the gap between Linux, Unix, Windows, and Mac operating systems. I think this could be the best thing that ever happened to open source. Let all the hardcore programmers keep writing in C, C++, Perl, Python, but now we have a way to bring it all together on the front end, without suffering through each hacker's idea of what a menu should look like. (I mean, there have been some seriously ugly GUIs developed by OS programmers).

First thing that we need to do is build an IDE for XUL, using XUL. That'll show em .

2. Now, going in a completely oblique direction, let me touch on what I consider a major reason why OSI stuff doesn't make inroads: there's no where anyone can go to check it out, play with it a bit, talk to an expert, etc... We need computer stores selling this stuff!. Maybe there are a few places in this country where one can actually play with Linux or FreeBSD on live computers, but most stores that sell Linux don't even have one live copy running on any of their demo systems.

I have been thinking for a couple years that if I had the resources, I would open up a little shop, in a nice central downtown business area, and just set up all kinds of cool stuff that people could come in and play with. You could blow people's minds if you did it right. In a store like this, one could sell computers more cheaply than any other local store, because of the software savings, and one could make plenty of contacts for consulting, application development, customization, etc... And, I think it would be genuinely fun to see people realize what can be done with open source software. It could be called something fun, like Compute the Difference.

I HEREBY DECLARE THAT THE ABOVE NAME (COMPUTE THE DIFFERENCE) WAS MY ORIGINAL CONCEPT. IF THIS NAME IS NOT TRADEMARKED OR IN ANY WAY RESTRICTED FOR USE AS A COMPUTER STORE, I GRANT THE RIGHT TO USE IT AS A STOREFRONT TITLE FOR ANY INDIVIDUAL OR ORGANIZATION SELLING COMPUTERS AND OPEN SOURCE SOFTWARE.

Fellow members, please be my witnesses .

Can we GPL a name? Whattaya guys think? Got any other good names?

Or we could call the store Free Beer, as in Free Beer. A nice recursive, in-joke.

By the way, Newsforge has a couple good links on this topic today:

http://newsforge.com/newsforge/02/1...53.shtml?tid=11 (notice that XUL gets mentioned in there?)

http://newsforge.com/article.pl?sid...1/1551256&tid=3 (One Sourceforge project is actually making money)

In response to rycamor's post above:We'd definitely get a lot of foot traffic from alcoholics and liquor license inspectors. But we could have a happy hour where ladies get software for free.



The first link rycamor listed had a piece of interesting feedback posted by a NewsForge reader about the CD full of free software:Maybe we always need to have a commercial version of each open-source project that is just a compiled version of the open-source code. Then we could please the people who can't sleep sound at night unless they've been gouged. Maybe we could call it Licensing 6.0.



I did a demo of some XUL stuff for my manager and another programmer where I work. They were very impressed. I just need more hours in the day to learn something new...

This really is testament to the fact that the wrong people are still making the technology buying decisions in corporate. It suggests that the legal ramifications of the software purchase are more important than the quality of the running code.

rycamor -

Nice idea with the store, and I'd definitely visit it . However, take it a step further. Get some local high school / college kid geeks to help you out and start offering free, hands-on learning with some sort of Linux system. Think about it: even as simple and pretty as RH 8.0 is, it's still a greek concept to the majority of regular computer users out there (permission? That's Read/Write/Hidden - yea?). Why not get them to the store with incredibly low prices on computers (man... build a (decent) RH 8.0 PC and sell it, with monitor and printer, for $450), then get them to actually USE them by offering them the chance to learn it in the presence of peers / teachers.

I have yet to start using XUL I've had so much school work (finals time of the year) that I haven't moved into ANYTHING new lately. I'm going to start working on that next though as I've heard nothing but good things about it from everyone here.