Abandonware, or 'Allaire Forums Open Sourced'

xtra sends news that Allaire has "open sourced" its Forums software (web-based threaded discussion groups). You can either rush off to check it out or read a bit of commentary below about abandoned software.

For those who don't know, Forums is a package of Cold Fusion templates which runs on a web server with Cold Fusion installed. You can see an example of it at forums.allaire.com.

I've been using Cold Fusion for a few years now, and my initial reaction was, "Why bother?" Allaire released the initial version of Forums in 1996, supported it for about 20 minutes or so[1], and began the process of abandoning it in favor of developing a lucrative "enterprise computing" package. Around 1998 or so, perhaps even 1997, people started asking for it to be open-sourced. In 1998 Allaire made a few bug fixes and released version 2 of the software, and in November 1999, they announced the software would be open-sourced. And yesterday, they actually did it. Of course, all of their Forums customers decided they were abandoned a few years ago, and found another product.

Now in one sense, Forums has always been "open source". Cold Fusion templates are interpreted, not compiled (and Forums was released before Allaire added even the weak encryption for templates that they now support), so anyone with a few weeks to kill could scrutinize the code and figure out how everything worked. Of course, redistributing modified versions of the code was a no-no, and if you made any modifications to it, even bug-fixes, then you lost all support from Allaire. Just like any compiled software product. In fact, I believe changing the templates was against the old license agreement, though I don't have a copy of the old license readily available.

So basically what they've done is make it freely downloadable. The license agreement is one of those bastardized we-took-our-standard-license-and-changed-a-few-words things that very obviously originated in the mind of a lawyer used to writing proprietary software licenses. I suppose it's open, but it sure isn't friendly about being open, know what I mean? It giveth with one hand, and taketh away with the other.

In any case, this is a classical example of the "abandoning a product while trying to keep our customers from feeling abandoned" open-source motivation. I'm dubious about its success, in this case or in the other cases where this is the motivation behind opening up a set of code. If Allaire had actually done this maybe three years ago, Forums would probably be a robust and stable product by now, and it would probably be driving a fair number of sales of the Cold Fusion application server. Instead, Allaire collected ($400 * #number_of_sales#) and pissed off (0.95 * #number_of_forums_customers#) by selling them a product with zero support.

Are people really going to flock to it now, spend a few hours parsing the license and trying to figure out if they can do anything useful with the code, and spontaneously develop a thriving user-group to support this thing? If Allaire can't even support it, why do they expect others to?

Maybe I'm wrong. Maybe it'll be a huge success, it'll turn into a beautiful open-source product and every Cold Fusion site worldwide will want their own set of discussion groups. Or maybe it'll just turn into another ghost site, lights on, but nobody home.

The open source/free software community is likely to see a lot more of these sorts of projects in the future. When your car gets old and feeble, and you don't feel like fixing it up anymore, you can donate it to Goodwill or to your local school for their auto shop course. Or you just take the plates off and drive it down to the waterfront, roll down the windows, leave the keys in it, and walk away.

When you don't feel like supporting your old software anymore, you dub it "open source", send out a press release touting your bold move, and dump it in the software burial grounds. It's a little better than previous burial methods (which involving interring the software in Yucca Mountain, permanent disposal), but maybe not very much.

Now Emmett is a little more optimistic. He notes that if even one person does something useful with the code, it's a net gain. And I suppose he's right. But the community is going to have to learn how to deal with "open source" code that is actually just a cynical move to dump some unsupported product and talk about how you're supporting the open source world.

Emmett: I agree with you, but there's always the point that someone will probably find a good use for it, even if it's to test it once and throw it away. OOP means that talented developers are talented in the reuse of code. I mean, if only one person picks it apart, takes 200 lines of code, and uses it to build some better, more efficient system of some sort, and GPL all of it, isn't that worth it? I think so, and I think it's the 'one person, somewhere' belief that keeps everything going. If they were serious about doing something, they would have used the GPL and been done with it.

The key here is that Allaire isn't thinking 'one person, somewhere,' they're thinking, 'good PR,' while they'll turn this into a PR extravaganza and say they're thinking 'one person, somewhere.' I think we're on the same page when I say we're both thinking 'too little, too late.' Don't confuse clever marketing and free advertising for innovation.

[1] A slight exaggeration. I believe Allaire actually supported the product for at least a week after they rolled it out.

This is not a Bad Thing

Frankly, I'd say it was a overall Great Thing, at least if it catches on. After all, what are the primary methods of learning how to program, gaining the tricks and techniques of the trade?

• Take a class in it, which will teach you syntax and not much else
• Figure it out yourself from the techniques you already know and the documentation you have
• (this is the big one) Learning how someone else did it

Way back when I was a wee tyke, when I was first getting the hang of Pascal, I would have given my left kidney to access the wealth of open source code today. Not because it's a free product, but because I could have learned things like fast hash functions and various optimizations without having to munge through the concepts myself. Something may be gained from the experience, I suppose, but the pain is considerably more than the gain.

If software companies want to abandon software by making it open source, then scores of programmers in their crysalis stage are going to benefit. The customers were screwed already from the companies poor support policies, so they don't even enter into the equation.

With an eye to the future, I'd say that anything that will help people become programmers is a worthwhile endeavor. Isn't that what the geek community is about?

Now if they would just do that to Cold Fusion

I currently use Cold Fusion at work. I may not have to for long. For months I tried to convince my boss that Linux and PHP was a much better solution than Solaris and Cold Fusion. (And *much* more cost effective.)

Then I installed RedHat 6.2 on his Sparc 2.

He has changed his mind.

Now that he has seen what it can do on his home machine, he is more than impressed. PHP runs rings are ound Cold Fusion when it comes to features and timeliness of updates. I am *still* waiting for our 4.5.1 bug fix release for Solaris. I will probably have to call them and scream until I am blue in the face to get it, even though we have a paid subscription for the software.

Linux and PHP are just a better choice all around.

Allaire in 96, iFactory in 98, Crack.Com in 99, ..

Back during 1996 I took a job as a web developer. The product the website was written in was ColdFusion 1.0 with some specific areas in 1.5. Coming from a programming background in C/C++ and pascal I found the shit to ColdFusion rather a pain.

Complete lack of programming structure, and the so called DBM/DBML files with their bad attempt at integration into HTML tag structures. With this language there was much that could be done and I decided to finish the site and clean things up. Part of this clean up with the installation and bug fixes for the Forums product.

The attempt was being made to use Forums to provide technical support, when they worked, for our customers. I remember recieving copies of updates and bugfixes on a semi regular basis, most of which broke something new and different each time.

After dealing with this for a short while we (ok I) decided to yank them down and re-write them because I was getting tired of users (customers) calling in and complaining about not understanding how to use forums.

I still wonder just where they got their initial design for the forums software. It would have been nice if they could have just ripped the perfectly usable and understandable system from Renegade or Maximus BBS systems. Atleast those I could very easily convice the users on how to use since it resembled the 3270 and 5250 green screens they were used to.

Also performance was horrid. On the Pentium 133, 128MB RAM (and if I remeber it cost us a few grand) with 2 4GB SCSI drives running Windows NT 3.51 (yes, real performance software I know) we could only get 1 page view every 3 - 5 seconds if someone was using the forums. After re-writing and converting/upgrading to ColdFusion 2.0 (which ran as a service instead of multi MB CGI) performance started increasing. Performance wasn't acceptable until leaving ColdFusion and another site re-write in InternetFacotry's SMX (running CommerceBuilder for the www server).

Now there is a good case for an abandond product as well. CommerceBuilder provided speed and integrated programming language, and much of what Apache and PHP on linux now provide. Too bad it too went the way of the scrap heap back in 1998.

Now merchant builder was a fun product to run, complete store with source and it was fast. It ran in code inside the server and was pretty. After the webserver war of netscape and microsoft started they became one of the casualties. Sure they ported their SMX to a IIS ISAPI module and moved merchant builder to it, which is all good, but they stoped supporting Commerce Builder.

Then Crack.Com in 99. They run out of money and "do-the-right-thing"(tm). They opensource their game with a $100,000 musical score and work on it, on their own dime, after the company ran out of money to finish it. Most famous for Abuse and getting many of us started using Linux (Mostly because slackware 2.x included Abuse and Doom), it is kind of sad to see this company go.

This pretty much shows the three basic ideas for a product when support is no longer available. Allaire drops it on the opensource people well after it could have been usefull and after alienating those long standing customers. Internet Factory in just completely dropping off the face of the planet and telling the customers to microsoft for their fix. And finally Crack.Com which gave their game away to the public domain.

I just hope Allaire in releasing this code realizes what they are doing and the PR appearnce they get from this. Not the kind bunch of every day geeks (a la Crack.Com), but the arrogance of holier-than-thou-I-got-rich-quick-and-you-didn't-m y-product-sucks-your-on-your-own-you-fix -the-problems attitude I have come to know and loath from Allaire once again.

PS. Yes I admit it, I still write applications in ColdFusion, if and only if I get hourly pay and 50% on top of that for headache medicine in order to deal with the language's lack of programming structures. (Yes I know it is starting to get some structure with the 4.x branch, but it still is no where close to speed, performance, and stability as Apache/PHP and dare I say it IIS/ASP.)

Re:Cold Fusion

In the PHP FAQ [php.net], there's a section [php.net] on PHP vs. Coldfusion. Also, there's a link to an article by a CF and PHP developer, making the comparison [theaimsgroup.com].

I found it REALLY useful.

Re:Cold Fusion is better than PHP, at least right

I'm really curious about what you don't like about PHP's documentation; I find it to be outstanding. I had an appointment book/calendaring application I had written in Python, and decided (as a learning experience) to redo it in PHP. Using the online documentation, I had most of the functionality duplicated in a weekend. You can't beat that with a stick.

I should also point out that while I know how to program (to an extent, anyway) I am no programmer; I don't do it for a living and never have. I have a passing familiarity with C, I know no Perl, and PHP was a dead cinch to pick up. Have you actually given it a try, or are you relying on something someone else told you? I should think anyone with even a little programming background (and since it appears you do this for a living I would hope that includes you) would find it easy.

Just Selling Their Interpreter?

As someone pointed out, Cold Fusion has always been an interpreted language. Therefore, giving away the source for Forums will actually generate revenue in sales of their interpreter, since you need to buy that for this for your web server, right?

So, they get to be "good guys" for giving away source to something you can't run without buying their product anyways, right?

Roy

Why the negativity?

I'm having trouble understanding this. I know that the open source community has, in the past, _begged_ and _pleaded_ for old, uninteresting software to be released to us. And now that it's actually starting to happen, it's just a PR stunt? That's the friggin' point! We've finally managed to give them enough of a good reason to actually do it. YAY.

PHP vs. Cold Fusion

Can anyone point me to a nice, level headed comparison between PHP and Cold Fusion? Does one have a higher overhead? Does one cache database results? We're going to be choosing a web application development language at work soon, and if I can say "Well Cold Fusion is nice, but..." I'd be happier with the open source solution.
--

Abandonware Wonderful!

If you've ever worked at a business that purchased a commercial product to do a basic task, and found yourself in a situation where the company no longer supported it but also was unwilling to release the source code so that you could make minor modifications to "keep it alive" until you could find a suitable replacement, the idea of a company open sourcing a product instead of abandoning it makes you feel all giddy inside.

I've found myself hex editing code or writing complicated wrappers in order to support extremely minor environment changes (upgrading patchlevels of another product) that could have been changed in seconds had I the source code to rework. I can't count how many abandoned but perfectly good packages were thrown out due for Y2K because the company was unwilling to test it and we had no facilities too.

I've advised that companies who buy expensive software contracts or packaages build in an option to purchase that software source should the developer choose to abandon it.

Re:Cold Fusion is better than PHP, at least right

Personally, I love the PHP documentation. It's concise and to the point--I rarely spend more than 15 seconds finding exactly the information I'm looking for, because it's so logically laid out, and if it takes any longer than that, I just grep for it.

It's not a tutorial, it's not an introduction, it's not a style guide or a talking paper clip. It tells you exactly as much as you need to know if you understand the basic idea of what server-side scripting is but never heard of PHP.

And if you've ever programmed in C, you'll find that you rarely have to look at the syntax portion of the manual at all. One read through should do the trick, once you've finished ooh-ing and aah-ing at all the high-level things you're allowed to do... :-)

Incorrect URL in story

The Correct URL is http://www.forumspot.org [forumspot.org]

Complain, Complain...

Granted the "Open" license that Allaire is using is not exactly wonderful. And, granted, Cold Fusion is an application that a lot of coders find excrutiatingly useless.

However, we should be happy that it's a step in the right direction. I have seen literally hundreds of posts on Slashdot expressing the desire for companies to open up products that are aging, instead of just shelving them. Here, Allaire has done this (sort of), and they're being criticized for the manner in which it was handled.

I agree that Allaire deserves some criticism for their half-hearted approach, but we should remember that they didn't have to do anything: they could have just abandoned this product. Someone there is facing in the right direction, so let's not be so hasty to beat them down for thier blunders.

--

Cold Fusion

So I like many others in my position use Cold Fusion in my office, and would like to get some momentum behind a movement to get off of Cold Fusion and onto PHP. The one problem tat I am finding is that despite their similar natures, there is almost never a comparison between the two. PHP is mostly compared to ASP, JSP & Zope, and Cold Fusion is mostly compared to ASP & other proprietary engines - but I never seem to find any information on how the two compare.

What I would really need to convince the PHBs is:
a. a comparison on how they hold up under heavy loads
b. a comparison of features
c. a comparison of how interoperable they are with other languages/toolkits

PHP looks better, but without any proof that it can hold up under the loads that cold fusion can (or beat CF) I can't get any movement.....

thanks.

Cold Fusion is better than PHP, at least right now

I've used Cold Fusion for 3 years now, and have seen the product since 2.0. In my opinion, Cold Fusion is a lot better than PHP because of the way it works. The tag based metaphor focuses on the Output of Database information into HTML. You can just type in an SQL query and output it without having to deal with all that Command/Recordset and object crap. It's a great mix for people with moderate programing background who do not like the low level C++/Java stuff. You may have some criticism about Cold Fusion having bugs, but they make up for it in a lot more things. Cold Fusion has much easier concepts to learn and use, and while PHP may be powerful--the documentation just sucks. That is a problem. Every Open Source project I see is not geared towards training the language to people. I have been initially excited about PHP, Zope, and Perl, but I have found the documentation to be written for "the geek", while the books on Cold Fusion, both from Allaire and from Ben Forta, focus on training new people, as well as teaching advanced concepts. Only now with a few books on PHP on the market can I really start working with it. Many of us want to deal with stuff that's more or less as easy to use as a 4GL, or deal with complex configurations. I'd rather use CF's web based administrator and CFML on IIS and Windows NT than deal with PHP's INI files on a Linux box. Allaire may not follow the GPL or the Open Source Philosophy, but they are a lot better than the Microsoft path, and they are less insular.

Abandonware?

First, cute name - I like it. Second, open source to businesses is still largely a PR stunt. Witness Sun's foray into it, or Apple's - both had absolutely atrocious licensing under the "open source" banner... both were (largely) resolved after community outcry and some work by our esteemed Bruce Perens, but many other companies have not been subjected to that level of scrutiny.

Most businesses still cannot (and will continue to for some time) grasp the basic idea of giving something away. In their world, if you give something away, you earn no money in return. This is true. They also presume support will continue to derive the same income, as the number of users hasn't increased much over a period of N. Hence, the logical conclusion is that without that source of revenue, they need to cut back expenditures to survive. Not very optimal - they lose profits. The flipside that they don't realize is that by open sourcing something, you can decrease your maintenance costs and increase your user base (if properly executed).

Companies don't see the long-term benefit. And, to be honest, open source is risky - there is no guarantee of high market penetration, as always.. and companies do NOT like risk, especially an unknown one - and there is unsufficient data right now to make a long-term commitment for many companies.

I know everyone is now shouting "FUD! FUD!" but this is how companies think, so you'd better warm up to the idea. Open source developers for the most part aren't getting rich - Microsoft developers are. That's a pretty powerful statement. So, if you want companies to open source a live product, instead of one that is no longer generating revenue (at which point it doesn't matter much what happens to the software, so the decision to go open source is easier) maybe we should concentrate on producing some hard numbers on companies that have taken the plunge.. and what happened 5 years later.

Better URLs

Better URLs for information about this seems to be http://www.allaire.com/products/forums/ [allaire.com] and http://www.forumspot.org/ [forumspot.org]

Re:Didn't the same thing happen with Interbase?

Well, with Interbase, it COULD have ended up as "abandonware," as it had been pushed pretty far down on the priorities list at Inprise. Now, however, it's anything but. Inprise has formed a spin-off company [interbase.com] (along with some VC money) to support and develop it further. Now Interbase is getting the publicity, development support, and openness that it needed all along. Version 6 should be a huge improvement, especially on Linux (which is now a primary development platform, rather than an afterthought).

--JRZ