Cal Jacobson's Blog

July 18 2007 Update: It appears that the thread in question (linked below) has now mysteriously dissappeared from the TV3D site, after remaining online for God knows how many days. What an amazing coincidence!

Preface
I’ve written about the Truevision3D (TV3D) engine before, and I’d like to stress again that my disclaimer still applies. My inability to find a good, concise word to explain my fascination/frustration/hope/dispair with TV3D hasn’t changed, either.

I stopped by the TV3D forums today in the hope that version 6.5 has finally made it out of the closed beta it had been languishing in since early 2005. Unfortunately (and perhaps unsurprisingly) it was still exclusively in the realm of paying customers.

A few discussions caught my eye, though, and I was motivated to add my two cents after reading them. I’ve chosen to do so here again, rather in the Truevision forums. I’d rather not be considered a rude guest in somebody else’s house, after all…even if I’m right.

This locked thread started off innocently; user birvin asked if there were any 6.5 updates, as he wanted to play with networking code. Four posts in, and cactus3d commits the cardinal sin of saying:

About 6.5 it’s not as simple as it seems… i mean 6.5 has been “nearing open beta” or “very stable” for nearly one year now… )c:

if you look in the forum you’ll find posts from september/06 saying that…

Oh-oh.

TV3D Project Manager darqSHADOW steps in a little bit later and says:

I have delayed the timeline over the past year for very valid reasons, both corporate and personal issues have dictated this. The engine has been stable since its creation, because of a solid design — however the needs of some customers have driven us into new technology additions, and 6.5 is now becoming a full 3D system with many tools, utilities, plugins, and the like — not an easy task for a team of 2.

I can’t argue that it seems like a lot of work for just two developers. Whether the reasons behind the delay are valid or not can be debateable, since darqSHADOW doesn’t go into further detail.

Things take a turn for the worse when Manjinder Singh Lamba makes the following post:

I have been waiting for the 6.5 dev but the time has come when I have to decide to buy another engine, because of the lack of a sdk which I could test. I have already decided upon the alternative Engine as there is no sign of 6.5 release. If darqSHADOW would see this and consider that the release is very near then please PM me so that I may wait. I would buy the license if only I see a sdk-demo and there are no restrictions like others cant release their work until its official.

He then goes on a couple of posts later and states (sic):

These are the two most important things that are needed for my or any games. 6.3 or any previous engine doesnt supports either of these. There is no way to know how easy and good are they in 6.5. I dont want a crippled license which imposes a restrictions on wether and when I can upload my games or demos.

Assuming that English isn’t MSL’s native tongue, I interpreted his posts as: I’ve waited for 6.5 to be released so I could see if it meets my needs, but I’ve run out of time for fact-finding. I’m going to purchase an alternative engine, since I cannot currently evaluate the 6.5 API nor can I release a product while it is in beta. If TV3D is close to release, I’d reconsider if somebody contacts me.

Maybe darqSHADOW (and where’s Sylvain, by the way?) didn’t interpret it the same way. He responds:

I find it suprising the number of people who say “I need to try it before I’ll buy it”. That has to be the most ridiculous argument I hear from people. I just spent $40k on Great Plains from Microsoft — do you think I got a free trial before I bought?

darq gets some gentle pushback from the community this time around, though. cactus3d states that he prefers to try before he buys, and Zaknafein goes so far as to point out that Microsoft GP had a 90-day trial available.

Manjinder weighs in again and says:

I consider this as your final opinion about the status of 6.5. I here from shall proceed to buy A7, and thanks for replying so considerately and honestly.

I have been testing many and every software that could fancy my needs. At the final stage only A7 and TV3d are the ones, which are perfect for my requirements. And as there is lack of 6.5, I shall go for A7 next week. Maybe some years after when there is need for a new engine then I may choose 6.5 (7).

I can’t fault him for his politeness, though he should have left well enough alone and not responded to ovek‘s suggestion so go ahead and buy version 6.3 to gain access to 6.5 since when its released he’ll “end up back here anyway”. Manjinder responds:

There is no point to rely upon such a claim. I am overly confident about my needs and it was just due to time I have spent with 6.2-3 that I wanted to consider my choices again before finalising the engine. I am now confident to the fact that it will take much longer for 6.5 than anybodies considerations. I am sure that a couple of years, till the death of DX9, and the sovergin rule of DX10, there wont be any 6.5.

By then XNA, Irrlicht will be strong enough to suffice my needs.

Good luck to the would be TV3D community, and success to all.

Bye,
MSL.

darqSHADOW feels the need to get in the last word, and he posts the following before locking the thread and silencing the debate:

I am sure that a couple of years, till the death of DX9, and the sovergin rule of DX10, there wont be any 6.5.
And you are wrong.

By then XNA, Irrlicht will be strong enough to suffice my needs.
I find this funny, since MS has been courting us to be their engine of choice in XNA for a year now.

You have made your engine choice, and you have no faith in Truevision3D — therefore there is no longer a discussion to be had here. I will not get into another debate on A7 vs TV3D, its been gone over by many users with experience in both in the engine comparison forum, suffice to say that your choices are limited in the 3D engine arena, and whether you believe it or not writing one is not easy. TV3D 6.5 is a top-tier engine and has been proven to outperform all competition (stats will be released with the new site) — given the community, the support, and the product most people end up back here sooner or later…

Now I can sympathize with DS a bit here…after hearing this for years now, I’m sure this gets rather old. However, here’s the thing: birvin started this now-locked discussion and never got an official answer to his question.

Strike one.

Here’s something else I noticed: MSL was polite to a fault, but was essentially turned away by darqSHADOW. Now maybe the TV3D guys aren’t interested in throwing a free beta entry to every polite guy that stops by, and maybe MSL could turn out to be some flake kid who lives with his mom and will never build a demo, let alone a game. I kind of doubt that based upon his discussion elsewhere in the thread about shaders, physics implementations and licensing issues, but maybe that’s the case. Still, it seems like just dumb business to me to brush somebody off who states his case in a civil manner and offers to work with you.

If he’s a goofball, then sure…blow him off. But in a public forum, where the whole world can observe your lack of customer relations skills? Dumb. Strike two.

And how about the XNA thing? This isn’t the first time darqSHADOW has crowed about talks with Microsoft but I have to wonder…if a multi-billion-dollar company is coming to you and asking you to be their ‘engine of choice’ for a framework to be used on both PC and XBox….why in God’s name haven’t you jumped on that? If it was a great deal, then you should have the capital to hire more people and get TV3D done. If it was a bad deal, why mention it in the first place? With no more info to go on, this just doesn’t make sense. Foul ball.

Finally, there is the mockery of somebody wanting to “try before they buy”. Gee, what nerve of people to be careful shoppers! How dare they be frugal with their money? And to use such a weak argument to prop up your position — one that is blown away less than 30 minutes later by one of your more loyal users, no less. Strike three.

Everybody has their bad days, and maybe I’ve just cherry-picked a bad one for darqSHADOW. If so, then I’m sorry. But it sure seems like the tune playing in Truevision3D land is the same one that was playing three years ago.

I stand by the statement I made in December 2006: if your product has been in a closed beta for well over two years (now over three), you are doing something wrong.

The TV3D community is not healthy: it is fragmented.

• Exhibit One: people are selling map editors for versions that are not yet available!
• Exhibit Two: resources for the old and unsexy engine are drying up.
• Exhibit Three: the Showcase is rife with 6.5 examples that can’t be run by anybody outside of the closed beta. And some of those are more than a year old!
• Exhibit Four: New users are asking the same questions others were asking over a year ago and they still can’t get a decent answer!

What’s the solution here?

I guess I should propose one, given that I’ve devoted enough time to belly-aching about the problem. My solution is this: open the Truevision3D 6.5 beta immediately. What’s the worst that will happen…people will go “oh no, this is a terribly unstable…uh….beta”?

More beta testers means more eyes on the product and one would assume, eventually fewer bugs that make it through to the final release. At the very least, your community is no longer split so drastically between the haves and the have-nots: nonpaying users can see what is in store for them, and the existing 6.5 beta testers get an infusion of fresh blood and ideas. Continue to forbid commercial 6.5 development if you wish, but make it happen now.

It won’t be perfect. We get it. Continue to piss away your momentum and community interest and the only people who care about 6.5 when it comes out are the small fraction of developers who have had it in their hands for months (if not years) now.

XNA, Torque, OGRE and a host of other competitors will not wait. Act like an agile two-developer team instead of a plodding monolith, and your community will forgive your mistakes.

UPDATE: If you can stomach my rambling after this post, here’s a new one on the topic from July 2007.

I’d like to preface this post with the following disclaimer, so as to deflect any low-hanging criticism before it appears:

What I’m about to write about is an incredibly minor, gnat-on-an-elephant’s-back annoyance. As far as I know, nobody will be harmed because of this issue. As an Internet freeloader and forum lurker, I’ve invested nothing but a little personal time into the subject at hand. Furthermore, the applicable developers owe me nothing. I do not wish this to be construed as a personal attack against them. I feel compelled to comment since I’ve followed this for a while and it seems so atypical of anything else I’ve seen in terms of software.

We all on the same sheet of music? Fantastic. Let’s have at it.

Several years ago I came across a great graphic engine called Truevision3D (TV3D for short). At the time I was getting hooked on the .NET experience and this engine not only supported the Framework (via COM) but had a nice collection of samples available so that I could quickly get up to speed with things. The community was small but spirited and helpful. Best of all, the engine was cheap: free if you didn’t mind a watermark, up to $500 if you wanted unrestricted rights.

I played with the engine, made some cool orbit simulations, and considered writing a game with TV3D. But on the horizon there awaited even more goodness: the incredible version 6.5. More than just three minor updates from the current v6.2, this new release would be a major rewrite of things and would feature shader support, an overhauled network engine, better terrain and model support and an integrated physics engine. As great as the current engine was, the new version sounded like the Best Thing Since Sliced Bread(TM).

I chose to wait, since I was busy at home and work anyways and had no burning need to crank out a game immediately. I felt that I could take my time, get even more comfortable with 6.2 and then start with 6.5 when it came out rather than try to retrofit older code. My time wasn’t too far away, I figured, since developer darqSHADOW announced in May 2004 that community members were to be slowly added to the beta program and I made it a point to sign up.

Months passed without word from the TV3D team about wider beta involvement. Still, most of the forum discussions focused on 6.2, even though the occasional 6.5 showcase thread would appear to keep us salivating. In December 2004 developer Arli announced that purchasers of 6.2 would automatically get into the beta; because I hadn’t purchased anything I didn’t get in on that action…but that was OK, since this most likely meant that a wider beta — or a full release — was in store for us in 2005.

By April 2005 it seemed like there were more and more questions about 6.5 appearing…the oft-asked “when is 6.5 coming out” forum posts (inevitably asked by new users) were receiving fewer and fewer polite responses by the forum regulars. September came and passed with little more than a new video posted by lead developer Sylvain along with the only “development diary” posted by the dev team for all of 2005. More and more 6.5 examples and sample code were being posted by beta testers; unfortunately this wasn’t as useful to the 6.2 folks.

By this time I wasn’t following TV3D as closely as I had before; whereas previously I visited the user forums on a near-daily basis, now I was coming around perhaps once a week.

Then, perhaps once or twice a month.

I stopped by in Spring 2006 and found that I wasn’t necessarily alone in my lack of optimism about 6.5′s progress. The community wasn’t quite as friendly as it was before: 6.2′ers were getting sick of seeing screenshots, videos and code samples exclusively for 6.5, and the 6.5′ers were sick of listening to the 6.2′ers whining about timelines instead of just ponying up a simple $150 to get in on the beta.

Each side, I think, had legitimate points: the 6.5′ers were mostly veterans of the forums and were sick to death of newbies asking the same questions; the occasional troublemakers and inflated-entitlement leeches didn’t improve the stance of the 6.2′ers either from what I could tell. On the other hand, the 6.5′ers tended to avoid such questions as “how can somebody publish a game using a beta?”

By Summer 2006 things continued their slow but gradual downhill slide. When one forum poster asked “Are they fooling us?” in regards to the ever-closed, never-ending beta, the conversation degenerated into the typical “they owe you nothing” dogfight that I’ve seen on a dozen other discussion boards (usually involving disgrunted game customers versus fanboys), with one developer chiming in and mentioning that the website was a key part of the 3D engine’s release. Then the discussion turned variations on “oh, well I’d rather wait and have something good than have a product that is rushed out” and “you’ll only complain louder if you don’t have proper documentation, which isn’t ready yet!”

It was at this point that I checked back out.

I do agree that in most cases it is better to wait for something great than have something barely adequate immediately. Small development teams may be forgiven for taking a long time to accomplish a task, especially if they are doing so on the side — that is, in their free time. Timelines for software are notoriously difficult. Building a 3D engine that itself is a product must be incredibly challenging and difficult.

But if your product has been in a closed beta for well over two years, you are doing something wrong.

First and foremost, you’ve Balkanized your community: instead of a vast majority that uses one version (with the occasional late-adopter dropping in on occasion), many of your advanced users and evangelists are using the newest toys while everybody else is in the slow lane. The gurus who would normally help the newbies are busy frolicking in greener pastures and are less inclined to “slum it” with the mundane problems of yesteryear.

Secondly, you’ve left some (perhaps many) of your bread-and-butter customers behind. These are people who started on their own projects a year (maybe two) ago and figured that as their own deadlines approached, your component would be ready for use. Maybe they’ve used the old product and their design requires a feature of your new product…or maybe they got in on the new version but can’t release their product because you are holding them up.

I can’t honestly say that watching the Truevision3D beta saga drag on angers me…because I really don’t have a dog in this fight. As I stated earlier, I haven’t invested a dime into this engine. A lot of hope, maybe, but no money. The words confusion and dismay are closer to what I’m feeling, along with disappointment and frustration. I’m having a lot of trouble comprehending how something with so much promise, with so many smart people working on it, can flounder like this.

I’d love to really find out why — but it’s clear that the TV3D forums aren’t the place for this discussion anymore, if it ever was. To have a voice, apparently you need to fork over at least $150…and at that point, you’re no longer interested in the opinion of anybody who hasn’t done the same. And I’d rather not pay somebody that much money for a product that could stay in beta for another two years.

Am I crazy?

This is a trivial thing — especially from my position — but does this situation strike anybody else as being normal or reasonable?

Postscript: the questions about 6.5′s release date continue. Two-and-a-half years later.

Postscript 2 (11 January 2007): TV3D user Zaknafein boldly posted a link to this entry in the Truevision3D forums, which explains my sudden surge of traffic.