?Houdini dev team size?

[wesserbro]

In every discussion about Houdini if someone points to its weak sides plenty H-funboys (on sesi and odforce) arise with two arguments:
-go find another software which will suit you better
-sesi is a small company to make big changes!

And recently i'v looked into theirs linkedin: 82 Side Effects Software employee in LinkedIn network. There may be also some not registered in LinkedIn employees.
Is it considered to few for a single-product company? I believe Softimage had like 5 and a half people in last years, or am i wrong? 5 is nothing of course , but 20-30 is a good number i thought...

Am I missing something? Can someone throw some info about number of developers on other products, like maya (not the whole autodesk, only maya devs), modo, etc...

P.S. I know not all of that 82 are developers, but maybe 50-60 are, or not?
P.S2. Ok, i'v counted them (yep, too much time ) - 16 actual devs not including trainees, QA and mathematicians. hmmm....

[Rork]

I rather have a small, lean and organized team of devs compared to the cluster**** of organization rings that must be AD.

Look at the team of Fabric, is it 'bad' that it is just a small group?
And looking at all the new stuff that is integrated very quickly in Houdini, I rather see that than a 'committee' voting for new features ;-P

Also, Modo's team isn't really big, nor is Cinema's.

So..... What's the point you're trying to make?

[NNois]

... I believe Softimage had like 5 and a half people in last years, or am i wrong? 5 is nothing of course ....

I agree software devellopers should inform customers about that.

And i will start a flame war but... look at Autodesk, just check the linkedin page of Guillaume Laforge, and be surprised, actually he did all the works on new Softimage feature from 2010, Crowdfx to Ice modeling...
I'm not saying he was alone, but the truth is very close i guess.

So i was wondering, how could Autodesk has the balls to sell a software to thousands of customers with nearly ONE feature developer, this is just crazy.

[wesserbro]

So..... What's the point you're trying to make?

The point is that if someone can throw some light on a DCC development: what is considered a big team, whats small. What is the time to implement some modeling feature to DCC (approx. of course). On Houdini they say it can take most likely several new versions before houdini modeling became at least "not embarrassing". At the same time there is one guy on sesi forum who claims he is making a plugin with all fancy modeling features by himself for several months and its almost ready. I'm just asking people who are close to this business to enlighten us, "tool-users", so we were more informed about tool-making.

I know, the question is too general, but any info would be appreciated. Like : "I knew a guy, who was working at <DCC name>, and he said that to implement tweak tool it took 2 years and 10 devs but then they realized that it makes build unstable and dismissed the feature".

[luceric]

Houdini's revenue is probably driven by big studios doing effects and their consulting business.

So if they have 20 developers, a lot of these people they're working on putting out fires for these people. It wouldn't surprise me if right now a few devs over there are spending all of their time fixing a some memory leak or a performance problem for some client or another. They're not just staring at the ceiling trying to find things to do.

Don't get sucked in the internet message board battle vortex. It doesn't matter how much time and resources it takes to write a "tweak tool". They're not working it because it's not a top priority. Sometimes a programmer does squeeze in something that becomes a feature, and some other times a large client will pay or otherwise demand something that wouldn't otherwise get done. It's hard to predict, but generally you can see what the product focuses on. Houdini just doesn't focus on modelling. Softimage didn't focus on ArchViz, etc.

I remember a thread I got involved in where someone asked why softimage wasn't working on batchserv. The answer was lack of resources. The users weren't happy, saying that this other render farm tool was made by just one guy so what's the damn problem. We can debate that it was more expensive for Softimage to make/document/support a product than an individual, and that would be true, but at the end of the day the question is that you want to spend your resource time where the money comes from, because you're never doing enough. And you need to remember that if you take a resource to do something, it's not doing that other thing. So if you got a guy working on a tweak tool, and none of the paying clients are really modelling in Houdini but they have things they want that are not happening, then that's that difficult call to make.

[wesserbro]

This explains alot, luceric, thanks! Very mature answer to very immature question )

[Werner]

Houdini's revenue is probably driven by big studios doing effects and their consulting business.

So if they have 20 developers, a lot of these people they're working on putting out fires for these people. It wouldn't surprise me if right now a few devs over there are spending all of their time fixing a some memory leak or a performance problem for some client or another. They're not just staring at the ceiling trying to find things to do.

Don't get sucked in the internet message board battle vortex. It doesn't matter how much time and resources it takes to write a "tweak tool". They're not working it because it's not a top priority. Sometimes a programmer does squeeze in something that becomes a feature, and some other times a large client will pay or otherwise demand something that wouldn't otherwise get done. It's hard to predict, but generally you can see what the product focuses on. Houdini just doesn't focus on modelling. Softimage didn't focus on ArchViz, etc.

I remember a thread I got involved in where someone asked why softimage wasn't working on batchserv. The answer was lack of resources. The users weren't happy, saying that this other render farm tool was made by just one guy so what's the damn problem. We can debate that it was more expensive for Softimage to make/document/support a product than an individual, and that would be true, but at the end of the day the question is that you want to spend your resource time where the money comes from, because you're never doing enough. And you need to remember that if you take a resource to do something, it's not doing that other thing. So if you got a guy working on a tweak tool, and none of the paying clients are really modelling in Houdini but they have things they want that are not happening, then that's that difficult call to make.

Seems like they went and did just that Luceric.

[Hirazi Blue]

But how were resources allotted in "integrating" code like your very "own" (IIRC) FXTree?
Isn't the industry "secretly overrun", from the perspective of the "smaller players",
by the demands of the "big players", who could easily afford mixed pipelines anyway?


(we lack an Illuminati smiley )

[luceric]

the fxtree was a gamble, and a rather inexpensive one, given that I did everything as a one-man army. (I think I was making 35k a year at the beginning!) We add a second person at one point, perhaps for 8 months. As a bonus, it delivered a early promise to have compositing in XSI. And on top of this, I enhanced some core services of XSI as part of this project, while managing the UI team.

[luchifer]

XSI FXTree was yours? Just want to say THANK YOU!

[Hirazi Blue]

XSI FXTree was yours? Just want to say THANK YOU!

Indeed +1