PDA

View Full Version : How is LW programming development being managed at Newtek?



robertoortiz
07-23-2014, 07:28 AM
Hi guys,
I have been wondering about this for a while… (just speculating here guys)

I wonder if part of the problem we are experiencing right now with the painfully slow overhaul for LW is
is HOW development is being managed at NT.

If the company has become a bit top heavy in terms of management and approval structures?

The reason for asking this is that top heavy organizations tend to favor development faddish practices development from the late 90's and 00’s that frankly have proven time and time again that they DONT WORK and stifle innovation.

How can you experiment and innovate when you have to explain EVERY DESICSION to a chain of command. And half those people in that chain of command are suits?


I honestly wish them luck, since they a have time and time again to be highly ineffective in allowing for creative coding. And frankly developing using such rigid development process tends to lead to
At best lackluster and unimaginative features at worst, failure.

This is why in recent times most IT innovation comes from VERY SMALL teams (hell sometimes ONLY ONE GUY, like the developers of 3D Coat or LW Cad) since they are not bound development guidelines that benefit only a manager in a bureaucracy.

I do wonder if part of the problem is HOW development is being done at NT.


What do you guys think?

-R
More on what I am talking about:
http://businessjournal.gallup.com/content/152429/Cost-Bad-Project-Management.aspx
http://www.hks.harvard.edu/m-rcbg/ethiopia/Publications/Top%2010%20Reasons%20Why%20Systems%20Projects%20Fa il.pdf

pinkmouse
07-23-2014, 07:58 AM
Indeed, speculation, and pointless to boot. LW3DG don't even share any info on what does or will directly affect users, so why would they share any info about that?

And even if we did know, what difference would it make?

MarcusM
07-23-2014, 08:08 AM
In my work we using JIRA Agile. We have so many projects and tasks that without this we would be much slower as a team in creating products.
https://www.atlassian.com/software/jira/agile

For sure LW3DG using something similar. They have programmers in many countries and skype and google accounts/files are not enought ;p

robertoortiz
07-23-2014, 08:16 AM
In my work we using JIRA Agile. We have so many projects and tasks that without this we would be much slower as a team in creating products.
https://www.atlassian.com/software/jira/agile

For sure LW3DG using something similar. They have programmers in many countries and skype and google accounts/files are not enought ;p

Ok you got my attention.
Tell us a bit more about this implementation of Agile development..
Is is true agile or (lets have a daily meeting about what you are doing type of agile.
( I have seen this stupid implementation of agile done VERY BADLY and it has flamed out in a spectacular fashion)

MarcusM
07-23-2014, 08:26 AM
Ok you got my attention.
Tell us a bit more about this implementation of Agile development..
Is is true agile or (lets have a daily meeting about what you are doing type of agile.
( I have seen this stupid implementation of agile done VERY BADLY and it has flamed out in a spectacular fashion)

In short, from time when my team grow up and more projects star coming we start using exactly this system from upper link. Once time on two weeks we showing our work to all team and planning next tasks. Every morning we have short meeting and telling what we done day earlier and what we plan doing today. In browser we have control on development and se progress or problems. I must admit that on beginnig i was sceptic but after few months everybody accustomed. However, this system must be in some parts adapted to type of work and team.
We using "scrum poker" on smartphones ;]
http://www.youtube.com/results?search_query=agile+jira

gclayton
07-23-2014, 10:21 AM
Ok you got my attention.
Tell us a bit more about this implementation of Agile development..
Is is true agile or (lets have a daily meeting about what you are doing type of agile.
( I have seen this stupid implementation of agile done VERY BADLY and it has flamed out in a spectacular fashion)

Roberto, I've got to agree with you about Agile. I worked for a company that implemented Agile and all we did was to have meetings, every day to explain what we did the day before and what we were going to do that day. I retired to get a way from it.

robertoortiz
07-23-2014, 02:27 PM
Also part of the problem with recent development methods is that for me it seem more obsessed about process than actual goals.
All of the sudden you got people with tittle of "System Architects" telling developers how do their jobs.(You can tell I am not a fan)

In effect the reason of why the work actually done is lost in the shuffle and instead of being something that is there to aid development it become
THE REASON why the development is done the way it is.

In plain English, how the hell you can innovate and experiment with idea when you have to account in detail what you are doing and what you WILL BE doing.

This might be of interest:
Dr. Dobbs Journal: The Corruption of Agile
http://datavizblog.com/2014/03/21/dr-dobbs-journal-the-corruption-of-agile-part-1-of-3/
http://datavizblog.com/2014/03/25/dr-dobbs-journal-the-corruption-of-agile-or-what-the-agile-manifesto-really-said-part-2-of-3/

BTW here is the original AGILE manifesto...

Principles behind the Agile Manifesto



We follow these principles:
Our highest priority is to satisfy the customer
through early and continuous delivery
of valuable software.

Welcome changing requirements, even late in
development. Agile processes harness change for
the customer's competitive advantage.

Deliver working software frequently, from a
couple of weeks to a couple of months, with a
preference to the shorter timescale.

Business people and developers must work
together daily throughout the project.

Build projects around motivated individuals.
Give them the environment and support they need,
and trust them to get the job done.

The most efficient and effective method of
conveying information to and within a development
team is face-to-face conversation.

Working software is the primary measure of progress.

Agile processes promote sustainable development.
The sponsors, developers, and users should be able
to maintain a constant pace indefinitely.

Continuous attention to technical excellence
and good design enhances agility.

Simplicity--the art of maximizing the amount
of work not done--is essential.

The best architectures, requirements, and designs
emerge from self-organizing teams.

At regular intervals, the team reflects on how
to become more effective, then tunes and adjusts
its behavior accordingly.

robertoortiz
07-23-2014, 03:44 PM
Sometime later I read My own post and I say to myself...Wow that was a LONG post. Sorry for that guys.

COBRASoft
07-23-2014, 07:56 PM
I hate daily meetings. As a software manager/analyst/developer, I stay far away from things like SCRUM and Agile development (in pure sense).
My system is a bit different. We have a meeting or two in which we decide who does what, then we estimate the time needed to develop that chunk of a program. Our company structure is quite simple: 1 software manager, no extra architects, but some good analysts/developers (not just code writers).
If a developer is stuck, he can contact another one for help/explanation for some time (calculated risk). If that doesn't help or the time has run out, the problem becomes escalated towards me to help fix the problem. Schedules are adjusted as needed.
Along this process, we contact the customers regularly to discuss the current status and what should be done different/better. Customers mostly don't really know what they want, so we have to be able to adjust very fast :). Once I had to release more than 3 version in 1 day due to customers opinion changes, crazy world!

robertoortiz
07-24-2014, 08:25 AM
Once I had to release more than 3 version in 1 day due to customers opinion changes, crazy world!

Back in the 90 this sort of development was MORE common, and frankly more systems were created faster becauseo of this mentality.
developer were allowed to fail and share their results with a small group of users. And multiple revisions of the program would be encourage until the client got what he wanted.
When done right(With limited number of revisions) this was smart because you ended with an app that clients wanted to use.


I meant lets look at all the 3D apps that came from that period.
LW, C4d, Maya and Houdini and Soft. Durign the 90 the application were allowed to grow.

And I think honestly that the development philosophy FOR CORPORATIONS from that time was more conductive for innovation.

It was not about following the latest development fad, it was more about TRYING TO REACH A SPECIFIC GOAL
( I want to do Character Animation in 3D, I want to do Hair) and work towards it.

for some reason, in the 00 the philosophy changed towards more architecture development.
Which is fine and good for certain systems, NOT ALL SYSTEMS.
The problem of this kind of developmenti s that it can be a rabbit hole of time. You spend most of your time thinking of possible future capabilities intead of, you know, developing programs. Next thing you know yo uhave been developing a system for over 4 years..

Hell we developed systems back them (that are still in OPERATION BTW) that were developed in months.


Now I feel that LW development is stuck to much in abstract targets (improvements in VPR rendering, Ability to export STL, New nodes! jay!)
instead of offering more clear goals.

And by that I mean
Let make a program that makes Character Animation easy
Lets make a program that makes 3D printing Easy.
Hell I want to create custom 3d models of Tools at Home Depot (they are offering 3d Printers now) how can you help me?
Can your program that allow me to do stuff for MY INDIE Game easy AND FAST.
I am doing LIVE broadcast graphics, I want a program that allows me to do 3D text FAST.

COBRASoft
07-24-2014, 10:22 AM
Yes, indeed. I'm rather oldschool towards modern developers. But I get the job done in time.
Today, many teachers/developers say OO is the only way to develop a good program. WRONG! We have many students trying to develop something in their last year in our company. Most fail miserable due to the lack of knowledge of simpler solutions. They are forced in the OO paradigm and forget that their solution actually has to work in the end within 3 months time.

Ok, I agree OO has advantages, but a program becomes slower due to OO, only maintenance becomes faster and easier in some cases. My own ERP application is now 11 years old and it has been largely rewritten over the years. The old code still works side by side with new code because I have foreseen this possibility from the start (with interfacing). Old code can call new code and vice versa. This concept will work for the next 10 years without a problem. I say, use OO where it is handy and a big time saver, don't use OO where it is used just to be OO.
Let's give a sample: a light, camera, ... can be OO based (definition). The render engine... Well, let's hope not for speeds sake. But the render engine has to be able to read the OO classes thanks to interfacing and inheritance. This way, the render engine stays fast and future objects stay possible as long as they implement these interfaces.
The other way is also possible. A new render engine can be made, based on the same interfaces and future ones.

To be clear, an interface is like a contract. It specifies what properties must be implemented. This could be color, position, rotation, ... This forces the developers to maintain some development agreements without the need of a meeting every day :).

About LW... Well, we don't know what they're doing right now concerning LW12 and their future development. I remember a post where I asked why Rob Powers became head of a development team without being a developer himself. He gave nice and clear answers to 'defend' his position. In the coming years we'll see how successful this will be.
Overall, I think he has done a good job until now and he deserves credit for that (together with his team). But the time has come for the new stuff.

evolross
07-24-2014, 07:02 PM
Not sure how the development is actually managed, but I'm pretty sure a lot of the developers/programmers are scattered in various parts of the world. And I believe each developer focuses on a specific feature and owns its development.

jasonwestmas
07-24-2014, 08:48 PM
Oh, are we on the "easy button" argument again.

lightscape
07-25-2014, 02:02 AM
About LW... Well, we don't know what they're doing right now concerning LW12 and their future development. I remember a post where I asked why Rob Powers became head of a development team without being a developer himself. He gave nice and clear answers to 'defend' his position. In the coming years we'll see how successful this will be.
Overall, I think he has done a good job until now and he deserves credit for that (together with his team). But the time has come for the new stuff.

Was Jay Roth a developer? How about Brad Peebler when he was head of Luxology? I think they weren't not sure.

robertoortiz
07-25-2014, 06:09 AM
Oh, are we on the "easy button" argument again.

Yep. For me the proper terms are good usability and workflow.
The goal is to make the different tasks, using your toolsets, flow well.

This is the advantage of GOAL ORIENTED DEVELOPMENT, that your job as a developer is to streamline the program in a way that users with different levels of experience can do their jobs as fast a possible.
That is it. PERIOD.
Believe or not the bells an whistles of a tool is arbitrary , because according to that philosophy in development, if a tool is too complicated to use it might as well not be there.
(IKboost I am looking at you).

To test how well an app works we would interview and record users using our apps to see what they were doing with the program, and usually we were stunned on what we found.
It is amazing how fast a developer can get "tunnel vision" because they are so close to the code.
and not see that a tool is hot mess of menus to use.
Rule of thumb, if you needs to do multiple 10 minute tutorials to use a tool, then the tool is too complicated in the first place.


Was Jay Roth a developer? How about Brad Peebler when he was head of Luxology? I think they weren't not sure.
A non developer can be a great Project manager. And Rob Power has proven to have very good ideas about the evolution of LW.


And to all LW dev reading this.
We need to thank you for all the hard you guys are doing.
Also don't be afraid to innovate, Chornosculpt showed us that you guys have some great ideas in reserve..

jasonwestmas
07-25-2014, 01:33 PM
"This is the advantage of GOAL ORIENTED DEVELOPMENT, that your job as a developer is to streamline the program in a way that users with different levels of experience can do their jobs as fast a possible.
That is it. PERIOD."

I just don't see how this is possible. That's like expecting two people to score the same on a written language test but one studied more than the other.

BTW Did you know that you can save 15% or more by switching to geico?

COBRASoft
07-25-2014, 04:09 PM
I'm talking about my own experience here. I see many developers struggle at some point and nobody to help them. Yes, 'Google' has helped a lot, but most code out there is scholar code and not real business proven stuff.

I've seen meetings where the manager has to decide what path to follow. This can be code patterns, language choice, platform, ... Most of the time non-developer managers have no clue what their developers are talking about and leave the decision towards their best/lead developer(s).
A ship can have only one captain and this doesn't have to be a developer at all, but let the development decisions over to people who know what their talking about :).

Now, you have several degree of non-developing managers. Some have no clue how a computer actually works and some know what logic is and understand the pitfalls that come with software development.

As said before, I agree Rob and his team did a great job so far, but it is time to see the results in the coming 2 years (LW12...).

jasonwestmas
07-25-2014, 07:06 PM
I really wouldn't know what to compare Lightwave to. It's such an odd app in good and bad ways. So I'm kind of surprised when people have no doubt Lightwave is going is good direction.

COBRASoft
07-25-2014, 08:23 PM
I have to ask, what is the good direction then for LW?

- Is unification the most important thing
- Making sure that every tool works in a much better way (incl. snapping and the like)
- Speed improvements all the way?

These are 3 total different directions and decisions to make. The first requires a time consuming rewrite, the second requires detailed focus per tool and the third requires an overhaul in the underlying system and an extra check if every tool still does what it does today.
I hope for the third. I'm more found for better foundations and keep the rest working and later improve every detail to new standards.

hazmat777
07-25-2014, 09:13 PM
I have to ask, what is the good direction then for LW?

- Is unification the most important thing
- Making sure that every tool works in a much better way (incl. snapping and the like)
- Speed improvements all the way?

These are 3 total different directions and decisions to make. The first requires a time consuming rewrite, the second requires detailed focus per tool and the third requires an overhaul in the underlying system and an extra check if every tool still does what it does today.
I hope for the third. I'm more found for better foundations and keep the rest working and later improve every detail to new standards.

I guess it just depends on what LW3D Group think. To me, undo, graph editor, and unification are the improvements I've been looking for since 9.6.1 ... Oh well...

jasonwestmas
07-25-2014, 09:42 PM
I have to ask, what is the good direction then for LW?

- Is unification the most important thing
- Making sure that every tool works in a much better way (incl. snapping and the like)
- Speed improvements all the way?

These are 3 total different directions and decisions to make. The first requires a time consuming rewrite, the second requires detailed focus per tool and the third requires an overhaul in the underlying system and an extra check if every tool still does what it does today.
I hope for the third. I'm more found for better foundations and keep the rest working and later improve every detail to new standards.

For starters. . . .not bolting new ideas onto old rickety foundations.

COBRASoft
07-26-2014, 05:24 AM
I agree, new foundations are a must. This takes the longest time to accomplish :). The hydra system is only a part of these new foundations, all be it a very promising one!

ianr
07-26-2014, 07:15 AM
! Did he say the 'H' word .?

Yep, Yep he did!

Well, last time I mentioned that
they slapped a 'D' notice on it!

Did they really.?

Yes they did!

an excerpt from: The App that Came in From the Cold

robertoortiz
07-26-2014, 07:32 AM
Part of the problem right now with LW dev is one of OPTICS.
When you do architecture re- development while dealing with a legacy system
development can crawl down do a glacial speed.
And from the point of view of the user base nothing seem to be going on FOR YEARS.


This was the the same problem Softimage had in the 90's when they decided to rewrite their program
Softimage and do a complete overhaul of their program called XSI. (AKA the best 3D program from the 00's)

That delay cost them a bug chunk of their market share and Maya just jumped right into the void.
The irony is that the developers of Soft, right towards the time before they got bought by Autodesk where experts of Goal Oriented Programming.

Their SPECIFIC Goal was to develop a world class Animation system that small studio could use.
And the developers keep an bidirectional communication channels with the user base in the development of XSI.
BTW a lot of the tools they developer for Soft have made a migration to Maya after they got bought out.

The incredible thing is that when they were eliminated by Autodesk they had bar none the best product in the market.
What killed them is honestly market momentum, nasty backroom deals Autodesk had with studios (Don't mention X program ever in interviews and you will get a volume discount)
and total dominance of the educational market. (BTW that last point is why they are VERY SCARED of Blender, something that I bet they are working to "Fix")


The reason I am so hard on LW to "make it easy" is that the market practically demands it.
The market base that would be willing to invest in an expensive (and it is in their eyes) app like LW are looking for a product that provides a lot of return for their investment.
They just need a reliable platform that can help them develop multiple varieties of content with speed. The goal I mentioned before are the kind of thing these users would demand.
In terms of the people looking to buy a 3d App things have changed a lot. We the not in the 90's hell we are not in the 00's where people from more technical mindset are the only ones doing 3d.
Now traditional Artists are the ones pushing the market.
You cannot to allow yourself to scoff at the userbase DEMANDING for "Make it Easy " features.
A company like Maxon and freaking Adobe Gets thi and they are laughing all the way to the bank.

jasonwestmas
07-26-2014, 07:58 AM
I agree, new foundations are a must. This takes the longest time to accomplish :). The hydra system is only a part of these new foundations, all be it a very promising one!

I have and will be further investing in Chronosculpt. I have had little interest in Lightwave classic these past two years. I think I've seen as far as it can go without it being gutted out.

robertoortiz
07-26-2014, 08:02 AM
To be blunt, if I was Rob I would EOL Lightwave and put all my efforts into Chrono.
BUT what scares me deeply is that Chrono will be going only after the high end market.

COBRASoft
07-26-2014, 08:41 AM
Let's not forget 1 thing: VPR.

Remember the days where we were told it was impossible in classic LW. Some internal shifting happened (not to be blunt) and suddenly we got the VPR inside LW and at great speed too. This alone gives me hope that they are indeed doing a lot of work under the hood, waiting to be revealed when the time is right.

jasonwestmas
07-26-2014, 08:59 AM
To be blunt, if I was Rob I would EOL Lightwave and put all my efforts into Chrono.
BUT what scares me deeply is that Chrono will be going only after the high end market.


I think the whole "this is high-end" because it has capability "x" and not "y" is going to eventually disappear. What truly makes a 3d application popular is the ability to SCALE your projects up and down with it. With intelligent design, a 3D package doesn't have to sacrifice tight deadlines in order to also serve the larger scale projects. Putting together nodal or layer tree systems, order of operations "stacks" like they were building blocks has worked in the past and have been very popular. Softimage was a good example of that. Something as flexible as a sculpting application (there are many different types of those now days) like chronosculpt can be used for any type or scale of project that requires an animated 3d image output and you would really like to reshape your character's profile over time, not just for simple cleanup.

So where exactly does deep fear come into the picture here?

jasonwestmas
07-26-2014, 09:05 AM
Let's not forget 1 thing: VPR.

Remember the days where we were told it was impossible in classic LW. Some internal shifting happened (not to be blunt) and suddenly we got the VPR inside LW and at great speed too. This alone gives me hope that they are indeed doing a lot of work under the hood, waiting to be revealed when the time is right.

That didn't surprise me as much later on, mainly because there was a lot of work done to enable 3rd party rendering. ALSO, it is not uncommon these days for other applicatons to support full integration of multiple render engines so you wouldn't have to use the default shader or lighting system at all if you didn't care to. So creating a brand new progressive render engine does have its appeal but it's not unheard of at all to integrate such things these days. What I don't see a lot of are plugins that completely replace the internal animation AND mesh handling system. Seems to me this is a lot more difficult.

robertoortiz
07-26-2014, 09:57 AM
That didn't surprise me as much later on, mainly because there was a lot of work done to enable 3rd party rendering. ALSO, it is not uncommon these days for other applicatons to support full integration of multiple render engines so you wouldn't have to use the default shader or lighting system at all if you didn't care to. So creating a brand new progressive render engine does have its appeal but it's not unheard of at all to integrate such things these days. What I don't see a lot of are plugins that completely replace the internal animation AND mesh handling system. Seems to me this is a lot more difficult.

Ont thing to consider is the time it took the LW group to develop Chronosculpt 1.0 using the Hydra infrastructure...Only a few months. And that is quite impressive.



So where exactly does deep fear come into the picture here?

I guess a misplaced sense of loyalty... LOl
I honestly feel that the high end market is at a stage right now that it is cannibalizing itself. And frankly it worries me that I see to much abstraction in terms of the 3d Tools being developed when I feel strongly that more intuitive tools are the way to go.

jasonwestmas
07-26-2014, 10:01 AM
Ont thing to consider is the time it took the LW group to develop Chronosculpt 1.0 using the Hydra infrastructure...Only a few months. And that is quite impressive.

It is impressive, and who knows, maybe some of this new tech that was put into Lightwave (VPR) can be brought into Chronosculpt.