3DCoat needs a bug tracker

[Taros]

Hi.

Today I planned to make some selftraining in 3DC, because I am presenting the current version on the biggest softimage User Meeting in germany, tomorrow.

But unfortunately I realised that some of the really important tools don't work correctly. Then I've installed the main release (3.1) and the same problems happened.

Finally I must say, that the texture tools feel like a alpha stage. This is really not what I've expected. I am a little bit disappointed about this.

I know that 3DC have a lot of tools and is a complex program, but the texture part seems to be really stable for me. OK, now I see, that I am wrong.

I know, that the programmers can't see all bugs and test them. So I think it is time to ask for a serious beta team. I mean some people who get tasks and test only special parts of the program. Currently bugs can be found only "random", and this is the wrong way.

So I appeal for launching a beta team.

Be creative

Chris (Not happy)

[erklaerbar]

I see where you are coming from. Sounds like a good idea, however i cant really see how it could work in practice when considering the speed 3DC updates and changes. It would either be very expensive-if paid, or an almost impossible task for volounteers-due to time needed.

[polyxo]

Yes, I also think that a team of testers made sense.

Certainly: Andrew is very busy hunting down every bug which gets reported.

I still have to admit that I run into situations where I do not know where to start reporting...

I had to report a new problem every minute basically, I am not kidding!

These mostly are harmless quirks, things which still are irritating or ambiguous and brake down the workflow.

Often enough I know to work around and then I forget to report. I'm sure this happens to others too.

I wonder however if Beta Testers were enough...

I sometimes ask myself if it was possible to spread the programing load on more sholders.

So that Andrew could define the great picture and others refine the implementations of single features.

But a man of his capabilities will know best what is needed...

Holger

[Taros]

I don't think, this is not possible. There are a lot of examples that a beta test works well. Look on the games, there are really big beta phases.

I guess, here are a lot of enthusiasts who would like to test the tool serious. I don't wanna be payed, I just would like to see a working tool and the right decision, when and what kind of "idea" have to be implemented or which bug has priority. And if you would like to have some advantages, why don't talk about this?

There can be rules. Not everybody can be a beta tester, only a small group of people who have to test in periodical times.

You could work with mantis or other tools, that work online and offer user accounts.

I can't decide this, but I am shure Andrew is not averse to offer some "special deals" for such people.

However, this tool needs more than a chaotic forum...

Don't missunderstand me: New Ideas are good and bug reports from all the people out there too, but some "special people" should have more "weight".

Be creative

Chris

[kay_Eva]

I don't think, this is not possible. There are a lot of examples that a beta test works well. Look on the games, there are really big beta phases.

I guess, here are a lot of enthusiasts who would like to test the tool serius. I don't wanna be payed, I just would like to see a working tool and the right decision, when and what kind of "idea" have to be implemented or which bug has priority. And if you would like to have some advantages, why don't talk about this?

There can be rules. Not everybody can be a beta tester, only a small group of people who have to test in periodical times.

You could work with mantis or other tools, that work online and offer user accounts.

I can't decide this, but I am shure Andrew is not averse to offer some "special deals" for such people.

However, this tool needs more than a chaotic forum...

Don't missunderstand me: New Ideas are good and bugs from all the people out there too, but some "special people" should have more "weight".

I don't think it's fair to compare 3dCoat to game development as 3dCoat is much more of a 'living document' than games are. 3d Coat evolves according to the need as well as according to Andrew's genius - it's not simply a matter of releasing a final forum. I think 3dCoat will constantly change until the day it dies.

That said, it makes sense to give certain people testing responsibility but let's face it those professionals most qualified are probably too busy to do that kind of thing. Anyways I like the current approach of an open beta. I think it's a much more evolved process than a lot of other software betas I've seen. In fact, I often find myself referring smaller software devs to this forum in order to cite Andrew's current approach as a GREAT and wonderful way of doing things.

It feels so great that andrew even listens to an amateur/hobbyist/student such as myself - he even made permanent a suggestino or two that I made, whereas a lot of these other devs completely look over my head. Even when, well perhaps I'm biased in saying this, even when I'm right.

Anywho, that said. I think the Japanese users are the most open minded when it comes to software use and software workflows. I vote that Japanese professionals gain instantaneous Beta "special status". And that's not a joke.

[JamesE]

I am just making a habit of reporting everything I find - bugs, missing features, workflow enhancements. Seems to be working so far. I think it's amazing to see something I request or report addressed even in the next build! I'm using 3dcoat in production and in my spare time. I don't always report everything right away and sometimes miss stuff, but even then, dedicated beta testing is the same so I would see it as no different. I have to say I much prefer Andrew's open approach over any other beta testing experience I've been part of, especially since it allows anyone with the interest and drive to see this program improve to be involved. Also, I believe he does have a small beta testing team already.

I don't think things would be any different if he had more beta testers. His development approach really does depend on his users giving feedback. So give lots and give it often. =]

As for the texturing side of things, I think this part of the program has probably received the least amount of user focus since the release. It was also one of the last things to go in, so I think it should be somewhat expected that it will need improvement. But that can only happen as people use it. Even for me, I'm just ramping up with the texturing tools as most of my initial focus and interest was on the sculpting and retopo side. That is changing though now that I have projects in and approaching texturing stage.

[T.H. ROCK]

I like Andrew's current method of development.

He implements more rapidly than any developer I've come across and is more than responsive to changes that users notice.

As the old saying goes, if it's not broken - don't fix it.

[Andrew Shpagin]

Official beta team is very difficult thing for 3DC. In general we have 1 beta tester, but it is not effective at all.

Reasons:

- income is enough for development but not enough to maintain full-time beta team. It will just kill development.

- beta team can't see so much as users see. And really narrow problems will not be found by beta team.

Why you are trying beta versions? I think because it is interesting to be in process of growing and creation. You asking - and you are getting.

[PalSan]

I would like to express my opinion concerning developing and program testing.

I agree with Andrey, that the beta testing command cannot see, pick up and find all bugs. As regarding open testing it works much more effectively.

However I agree and with Taros that many bugs are casually. As repeat in other versions.

Meanwhile I would like to offer you such a concept of development of a package. For the certain period of time, the prior direction is chosen. For example, UV Tool is developed thoroughly. The developer's toolkit is improved. The work process is grinded. All bugs are caught as far as it is possible. There is a real "brain storm" of some active users of the program. Now there is a ready module.

Now we can move to the next direction. For example to the module rendering, etc. Thus, we receive a stable, strong and functional toolkit with the minimum quantity of bugs after each completed period. How to choose a prior direction? whether it is a public opinion poll or point professional users' point of view, it is the other subject. And I think, evrything can be solved easily.

There is a situation now, that there are a lot of utilities and powerful tools in the program, there are a lot of changes in all modules, which occur permanently and a lot of raw and insufficiently functional tools and a lot of bugs.

Andrew, quickly corrects them it is necessary to note, but I think that also it would facilitate process and to it.

thank you for attention)

Я хотел бы высказаться по теме разработки и тестирования программы.

Согласен с Андреем, что команда бета тестеров не может увидеть,выловить,найти все баги. и открытое тестирование в этом смысле работает гораздо эффективнее.

Однако,согласен и с Taros, что многие баги нахожятся случайно.и повторяются в следующих версиях.

Между тем хотел бы предложить вот такую концепцию развития пакета.

На определенный временной период выбирается приоритетное направление. Допустим UV Tool ,разрабатывается,разбирается "по косточкам". вылавливаются все баги, насколько это возможно. совершенствуется инструментарий. шлифуется рабочий процесс. устраивается настоящий "мозговой штурм" активными пользователями программы. Все есть готовый модуль.

После этого переход к следующему направлению. Например Render модулю и т.п. Таким образом, после каждого завершеного периода, мы получаем стабильный,сильный, функциональный инструментарий c минимальным кол-вом багов. Как выбирать приоритетное направление, опрос ли это, точка зрения профессиональных пользователей,это уже другой вопрос и тут ,я думаю, все можно решить достаточно просто.

На текущий момент же складывается ситуация ,что в программе огромное кол-во полезностей и мощных инстументов,изменения во всех модулях

постоянно происходят и много сыроватых и недостаточно функциональных инструментов.и много багов. Андрей, быстро правит их надо отметить,но думаю, что также это бы облегчило процесс и ему.

спасибо за внимание)

[Taros]

Hello.

I have the feeling, this discussion goes to the wrong direction and some people misunderstand my first posts. I am shure, this is a language problem. My english is not bad, but seems not good enough...

So I try it again in points:

1. I don't advocate for a payed beta team or something similar.

2. I don't want to end the current open beta, it is important.

Now I will try to give my proposal in detail again:

Andrew said, he have already a tester, and I understand the reason, why it is only one guy. That's OK.

But why don't do the following:

Your tester stops to test the tool by himself. Yes, I am not kidding. You install mantis bug tracker or something similar. Your tester starts to collect all bug reports from the forums and your inhouse bugs. He organise all the bugs in a bug tracking software, like a "lead tester" does. He reproduce posted bugs by the users and rewrite them, if necessary. You lead tester will be the one who have the overview over all bugs and ideas. He is the one, who sets the priorities in your bug tracking software, together with the coders. Your coders have a portal, where they can take open tasks and set completed tasks. Your tester will be the main guy, who communicate with the community and give feedback, if a bug report is affiliated into your list or if one is corrected.

I just please, that there is someone official who concentrate on collecting all the feedback into a software like mantis, to have a overview and do the right thing at the right time. Please correct me, if you already do that. Then I am quite. But then there must be other problems, why there are still problems with "finished tools", that are already implemented and officially not in developement.

And one more comment: I am working professional, and I have some time to give feedback. I am shure there are some more people who account their part.

I hope, you understand my wish now. I just wanna help you develop a better tool.

Be creative

Chris

[artman]

Your tester starts to collect all bug reports from the forums and your inhouse bugs. He organise all the bugs in a bug tracking software, like a "lead tester" does. He reproduce posted bugs by the users and rewrite them, if necessary. You lead tester will be the one who have the overview over all bugs and ideas. He is the one, who sets the priorities in your bug tracking software, together with the coders. Your coders have a portal, where they can take open tasks and set completed tasks. Your tester will be the main guy, who communicate with the community and give feedback, if a bug report is affiliated into your list or if one is corrected.

That part is very good I think

[Taros]

I would like to warm up this thread.

Please, is there a possibility to arrange a subwebsite with mantis for all users?

There could be an admin, who can organise the bug posts by priority.

I know, that averyone would like to post bugs, like here in the forums.

I know the people like the "nice" way to post bugs in threads, but till today the posts are not good organised and sometimes posted in the wrong area or with the half of information, that is needed. A bug tracking system could help to avoid this.

I bet for a bug tracking system, like the open source platforms, f.e. sourceforge.

What I really miss, is a feedback, if a bug is really fixed or not. I know, that a bug tracking system gives easier ways to organise the bugs and give feedback, than a forum. I understand, that Andrew do not always have the time to comment all bugs, but a short click on a "approved" or "fixed" button would help all a lot.

I know a lot of people love the "forum way", but believe me, if you have used a good bug tracker once, then you will never abandon it.

Best wishes

Chris

[Javis]

Hey Chris,

I completely agree. A bug tracker would be really helpful for everyone, users and programmers alike.

[philnolan3d]

That Mantis looks nice. FogBugz that NewTek uses for LightWave is also very good.

http://www.fogcreek.com/fogbugz/

[Taros]

That Mantis looks nice. FogBugz that NewTek uses for LightWave is also very good.

http://www.fogcreek.com/fogbugz/

Mantis is an open source project...