Patch reviewing, the long waits involved now and the impact on development.

Not only that, but have you noticed that very few additional people seem to get commit rights these days, imagine how much more Blender could have in areas like BI and the BGE if people like HG1 had commit rights in response to the number of patches they made that showed a competent skill in coding?

It seems to me that it is reasonable to conclude that helping people get to where they have direct access to Master is of a very low priority as well, it almost seems like volunteer driven development takes a permanent back seat to the Open Movies, where people voted with their wallets that there are other things they want to see when the funding for a full length movie fell short.

Does Ton himself see this as a possibility that this is why he doesn’t seem to get enough volunteer help at times? It is indeed possible that perhaps the only way to convince him on that would be for all on the cloud to vote with their wallets and drop their subscriptions, he would have no choice but to end his dreams of movie making for the time being and focus on things requested by the users such as helping aspiring volunteers. The message needs to be clear then that he will struggle with all future fundraising for a movie project unless he gives a higher priority to more neglected areas of development.

Who knows, we could see another acceleration of development of Blender above the current pace and allow for a new movie project of a caliber that he never even thought would be possible so soon.

Did HG1 ask for commit access? I cannot remember it, therefore also no rejection from our side.

What i dont understand is, why the Bisect modifier of kagegeo doesnt get in trunk. It is really needed!

I always thought this was automatic after a few successful patch review sessions, at least I don’t recall reading about a bunch of people asking for it.

Is that not the case then? If so, is there any actual indication on blender.org or developer.blender.org that lets a developer know about a good time to ask this or that it can be asked at all?

This isn’t quite right, it wasn’t some ‘accident’ - we just moved to a new bug tracker (phabricator) which doesn’t have the ability to export the data needed for the credits generator to reference the original authors (or there may be some way, but it would involve writing PHP).

Couldn’t it be done manually then (because throwing your arms in the air and deciding there’s no solution is definitely going to rub a lot of people the wrong way)?

I didn’t say it was caused by an “accident”, but that it happened “accidentally”, as in not on purpose, not planned, not malicious, not intentional.

Sure, but then someone has to do it manually, At that point we better get someone who knows PHP to solve.

Or just not have a credits page, we didn’t have one until recently and development went ok.

Might as well have ourselves sectioned now, to face the inevitable consequence of doing so :wink:

Basically a few thousand commits by 253 people have been accidentally attributed to the person who committed the patch instead of the author. So it would take a very long time to grep the logs for each of these authors and then fix each commit.

I wonder if they could just bring up GForce again (the old tracker) temporarily with the old data, run the old script to get the CSV file and use that to fix the entries.

I think this isn’t the problem really, we are very relaxed about handing out commit access to the addons repository, yet it doesn’t mean we automatic get a lot of addons development.

@Harley:

Your reset to defaults patch was working fine and they rejected it?

When I first started, that was one of the big things that annoyed me the most! I would mess around with a setting and then want to go back to the default, because I don’t know anything about it and I don’t want to mess anything up. If there was a working feture like that it would make it much easier for new users (I think).

It worked wonderfully. I don’t really mind that things don’t get accepted. It did sting to do so much work (185k patch!) and then be told that Ton suggested simply removing the “Reset to Defaults” item instead. But the real kicker is that the item was not removed and the “to do” item remains to this day. Someone else started working on it again in January of this year before they found out they were wasting their time.

It makes a person very wary of doing anything at all. It was asked for, reviewed, and still not accepted. I’m not likely to repeat that process…

Yeah, I don’t understand why you would get rid of that though, If it functions properly and doesn’t brake anything, why reject it?

I’m interested in getting into development but if it’s like that I’m not so sure. It doesn’t help that there isn’t much out there on learning how to get started (besides python), and the Wiki doesn’t seem to have up to date info on building from source.

I wish that building blender was like using quake army knife…

I am also asking for this. Really really, I am asking for this! ideasman42, can you implement it? :wink:

The question here is: which default value do we take?

As far as I remember, Harleys patch put all values into DNA code.
We actually have 3 kind of “defaults” though.

  1. Default values inside the code (constructors and other places).
  2. Default values in the inbuilt startup.blend
  3. Default Values in the user-saved startup.blend

So what is the correct value? What should happen when the users presses the Reset button? Should it reset to the value in the user-saved startup.blend or the inbuilt one, or should it take another way?

With all due respect though, that;'s a question you ask when someone has already got a 185k patch on the table and willing to work on it. It is not something you raise after the patch is requested, developed, then dropped on the whim on the project leader… then left in the TODO list for over a year.

The point I think being made here isn’t about the patch in particular but about how the patches are handled (or not) by the Blender Foundation. I don’t have anything along the lines of Harley’s monster effort, but I & others have watched patches requested, discussed, worked on post-feedback… then dropped on whim. It’s not an isolated incident and it clearly has a negative effect on people contributing to Blender’s development.

Quoted for agreement, +1 and all that. Countless times I had to just type in a guesstimate value to get started or in some worst case scenarios I had to delete the modifier/particle system/etc and then add anew, and try to remember what I did. Slows it down.

Unfortunately I have to agree with you.
On the other hand I believe it wouldn’t make sense to define a complete process about how patches need to be submitted, because it is unlikely that there are many enough of them, even if the whole thing worked better.
Two years ago when the patch was submitted, Phabricator wasn’t in place for Blender yet, if I remember correctly. Now it is there, and I believe there is some potential in it regarding contributions. Someone who is willing to create a patch, may e.g. get in touch with the developers on irc first before a task is created in Phabricator. Everyone could keep track of what is going on, be it the UI team, developers or anyone else and bring up relevant information, like “there are three kinds of defaults”. That kind of issue could easily be avoided and someone who is willing to contribute could be supported by pointing out the relevant aspects.