What's wrong with phpBB development

Discuss general development subjects that are not specific to a particular version like the versioning control system we use or other infrastructure.
bolverk
I've been banned
Posts: 280
Joined: Mon Feb 02, 2009 5:39 pm

What's wrong with phpBB development

Post by bolverk »

naderman wrote:RFCs are not meant to be used to collect data on the opinions of all phpBB users.
But you use them to justify your decisions anyway by generalizing that small sample's response and applying it to the whole user base. RFC's should not be used as opinion polls or general surveys, they should be used to initiate and drive technical discussion of a development change proposal and they should be based on a pre-designed template that requires initiators to base them on sound technical reasons (facts) rather than subjective opinions. Each RFC should include from the submitter:

Proposal detail
Pros - using data to support why the proposal should be implemented
Cons - due diligence to identify the potential negative consequences of the proposal
Scope -Impact, what and who would be impacted by the proposal again supported by quantifiable data
naderman wrote:Statistical data can be helpful, but often it leaves questions unanswered.
Statistics cannot answer everything but it is more reliable to base decisions on actual data than subjective opinions.
naderman wrote:We should think about what the consequences of our decisions are for all those people who do not participate in this process but in the end decisions here are made based on subjective reasoning and the interests of whoever is here to raise their opinion. This process is flawed but we will have to live with that.
That is what is so upsetting to me personally, you do not have to live with it, you are choosing to. You have so much untapped data available to you that you refuse to avail yourselves of, instead you choose to rely on subjectivity to drive the development process. Any one of you could have found the same data that nn- and I did, all you had to do was look. But you didn't. I really believe it never occurred to anyone to try and assess the impact which infers you did not think there was any/it was negligible, or that you didn't care what the impact was.

I will again try to drive home the most egregious error in any software development project that I have ever witnessed (by quoting myself):
me wrote:The reason that this decision was not sound is because you based the decision on boards that would not be affected by the proposal instead of boards that would be. You gave the voice that should have the least weight in such a decision the most power. :? You need to use the other data points to assess the true impact of the decision,
:shock: :shock:
Analogy: You posed a question regarding the continued existence of product A (SS) to an available group of people (available at that time and not a representative group of people) and the majority of responses were "drop it" however the majority of these respondents also admitted to not using product A themselves (for various subjective reasons) thus widely skewing the perceived impact as negligible. If you want assess the impact of anything you have to start by querying the people who would be impacted. It is safe to infer that anyone who does not like or use SS today would not object to it's removal in the first place.

User avatar
naderman
Consultant
Posts: 1727
Joined: Sun Jan 11, 2004 2:11 am
Location: Berlin, Germany
Contact:

Re: What's wrong with phpBB development

Post by naderman »

I'll try and say this again. phpBB is driven by what the people working on it care about. We choose to run phpBB that way. While we are interested in statistical data, we are not going to base our decisions of what is important or what needs to be done on statistical data.

bolverk
I've been banned
Posts: 280
Joined: Mon Feb 02, 2009 5:39 pm

Re: What's wrong with phpBB development

Post by bolverk »

An FYI to those who may be as confused as I was when I saw this topic...I did not create it. The title and location were not my choosing, it was apparently split off my reply to naderman's reply to me in the subsilver topic.
naderman wrote:I'll try and say this again. phpBB is driven by what the people working on it care about. We choose to run phpBB that way. While we are interested in statistical data, we are not going to base our decisions of what is important or what needs to be done on statistical data.
Well that was certainly crystal clear. I finally got the answer to the question I asked Acydburn last year. To summarize: phpBB is and will continue to be what the minute number of contributors envision it to be based on what they care about. Using metrics to identify areas where the software could and should be improved from the data available in the support forums and bug tracker, not in this lifetime. Got it. I never suspected that the most widely used OSS bulletin board script 6-8 years ago would end up here. I feel like I should share the emotional response reading your statement above elicited from me just now. I am actually a bit shaky, adrenaline rush from the shock I suspect. I feel like I just watched a suicide bomber blow themselves up. Thank you for the honesty as to where the users and their experience with the software really rank.

Oleg
Posts: 1150
Joined: Tue Feb 23, 2010 2:38 am
Contact:

Re: What's wrong with phpBB development

Post by Oleg »

naderman wrote:I'll try and say this again. phpBB is driven by what the people working on it care about. We choose to run phpBB that way. While we are interested in statistical data, we are not going to base our decisions of what is important or what needs to be done on statistical data.
Thank you for your honest response. Every now and again some of us get delusional and think that there is a community here based on mutual understanding and respect, but the record gets set straight eventually: the green people run the show, everyone else is allowed to stay only as long as they don't get in the way.
bolverk wrote: To summarize: phpBB is and will continue to be what the minute number of contributors envision it to be based on what they care about.
As a contributor, I think a more accurate statement would be this:
To summarize: phpBB is and will continue to be what the minute number of phpbb developers envision it to be based on what they care about.
I'm still waiting for phpbb developers to review patches I submitted in early March.
Using metrics to identify areas where the software could and should be improved from the data available in the support forums and bug tracker, not in this lifetime.
Wiki replacement comes to mind as another data point.
I never suspected that the most widely used OSS bulletin board script 6-8 years ago would end up here.
Actually most forum software is developed this way. phpbb (my understanding) was in the same camp until it recently started claiming otherwise, or at least did a good job convincing people of such.
I feel like I should share the emotional response reading your statement above elicited from me just now. I am actually a bit shaky, adrenaline rush from the shock I suspect. I feel like I just watched a suicide bomber blow themselves up. Thank you for the honesty as to where the users and their experience with the software really rank.
Of course, after being convinced of "open development" and "community involvement" the truth hits home that much harder.

Edit: fixed the month in which I submitted the patches.
Last edited by Oleg on Wed Jul 28, 2010 7:12 pm, edited 1 time in total.

Marshalrusty
Project Manager
Project Manager
Posts: 273
Joined: Thu Oct 27, 2005 1:45 am

Re: What's wrong with phpBB development

Post by Marshalrusty »

This topic comes at an interesting time because several managers and development team members just had a conversation about this yesterday. I generally agree with the points made here, and would rather not. The good news is that the development team sees the problems as well, and we are all going to start taking a critical look at the way we want decisions to be made in the future.

On the other hand, please don't sensationalise or exaggerate the problems, as that only serves to create argument (rather than constructive change).
bolverk wrote:Analogy: You posed a question regarding the continued existence of product A (SS) to an available group of people (available at that time and not a representative group of people) and the majority of responses were "drop it" however the majority of these respondents also admitted to not using product A themselves (for various subjective reasons) thus widely skewing the perceived impact as negligible. If you want assess the impact of anything you have to start by querying the people who would be impacted. It is safe to infer that anyone who does not like or use SS today would not object to it's removal in the first place.
I particularly agree with this. Not just for subsilver2 (as it is an outdated style and I'm not necessarily convinced that it should continue to be supported), but in general about the ways RFCs are handled.

User avatar
naderman
Consultant
Posts: 1727
Joined: Sun Jan 11, 2004 2:11 am
Location: Berlin, Germany
Contact:

Re: What's wrong with phpBB development

Post by naderman »

You are putting words in my mouth. What I explained is that the code in this project is written by volunteers. Now such volunteers don't typically like being forced to do something. So if you want to keep them working on your project, you have to give them the freedom to pick what they want to work on. Now most of these volunteers working on phpBB have an interest in improving the software for the users. So if you can convince anyone able to write code (be it a developer or someone else who can post a patch) that the idea you are presenting is a good one, you have a good chance of it happening. If you have a great idea but nobody is interested in implementing it, it just won't happen, even I can't change that. So statistics might help you convincing someone to implement something but they might also simply not help.

Now with the RFC process there are really two possibilities. Either the RFC is posted by someone who intends to implement the feature him/herself and the topic is used to get some feedback on whether the idea makes sense, wether some pieces of information were not considered in the proposal or whether some detail needs to be adjusted. So it's not a vote where the number of yes/no voices counts, but the result should be consensus amongst everyone who was interested in the topic that the correct way to go forward was found, even if that means not to implement the proposal. The other possibility is that the RFC was posted by someone who is not capable of implementing the feature. In this case the person has to present the idea in such a way that someone capable of implementing it volunteers to do so. Apart from that the topic is used in the same way as in the other case.

User avatar
naderman
Consultant
Posts: 1727
Joined: Sun Jan 11, 2004 2:11 am
Location: Berlin, Germany
Contact:

Re: What's wrong with phpBB development

Post by naderman »

Marshalrusty wrote:
bolverk wrote:Analogy: You posed a question regarding the continued existence of product A (SS) to an available group of people (available at that time and not a representative group of people) and the majority of responses were "drop it" however the majority of these respondents also admitted to not using product A themselves (for various subjective reasons) thus widely skewing the perceived impact as negligible. If you want assess the impact of anything you have to start by querying the people who would be impacted. It is safe to infer that anyone who does not like or use SS today would not object to it's removal in the first place.
I particularly agree with this. Not just for subsilver2 (as it is an outdated style and I'm not necessarily convinced that it should continue to be supported), but in general about the ways RFCs are handled.
Now if you look at the subsilver2 topic you will find that the arguments made in the topic after the decision are exactly identical to what was already brought up before the decision. The people contributing in the RFC process might not be using subsilver2, but they do not simply state whatever is in there interest, but they try to post concerns other phpBB users might have as well. Of course that process is not perfect, but at no point was any deliberate attempt might to hinder anyone from voicing their arguments in the topic. The decision was made based on those arguments, not based on how many people were in favour or against, or how many people use prosilver or subsilver2.

Oleg
Posts: 1150
Joined: Tue Feb 23, 2010 2:38 am
Contact:

Re: What's wrong with phpBB development

Post by Oleg »

Implementing features is one thing, removing features that people use is another. Read any sane backwards compatibility policy - feature removal is handled much more delicately than feature addition.

Typically people in leadership positions are expected to consider more than their own interests. For example, if I'm a mere contributor I can propose a feature that would benefit me at the expense of everyone else. Being only a contributor I can't force it onto the world though - it must be approved by the development team. If many people object to my feature, one would expect it to be rejected by the development team and not included in the product.

If now I am a developer, I can commit whatever feature I want. The normal expectation is that I have an understanding of how any proposed feature would affect the entire userbase and will not do selfish things for my personal benefit at the expense of users. Subsilver removal offers potential savings of effort, which may or may not matter, for phpbb developers at the expense of definite suffering of a number of users. Hopefully I don't need to explain the absurdity of statements like "if users really want subsilver, they are free to become developers and maintain it as a modification".
Of course that process is not perfect, but at no point was any deliberate attempt might to hinder anyone from voicing their arguments in the topic.
This reminds me of a conversation that I sadly can't find the link to right now. It goes something like this:

- Why haven't you told me about X?
- We have given the notice to everyone.
- The notice that was under a stack of papers in the bottom drawer of a locked filing cabinet in an unlit locked room in the basement filled with antique equipment and marked 'Sewer Access' on the door?
- Like I said, anyone could have read it.

If you care about users, you tell users about important changes in a forum users read. Even a forum topic would not be seen by all users. For something like subsilver removal, an announcement should be made. However, we established already that it's a bad decision so advertising it better is a moot point - it should simply be reverted.

User avatar
Ger
Registered User
Posts: 293
Joined: Mon Jul 26, 2010 1:55 pm
Location: 192.168.1.100
Contact:

Re: What's wrong with phpBB development

Post by Ger »

nn- wrote:However, we established already that it's a bad decision so advertising it better is a moot point - it should simply be reverted.
You have established it. Not everyone. ;)
nn- wrote:This reminds me of a conversation that I sadly can't find the link to right now. It goes something like this:

- Why haven't you told me about X?
- We have given the notice to everyone.
- The notice that was under a stack of papers in the bottom drawer of a locked filing cabinet in an unlit locked room in the basement filled with antique equipment and marked 'Sewer Access' on the door?
- Like I said, anyone could have read it.

If you care about users, you tell users about important changes in a forum users read. Even a forum topic would not be seen by all users. For something like subsilver removal, an announcement should be made.
I don't think that's fair. It's mentioned on the blog, explained after a question in the comments and spoken of here on Area51. Of course, not anyone would have read it and an annoucement would have been better, but your anology is totally unfair.

And hey, what about all the discussion here? The team is considering all the flaws in Prosilver and is (if I understand it correctly) reconsidering the options. So they do listen to the community. But listening doesn't mean resetting the course. Many people don't care about subSilver2 or even hate to work with it. Of course you don't hear them so much as the people who are afraid of losing it since they have nothing to loose and not much to gain.


My opinion* on all this: let's lay it to rest and let the dev's go back to developing in stead of hours and hours of discussion.


*) That's strictly my opinion, don't let it be mistaken for the opinion of others.
Above message may contain errors in grammar, spelling or wrongly chosen words. This is because I'm not a native speaker. My apologies in advance.

Oleg
Posts: 1150
Joined: Tue Feb 23, 2010 2:38 am
Contact:

Re: What's wrong with phpBB development

Post by Oleg »

Ger wrote: I don't think that's fair. It's mentioned on the blog, explained after a question in the comments and spoken of here on Area51. Of course, not anyone would have read it and an annoucement would have been better, but your anology is totally unfair.
I'm talking about notifying users before a decision affecting them is made and giving them a real opportunity to voice their opinions. All announcements you're referring to came after the matter had been "settled" and the code physically removed from trunk.

Post Reply