Thoughts On Preventing a 2013 Release of phpBB 3.2

[webmacster87]

Well, at long last, phpBB3 has been released, exactly 2079 days after the 4/4/02 release of phpBB2. We're all definitely excited that phpBB3 is finally out now (even if some of us still have an appreciative spot in our heart for phpBB2), but I'm sure some people will be asking why it took so long and how we can make sure that we don't have to wait until August 22, 2013 for the next major release of phpBB. (8/22/13 is exactly 2079 days after phpBB3's 12/13/07 release date.)

Now, the phpBB teams have always had a practice of not sharing a release date with the public because they didn't want to have something come up, resulting in public disappointment. However, when development started in early 2003, the developers weren't developing with a phpBB3 in mind; they were developing with phpBB 2.2 in mind, which they had hoped to release sometime in 2004. (After all, phpBB2 had only taken 16 months to develop, that probably seemed very reasonable.)

What happened was the development team got a bit over-excited and ended up completely rewriting phpBB. After all, a minor new version such as 2.2 is intended to indicate a release that would evolve off of 2.0, not completely rewrite it. It was after the teams became a bit self-conscious that they posted the momentous announcement that phpBB 2.2 had become phpBB3 "Olympus". In the almost three years since then, even more significant portions of phpBB3 have been rewritten, as it took a full 16 months to go from the "2.2 becoming 3.0" announcement until the first beta was published in May 2006, and then just about 19 months from the first beta in May 2006 through the final release in December 2007.

I mean, I did a week long series in late October making fun of Apple for taking two and a half years to release an update to Mac OS X. Quite frankly, it's easier to tackle two and a half years of world events than five and a half years.

Now obviously the development team is taking a nice breather, celebrating that they've finally gotten together something to declare "final", and will be going through the initial stream of bugs in preparation for the inevitable 3.0.1 release. But when they get started, we're definitely all hoping that they will be able to hold themselves back and limit 3.2 to an evolutionary upgrade, not a revolutionary upgrade, and be able to have it released soon. In fact, I would say that I think that 3.2 ought to be released before the end of the year 2008.

While I was thinking about all of this, I got to thinking about WordPress, another popular open source program. They actually have a fairly interesting release schedule. They try to have a new release out every four months (ignoring minor bugfix releases), and each new release contains a handful of new features, rather than the whole kitchen sink. Their roadmap is also wide open, and anyone can go and find out when their planned release dates are. Obviously, they indicate that the dates are subject to change, but what's important is that they have a self-imposed target. In their development process, they allow themselves the first three months in between releases to develop code, at which point they freeze the feature list for their next release and spend the next month with testing, internal betas, release candidates, etc. in preparation for the final release. It seems to work fairly well for them, and through that system they've had three releases (2.1, 2.2, 2.3) during the year 2007, and have 2.4 scheduled to be released at the end of January.

While I'm not necessarily saying that we need to have new phpBB releases every four months, which, granted, may be a bit frequent (especially to us phpBB users which aren't used to very frequent releases; heck, we haven't had a phpBB2 release in a full year!), I am suggesting that perhaps the teams should be a bit more open. I can understand the rationale for wanting to make a release date of "it'll be released when it's ready," but then who's to say when it's ready? Quite frankly, you can keep going with development forever; it just takes the will to decide when to stop developing and release something. What I think the development team should do is come up with a strategy, similar to how WordPress did, to make sure they don't go and repeat what happened to phpBB 2.2. I think it would be good for them to publish a "target window" as a projected release date, and then try their best to meet that target. Obviously, they could state that the target windows were only possible goals and completely subject to change, but I think that if the community at large could have a general idea of what to expect and the teams could have some idea of knowing when they should stop coding and start releasing, we would be able to see a more frequent release schedule for phpBB and not see phpBB3's codebase age the way that phpBB2's has.

Anyway, that's just my own thoughts and opinions. What do you think about that idea?

[John Hjorth]

Interesting thoughts, Douglas. However I think it´s quite diffcult to make any assessments on that. For me the problem is, that I can't find out which parts of the above above is based on inside information that you have.

- Perhaps we should start to make a call for the "story of phpBB 3.0" that has been promised by dhn on area51, in a way that does not create any kind of *pressure*, and thereby not *bugging* the developers?

[webmacster87]

I don't have ANY inside information at all, because I'm not a team member. What I wrote above is just my own opinions on what phpBB should do in the future and is not based on anything at all.

[John Hjorth]

Well, Douglas,

Thank you for clearing that up.

It was not my intent to insult neither you nor the contents as such of your very well thought out post. The *problem* for me with your post was, that everybody who knows just a little about the history of phpBB, knows that you have been a MOD Team Member

Actually, some parts of your post contain information, that is new for me, and I have read about everything to find *right places* about phpBB 3.0 within the last 1½ years - worthy of reading or not.

- I apologize, if I by my post have offended you, that was certainly not my intent.

What do you think about my suggestion posted above - should we make a move on that?

[webmacster87]

No, I didn't take it as offense, just wanted to let you know that I'm not posting that with any insider information that isn't already public, and I don't have any other insider info in mind.

[Techie-Micheal]

I don't know what the plans are for 3.2, so I couldn't give you any information even if I wanted to.

However, you should know that development of 3.0 basically halted for about a year due to issue after issue after issue. We didn't want that to happen, but it did. Then there were other, smaller fires, slowing down development even more.

That all said, what I do know is that the devs have made drastic changes to the process to speed things up. I know you've heard it before, but I can tell you with an absolute certainty that that is the case.

[John Hjorth]

However, you should know that development of 3.0 basically halted for about a year due to issue after issue after issue. We didn't want that to happen, but it did. Then there were other, smaller fires, slowing down development even more.

That year with all those *damages* was 2005, right?

[Techie-Micheal]

However, you should know that development of 3.0 basically halted for about a year due to issue after issue after issue. We didn't want that to happen, but it did. Then there were other, smaller fires, slowing down development even more.

That year with all those *damages* was 2005, right?

Sounds about right.

[Waleed]

Nice post, and yes I agree that it took a long time and yes it shouldn't have. But i think the dev team will learn from experience and not get carried away this time! I think one of the reasons why 2.2 was transformed into 3.0 was that the developers (and users) had so many ideas and features they wanted to see (and use)...

I know development of Olympus has caused some (if not a lot of) friction between some users and the dev-team, its a sensitive issue, with posts like "WHEN WILL IT BE DONE" and replies of "It'll be done when it's done," I recall the atmosphere of area51 getting somewhat tense at times.

I respect the phpBB Team for all the efforts that go into maintaining a project like this -- I'm sure everybody here (and at phpbb.com as well) feel the same way. But yes, I do agree with Doug about the benefits of a schedule for development. I'm sure phpbb users will feel even better if the dev team decides to share a rough outline of that schedule with us too.

But in the end its the team who have to do the developing and whatever makes them comfortable, I can take.
That isn't meant to be a suck-up, but a late release of phpbb hasn't exactly been a life-or-death situation for me. SOme people may feel strongly about this, different to my view. Just what I think.

[Acyd Burn]

To put it straight - giving a release date/time frame to the users and having one internally are two completely different things. Without having time frames in mind you are not able to develop and to reach the goal you set.

think it would be good for them to publish a "target window" as a projected release date, and then try their best to meet that target.

I do not think this will happen - you are not able to predict how much time each developer is able to devote to the project. It could be that one developer is able to put 10 hours into the project per week, the other one occassionally for 6 months only 3 hours per month, then being able to code the whole day, then not being able to code for 3 months. Such things are unpredictable, and since they are the main factor for calculating a real time frame we *always* miss our internal ones - still they are good to get an idea of when something needs to be ready.