suggestion to developers: please release beta-6

[Gumfuzi]

@code reader:
dont wanted to offend you, this were only my thoughts.

I think I got your thoughts (now), so let's see if there will be B6 or RC1 - for me, it's also no prob if there will be a B6. but I like surprisings, so I'll wait and see

[Jhong]

I want another beta release cycle.

I want to use the .MODS tab that hasn't been unhidden yet. I want to get practice using the automatic install and uninstall member functions of mod classes. I want to see if there are any bugs in the way these are implemented and prepare myself as necessary. An RC is too late a stage for this.

I want phpBB to improve the phpbb API documentation during this time, so I don't have to go to 5 different sources depending on what it is I want to do - use the auth system, create an ACP page, add a permission... etc etc.

These are all very selfish reasons (yes, the sentences all begin with 'I want'). I also want phpBB to be the best there is... it's taken this long to produce, another Beta cycle won't hurt anything. There ARE still bugs that need to be addressed.

[Lastof]

I want to get practice using the automatic install and uninstall member functions of mod classes. I want to see if there are any bugs in the way these are implemented and prepare myself as necessary.

You realise that isn't what the .MODS tab it about, right? (well, obviously not).

It is there to give mod authors a place to add their ACP panels. The tab is active, it just isn't shown by default since there are no modules under it.

[jumborex]

[...] what is the best/fastest/most efficient way to get the product to the desired "rc readiness" state?
one way (the one the developers seem to be taking now) can be to wait and solve bug reports produced mostly by users installing the cvs version, until "rc readiness" is achieved. [...]

I see in this claim from code reader some truth: I have seen how the new style gave to me some trouble in my test installation. So, the only way to make a good test is to install CVS for all people participating to the test project.
But, when you find something not working properly, immediately someone writes to you "CVS has no support!". Ok: so it's needed a Beta 6! I would like have some support for things apparently not working in beta 5 CVS!

[NeilUK]

Can't see that we need a Beta 6. Lets just leave the team to their work and wait for RC1.

[jojobarjo32]

I want to get practice using the automatic install and uninstall member functions of mod classes. I want to see if there are any bugs in the way these are implemented and prepare myself as necessary.

Automatic install and uninstall are not implemented yet and won't be before 3.2 IIRC. install() and uninstall() functions do nothing at the moment.
For the .MODS tab you talked about, as Lastof said previously, it actually "works" (well, it's just a category of modules...). It's only hidden because there are no mods using it. But if you want, you can already insert modules within it

Finally I think another beta cycle would be a good thing too (previous posters have said quite well why...).

[Handyman]

[...] what is the best/fastest/most efficient way to get the product to the desired "rc readiness" state?
one way (the one the developers seem to be taking now) can be to wait and solve bug reports produced mostly by users installing the cvs version, until "rc readiness" is achieved. [...]

I see in this claim from code reader some truth: I have seen how the new style gave to me some trouble in my test installation. So, the only way to make a good test is to install CVS for all people participating to the test project.
But, when you find something not working properly, immediately someone writes to you "CVS has no support!". Ok: so it's needed a Beta 6! I would like have some support for things apparently not working in beta 5 CVS!

The support factor doesn't really make a difference since Beta 5 only offers very limited support… such as trying to get it installed… but no support for running it live.
So even a Beta 6 wouldn't help if you're looking for support.
All good testers will download the CVS and see if the issues exist in the CVS before reporting it… so as far as I know, it hasn't been an issue so far.

[Nicholas the Italian]

All good testers will download the CVS and see if the issues exist in the CVS before reporting it… so as far as I know, it hasn't been an issue so far.

Rephrasing: "Both good testers..."

[code reader]

All good testers will download the CVS and see if the issues exist in the CVS before reporting it… so as far as I know, it hasn't been an issue so far.

if this is really so, then you have to ask yourself: what is the point of having beta releases in the first place? just open the bug-tracker, and continuously report and fix bugs that relate to the bleeding edge cvs, without any beta release.
i don't believe this is so.
my (entirely subjective) observation is that with every beta-release so far, there was a significant upsurge in reported bugs that lasted 2-3 weeks, and after this point bug-reports about the beta started to diminish, to the point that the 2nd half of every beta-cycle had mostly trickle of reports pertaining to the cvs version*. this makes the "bug is less" crowd happy, but, imo, this is because they can't comprehend the difference between bugs and bug-reports. (it is easy to extinguish all "bugs" this way: avoid testing!)
if you look at the quantity and quality of the reports, i believe you'll find this first month of each new beta cycle much richer than the 2nd (or 3rd) month.
it is my opinion that almost all beta releases were too late.
in the 9+ months since the first beta, if we would have 9 beta versions rather than 5, i think the testing process would have profit.


* with some exceptions: beta-2 took full 3 months, with the last month almost completely idle as far as testing activity is concerned, while beta-3 was only 3 weeks, which is a tad shorter than the optimal length imo.

[dhn]

edit: Nevermind me posted something that wasn't true