when will the 2.0.x -> 3.0 converter become available?

[code reader]

I will also note that I do not plan to ship the upgrade code for 2.0.x with Beta 1 as it stands at the minute, so those looking to actually migrate from 2.0.x to 3.0.x will have to wait a bit longer than that to do so.

imho, this will be a serious mistake.
first, it will mean that the upgrade script will not be tested as part of the beta.
second, it means practically limiting the beta to new installation. this means that any issue that will surface only with large forums will not be exposed by the beta.
third, it will prevent the mod creators (for mods that modify the db) from starting to work on the conversion of the data belonging to their mods


and lastly, i cant see any good reason for this. everyone will still know it is beta. why should you care to limit our ability to test the code? clearly making the converter available will enable some kinds of testing that will be impossible without it.

(ps: if there will be more than one beta, replace "beta" above with "beta 1". the logic still stands.)

please reconsider.
thanks.

[Yawnster]

There is RC stages for a reason.. as per the developers IIRC they will release a converter around RC1/2, so you still have a few cycles to iron out those bumps.. the BETA stages are for fixing the main primary bugs that will affect everything, and as it goes along of course more will be found, but they should be less serious and thus its safer to let boards migrate..

Yawnster

[Lastof]

The way I see it, it's to prevent people upgrading straight away when beta one is released, without sufficient backups, and then expecting the devs/support team to help if it all goes wrong.

[NNO-Stephen]

yeah, not having a beta upgrade script is just fine since they probably won't have any "upgrade" pack from beta to 3.0 so it makes perfect sense that they wouldn't release these upgrade tools till the release candidate stage for precisely the reasons just stated by the previous posts.

[code reader]

actually, there were no "reasons" cited by previous posts, only claims of the type "there will be no converter to prevent people from converting", "this is how it is going to be done", or "there is an rc stage for a reason".
none of these "arguments" actually argues any of my reasoning, and none explain anything.

as to your statement that "they probably won't have any "upgrade" pack from beta to 3.0":
it stands to reason that if there will be any structural change to the db among the beta(s) or from the last beta to the first rc, they will be minimal.
i will be very suprised if there will be no "upgrade path" between the betas, or from the beta to the rc. i dont know where do you take this information, but i never read such a claim from any of the developers.

imho, it will not be wise to expect the people participating in the beta(s) to start their board from scratch every time.
and, as i stated in my first post, i dont think it will be a good decision to force the beta participants to start from scratch even once. hence i think it would be best to release the converter script AT THE LATEST with the first beta.
personally, i dont even see the reason not to release it now.

[NeoThermic]

I'm sure Graham will post more on this, but I'll touch in here in the mean time.

actually, there were no "reasons" cited by previous posts, only claims of the type "there will be no converter to prevent people from converting", "this is how it is going to be done", or "there is an rc stage for a reason".
none of these "arguments" actually argues any of my reasoning, and none explain anything.

Beta is beta for a reason. You don't want to convert a whole large live forum over to a beta because if it breaks, you're rather stuffed. Also, you mentioned that you don't believe about there not being an upgrade path from the beta's to the final thing. It might happen, developers can not forsee major structural changes in the code. A beta might find that a certian aspect of the way the code is done is incorrect for normal forum operation, requirng it to be overhauled to become what is required.

and, as i stated in my first post, i dont think it will be a good decision to force the beta participants to start from scratch even once. hence i think it would be best to release the converter script AT THE LATEST with the first beta.

If you're downloading a beta of any software you'll have to accept that it is not fit for use. That might shock you, but beta isn't a stage at which you should be looking to test software if you're not up to things like starting from scratch to test. The whole point of a beta is to find major problems, critical bugs. Generally speaking (I'm drawing from my knowledge of design practises in games design) beta will contain some major functional issues which need resolving, but that can not be found, and should not be found by taking data in which might confuse the real location of a bug.

If you convert a forum over and suddenly quotes don't work in posts, is that a fault of the beta or the data you've converted? How will you know? You would have to create another test board, from scratch I might add, to prove or disprove that it has to do with your data. However, if you started from scratch in the first place and quotes did not work, you would immediatly know that the issue lies with the beta, and not your data. Thats just an example of why conversions into beta should not be provided on the grounds of bug tracking.

personally, i dont even see the reason not to release it now.

Now, I'll give you a good one. We get enough people asking for CVS support here. If we provide a converter, people will ask for support for that as well, and then you'll get all manner of people trying to use the conveter for various reasons, of which you're bound to have a few who don't grasp the concept of no support.

It has actually been stated that the converter won't be provided with the first beta before, such as here, and somewhat here (doesn't stipulate that it will come with beta).

Hmm, I wrote more than I expected


NeoThermic

[code reader]

you all (specifically neo) seem to be making the same (false) assumption:
if i upgrade my board from v2 to v3, it doesnt necessary means that i shut down my v2 board and start operating (with all my users) on the v3 board.
all it means is that when i begin my beta testing, my TEST board will be populated.
of course i will keep operating my "live" board on the v2. why shouldn't i?
but i may invite selected few of my users to go to the v3 board, and start frolic around.
this will make a much more effective beta than to open an empty board from scratch.
and if someone will actually switch completely to olympus with a large live board still in the beta phase, well, it shouldnt be that difficult to put a disclaimer telling them they are "on their own".
that why the

Beta is beta for a reason. You don't want to convert a whole large live forum over to a beta because if it breaks, you're rather stuffed

is just hogwash (forgive my french).
and neo, i would hate to begin a "mine is bigger than yours" kind of discussion, but please believe me, i developed software when you were going to the kindergarten, so please dont give me this patronizing

That might shock you, but beta isn't a stage at which you should be looking to test software if you're not up to things like starting from scratch to test.

starting a bbs system from scratch means you will be testing with miniscule boards only, which is a lousy test. any problem that has anything to do with scalability, either in the number of users, posts, pms etc. will not surface. allowing people to convert large (preferrably even huge) boards will make for a much better and more vigorous beta.
That might shock you, but running a large number of toy boards is no match to running a reasonable number of reasonable size boards, as far as testing is concerned.

and coming with all kind of justifications and rationalizations for decisions taken by other people is a lousy substitute for citical thinking.

[NeoThermic]

you all (specifically neo) seem to be making the same (false) assumption:
if i upgrade my board from v2 to v3, it doesnt necessary means that i shut down my v2 board and start operating (with all my users) on the v3 board.

You might not, but there are other users who will. I dunno about you, but I don't want to have to help people try fix a botched conversion.

and if someone will actually switch completely to olympus with a large live board still in the beta phase, well, it shouldnt be that difficult to put a disclaimer telling them they are "on their own".

Both you and I know that people don't read/care about things like that. There's a clear message in the CVS builds that the's no support in any way, shape, or form, yet you still get people asking for help with it. What good will a message that they are on their own do?

and neo, i would hate to begin a "mine is bigger than yours" kind of discussion, but please believe me, i developed software when you were going to the kindergarten, so please dont give me this patronizing

Yet you feel the need to patronise me? I'll just make a side note I never went to kindergarten, and never did the 1st or 2nd grades. I started in the 3rd grade.

starting a bbs system from scratch means you will be testing with miniscule boards only, which is a lousy test. any problem that has anything to do with scalability, either in the number of users, posts, pms etc. will not surface. allowing people to convert large (preferrably even huge) boards will make for a much better and more vigorous beta.

Now, now. It has been mentioned that the converter script won't come with the *first* beta, but who's to say that it won't come with the second beta or the first RC? Thats the real stage to check on performance, but I'll note that a fair few large forums are running from CVS code as of right now (this one included), and there's not been any performance problems or major problems. We already know 3.0 scales well.

That might shock you, but running a large number of toy boards is no match to running a reasonable number of reasonable size boards, as far as testing is concerned.

It doesn't shock me at all (yay for patronising again). I've been testing 3.0 myself with data from two very large boards without problem. This forum is large and, also, works without problem. Bokt.nl runs 3.0 without problem. eurofighter-typhoon.co.uk runs a copy of 3.0 without problem. I don't think it would be right to call Bokt.nl a 'toy' board now, would it? On the consideration its the second largest phpBB forum on the whole internet...

and coming with all kind of justifications and rationalizations for decisions taken by other people is a lousy substitute for citical thinking.

Ok, fine, if you really think that the converter should be with the beta, then give reasons, excluding scalability reasons, as we've covered those. If you really want to show critical thinking, this will be a great chance to, as currently you've just expressed opinion.

NeoThermic

[NNO-Stephen]

if you want to test how it will work w/ your board, wait till at least the RC stages before doing so. hell, you might as well go ahead and "test" the CVS versions now. the beta phase isn't to test how it'll work w/ your forum, it's to test how it works PERIOD. you can test how it'll work w/ your board once it's established that it works in the first place.

patience is a virtue.

[warmweer]

That's the whole point in the difference in names between a beta and a RC (Release Candidate). The latter being tested sufficiently and no major code changes are expected anymore so finalisation of the converter is also near (and probably changes in converter code - between the various RCs - will (probably) not require any major overhaul.

You could just as well start with phpBB3 1.0 and update everything as we go along (and now would be phpBB3 1.23 with just as many upgrades of the board and the converter).

RC stages are about getting the product "in production" and a converter is necessary for that, NOT in beta stages.