Drupal 7 RC1 Release Countdown!

Subscribing, could you provide the Spreadsheet file please?

Wow, very cool.

nifty. If you have a twitter account, I would follow it, might get more exposure than this issue.

This is so insanely helpful. Thank you!!

Now let's get crackin'! ;) *whppsshh!*

+1 for getting a twitter account

@drupalforecast ? ;)

Great work!

Note that there are also 4 non-critical release checklist issues that are also release blockers - http://drupal.org/project/drupal

@webchick: Could you please give a quick evaluation how well these projections correlate to what's happening on the ground?

For instance, does 7/13 align with core developers' time lines? How likely is it that some upcoming code sprint will radically disrupt these projections?

Fascinating projection and process...assumes all critical issues are the same? ignores the principle of "the long tail"? makes us all itchy with anticipation anyway??

go core team!

How likely is it that some upcoming code sprint will radically disrupt these projections?

Very unlikely. When we had 300 critical issues, it was easy to make a dent by sprinting, most of what's left is slog now.

The only thing which will make a significant dent in the current numbers beyond knocking them down one by one is #669048: We should have four priorities not three (introduce a new 'major' priority) - there are issues in the critical queue which aren't really release blockers - for example many of them have existed since Drupal 5 and we've had two releases come out with those bugs in, however they're either very annoying, or very serious when you hit them (but only 0.01% of sites might etc.), so demoting them to 'normal' isn't great either.

At the moment we abuse 'critical' to mean two things - 1. release blocker 2. 'we really should sort this out it's horrible'. Moving the latter category to 'major' would let us differentiate better and see what the actual release blockers are. For example there or more than 50 'critical' issues against Drupal 8 already, despite no Drupal 8 code actually existing yet - http://drupal.org/project/issues/search/drupal?text=&assigned=&submitted...

Also for people following this issue, note that 0 critical issues = release candidate, we may well have 2-3 of those before a release.

subscribing, very exciting countdown =)

lol

This countdown is a very good idea.
Though I hope, that Drupal 7 will be ready sooner. ;-)

The forecast is increasing every day. Now we can do the forecast of the forecast.

The idea of a forecast-forecast is delightfully ridiculous :) I was thinking of doing the same thing a few days ago.

Maybe a better option would be to do an exponential fit on the original critical issue data rather than the linear fit that has been used. That would provide a more accurate "forecast" considering that the rate of issue closures is slowing and, as was pointed out earlier in post #30, account for the long tail.

I really like this thread and I've been excited to see the issue count drop a good bit in the last two days.

-Josh

Actually, the best option of all would be to review and re-roll some patches to bring that critical queue down instead of dorking around with numbers. ;) ;)

var issue = 'awesome';
issue.do('subscribe');

:D

I detect a flattening in the curve...

There are two issues why progress is slow:

1. A big backlog of non-critical RTBC issues, there's been over 50 since more or less the new year. If you're following this thread and want to help, the best way would to be to work through that list oldest first, and send the most recent patch for re-testing (and possibly review the issue to see if it's still relevant and whether the patch is actually rtbc).
http://drupal.org/project/issues/search/drupal?page=1&text=&assigned=&su...

Cutting that list down to a manageable size (by either a commit spree or marking enough of them 'needs work' to allow a commit spree to make a bigger dent) would likely help the velocity of critical issues since it'd be easier to focus on just those (especially now there's not much of an rtbc critical backlog any more).

2. The remaining critical issues are in the main quite hard, in some cases there's not much actually needs to be done to get them committed, but they're long and complicated discussions in most cases which are hard to pick through and figure out what needs doing. There's not really a good solution to that. However people wanting to help, find me in irc and I'll try to find an issue that suits you.

subscribing

can I be the first one to say that this thread is actually jinxing the critical bug count. :-)))

the first estimate predicted June 20th...by which time it's looking like there'll still be 60-70 bugs.

subscribing

At the following URL a longer list of the bugcount is shown http://rym.waglo.com/drupal-7-status

Nice effort @auzigog, but couldn't you have built it with Drupal?

Nice! BTW, I also have all the bug and patch data available in csv since early November 2009 at the above URL. Oups, I was replying to http://drupal.org/node/770776#comment-3036324

The predicted release date seems to move a few days further away with every update... you sure this thread isn't jinxing it? :-)

please see #55
8-)

Yes, the lineal prediction it's a bad choice, should be potential.

Lineal correlation (R^2 = 0,8915) < Potential correlation (R^2 = 0,9493)

Do you think using older historical data would change your forecast?

there's so much I love about this thread, mainly that it shows the limits of "science". The very first post in the thread posited that it would take 2 months and 6 days to clear 110 critical issues. Now, we see the prediction that it will take 2 months and 6 days to clear 67. Superb.

I have a duck. If it sleeps under the stairs, there'll be no drupal 7 official release tomorrow, if it sleeps in the sink, there will be. Tonight it's sleeping under the stairs again. Official Duck Forecast - no drupal 7 official release tomorrow. You can sleep peacefully tonight.

of course, I also have a logarithmic duck forecast but I have to look between its legs for that.

keep up the good work! :0)

Subby.

sub-scrib-ing!

I think that's the first time I've seen the date move earlier.

interesting to note that D5 took 11 weeks between beta 1 and full release. D6 took 21 weeks for the same journey.

I know the D7 team wanted to release a beta this week (http://drupal.org/drupal-7.0-alpha5)

but there are many issues stopping it. So let's say they resolve those in a few weeks (july 20th)...then add 20 weeks to that.

My prediction: December 15th (that's for the full release...not for when the critical count gets to 0)

let's see if that's more accurate than using the matrix to make a prediction! ;-)

edit: silly typo

Wordpress 3 is out already, upgrade only took a few seconds. While Drupal is snailing behind with every new release, Wordpress is pounding out new releases like there's no tomorrow. All I can say is wow. Hopefully by the time the long awaited D7 gets released we won't be too flabbergasted from all the anticipated hype.

It isn't a race. If Wordpress is coming out with new releases, great. All the best to them.

Let Drupal 7 be released when it is ready.

As for the hype, what hype?

While Drupal is snailing behind with every new release, Wordpress is pounding out new releases like there's no tomorrow.

Apples and oranges. They actually both release pretty frequently (Drupal sends out minor releases every 1-2 months), but WP includes changes, additions, etc., into minor releases (e.g. 3.1 to 3.2) whereas Drupal will only include bug fixes in them, saving any other changes or additions for major releases. Can't really compare the two. Also, why would you want to?

Just want to echo @catch from above that 0 criticals is the criteria for the first release candidate (which might be followed by any number of other release candidates). Its not when a final release is made. We did not even reach having a beta yet.

Hard to believe someone would go through all the trouble of registering a domain for drupal7releasedate.com was it really necessary ?

#136 @espirates - never be surprised by the dedication of the Drupal community!

http://drupal7releasedate.com/graph.php shows there are 0 (ZERO) critical bugs as today. O_o

I wrote a small PHP script to predict the release date. If you want to skip the boring details, conclusion: Sep 30th, 2010. 95% confidence interval: September 13th, 2010 - October 17th, 2010.

Assumptions:
- Daily changes in the critical issue count vary approximately Normally - seems to be true, a few outliers but nothing dramatically problematic.
- Daily changes in the critical issue count are independent of each other - this is a more risky assumption, but the correlation between the change in critical issue count and the previous change in critical issue count is not statistically significant at the alpha = 0.05 level.
- The mean change in critical issue count thus far will remain essentially unchanged through the release date of Drupal 7 - this may not be the case if the issues that we're left with are tougher to solve; if this assumption is incorrect then the whole prediction is meaningless.
- I wrote my code correctly, both programatically and statistically - I attached the source code so you can judge for yourself, patches/improvements welcome.

Results:
Ran 10000 simulations using a mean daily difference of -0.7746 critical issues with a standard deviation of 3.1972 critical issues, starting point from August 14th, 2010, with 36 critical issues. The percentile lines represent the 0th (best-case), 5th, 25th, 50th (median), 75th, 95th, and 99.99th (worst-case) percentiles.

$ ./release-date.php -0.7746 3.1972 "August 14th, 2010" 36 10000 0,5,25,50,75,95,99.99
44                                                                                                          
43                                                                                                          
42                                                                                                          
41    .                                                                                                     
40   . .....                                                                                                
39  .       .. .... . . .                                                                                   
38            .    . .                                                                                      
37  ...                . ....                                                                               
36  +  ..                    .                                                                              
35  +++  ..                                                                                                 
34  .+*++  ..                 .                                                                             
33     * +   ..                .                                                                            
32  ..++* ++   ..               ...                                                                         
31    . +** +    ..                .                                                                        
30     . + * ++    ..               ....  .                                                                 
29   .    + *  +     .                  .  .                                                                
28      .  + *  ++    ..                 .                                                                  
27       .  + **  +     ..                  ...                                                             
26    .   .  +  *  ++     .                    .                                                            
25     .   .  +  *   +     ..                   ..                                                          
24          .  ++ **  +      ..                   ...       .                                               
23      .    .   +  *  ++      ..                         .. .                                              
22       .    .   +  *   +       .                       .                                                  
21        .    .   +  **  ++      ..                 ...      .                                             
20         .    .   +   *   +       .                   .      .                                            
19               .   +   *   ++      ..                         ..                                          
18          .     .   +   *    +       .                          .                                         
17           .     .   ++  **   +       ..                              .                                   
16            .     .    +   *   ++       .                         . .. .                                  
15                   .    +   *    +       ..                      . .    ..                                
14             .      .    +   **   ++       ..                             ..                              
13              .      .    +    *    +        .                              ....                          
12                      .    ++   *    +        .                                                           
11               .       .     +   *    ++       ..                               .                         
10                .       .     +   **    +        ..                                .                      
9                  .       .     +    *    ++        .                             .. .                     
8                           .     +    *     +        ..                               .. .                 
7                   .        .     ++   **    +         .                                .                  
6                    ..       ..     +    *    ++        ..                                                 
5                               .     +    *     +         .                               .                
4                                .     +    **    +         ..                                              
3                      .          .     +     *    ++         .                             ..    .         
2                                  .     ++    *     +         .                              .. . ..  .    
1                       ..          .      +    *     +         ..                              .    ..     
0                         .          .      +    **    ++         .                                     ... 
Day  15               Sep 1         15              Oct 1         15               Nov 1          15        

Sep 3 2010  
Sep 4 2010  
Sep 5 2010  
Sep 6 2010  +
Sep 7 2010  
Sep 8 2010  
Sep 9 2010  
Sep 10 2010 +
Sep 11 2010 ++
Sep 12 2010 +++
Sep 13 2010 +++
Sep 14 2010 ++++++
Sep 15 2010 ++++++++++
Sep 16 2010 ++++++++++++
Sep 17 2010 ++++++++++++++
Sep 18 2010 ++++++++++++++++++
Sep 19 2010 ++++++++++++++++++++++
Sep 20 2010 ++++++++++++++++++++++++
Sep 21 2010 ++++++++++++++++++++++++++++++++
Sep 22 2010 ++++++++++++++++++++++++++++++++
Sep 23 2010 +++++++++++++++++++++++++++++++++++++++
Sep 24 2010 ++++++++++++++++++++++++++++++++++++++++++++
Sep 25 2010 ++++++++++++++++++++++++++++++++++++++++
Sep 26 2010 ++++++++++++++++++++++++++++++++++++++++++++++
Sep 27 2010 ++++++++++++++++++++++++++++++++++++++++++++++++
Sep 28 2010 +++++++++++++++++++++++++++++++++++++++++++++++
Sep 29 2010 ++++++++++++++++++++++++++++++++++++++++++++
Sep 30 2010 ++++++++++++++++++++++++++++++++++++++++++++++++
Oct 1 2010  +++++++++++++++++++++++++++++++++++++++++++
Oct 2 2010  +++++++++++++++++++++++++++++++++++++++++++
Oct 3 2010  +++++++++++++++++++++++++++++++++++++++++
Oct 4 2010  +++++++++++++++++++++++++++++++++++++
Oct 5 2010  +++++++++++++++++++++++++++++++++++
Oct 6 2010  +++++++++++++++++++++++++++++++++
Oct 7 2010  +++++++++++++++++++++++++++++++++
Oct 8 2010  +++++++++++++++++++++++++++
Oct 9 2010  +++++++++++++++++++++++++
Oct 10 2010 +++++++++++++++++++++
Oct 11 2010 +++++++++++++++++
Oct 12 2010 ++++++++++++++++
Oct 13 2010 ++++++++++++++
Oct 14 2010 ++++++++++++
Oct 15 2010 +++++++++++
Oct 16 2010 ++++++++
Oct 17 2010 ++++++++
Oct 18 2010 ++++++++
Oct 19 2010 ++++
Oct 20 2010 +++++
Oct 21 2010 ++++
Oct 22 2010 ++++
Oct 23 2010 +++
Oct 24 2010 +++
Oct 25 2010 +
Oct 26 2010 ++
Oct 27 2010 +
Oct 28 2010 +
Oct 29 2010 
Oct 30 2010 +
Oct 31 2010 +
Nov 1 2010  
Nov 2 2010  
Nov 3 2010  +
Nov 4 2010  
Nov 5 2010  
Nov 6 2010  +
Nov 7 2010  
Nov 7 2010  
Nov 8 2010  
Nov 9 2010  
Nov 10 2010 
Nov 11 2010 
Nov 12 2010 
Nov 13 2010 
Nov 14 2010 
Nov 15 2010 
Nov 16 2010 
Nov 17 2010 
Nov 18 2010 
Nov 19 2010 
Nov 20 2010 
Nov 21 2010 
Nov 22 2010 
Nov 23 2010 
Nov 24 2010 
Nov 25 2010 
Nov 26 2010 


Avg: Sep 30 2010
Std: 8.84 days

Min: Sep 6 2010
Med: Sep 30 2010
Max: Nov 24 2010

The script is attached for peer review. I'm just a statistics student, so it's quite possible that something statistics-related is incorrect with my code, or that there's an easier way to accomplish what I've done - please let me know!

Everything seems about right, though. Thanks to arlinsandbulte for collecting the data!

It will be ready by 2012 in time for the end of the world.

@auzigog Thanks for the SimpleTest tip ;)

I too have been monitoring bugs since last year. I got caught a few with ago with a subtle change to the html code and now this, which I am just noticing. One thing I was doing (nothing pretty, hey, that's me!) was sharing the data I was gathering. Could you do this and provide a csv file? It would be great if you could grab all 4 data points too. You don't have to show them in the graph, just handy to have in the csv, since you're parsing the page anyway).

Let me know either way. If you do provide the data, I will simply point my page to yours. And if you don't, I will fix my script.

@auzigog, ah, doh, never mind, my bad. I will point my page here in that case. Oh, and I'll add a link your site too while I'm at it.

I'm gonna go out on a limb and say the first week of September. But another concern, almost as big as when Drupal 7 will ship, is when its most popular modules (with or without the #D7CX pledge) will ship their new versions. Views, for instance, has not pledged to release a D7 version on the same day D7 ships. Almost certainly many of the modules that have pledged will not actually release a stable D7 version that day. I'm sure many people would love to start migrating their websites to D7 as soon as it comes out, but won't be able to because the corresponding modules aren't ready yet. Of course if the site has been working thus far then there's probably no real urgency to upgrade to another major release, but it would still be nice.

Edit: Uh, why did I say "first week of September" when I was posting this on the first week of September? The world may never know.

Isn't views built in to D7?

daviesap: nope ... most of CCK is though. There are few basic (read *very basic*) features of views covered by the new field api but 99.9% of D7 sites will still use views.

exactly...that's why Views not being D7 ready will be a show stopper for so many, as bad as D7 itself not being ready. At least the D7CX move should prevent what happened with D6.

It's only The Most Awesome Bug Ever (TM)!

<<>>>
Free Drupal 7 From Bug Prison

=)

Good Luck, Drupal!