Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Kreuvf
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

bugs buggy wrote:
  OTOH announcing the 2.2 releases on more than just our website (Freshmeat 
 has
  only 2.1.3, the Linux Game Tome is at 2.1.1 for example) will reach a far 
 wider
  audience and thus potential testers.
 
 Last time I checked, I don't think we have ever done such a thing
 before, and I wouldn't really know how to 'announce' a new release to
 them.
When I announced it to several sites (last time I did is probably a year+ ago) I
just wrote a short e-mail summarising what's new and linking to the downloads +
change-log. That's all :D

- - Kreuvf

-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFKBouG4y86f1GXLDwRAgKdAKCXGLxa9StIuD9N8lbb9jPsXDGi3wCghqSx
DduJT23wZSBWPWMlfTeWXrg=
=3VL9
-END PGP SIGNATURE-

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Christian Ohm
On Saturday,  9 May 2009 at 15:12, bugs buggy wrote:
 The whole purpose of doing a release candidate (as opposed to another
 beta) is that to me, it is just as stable or more stable than 2.1 ever
 was when that was released.
 (Left a 4p AI game go on for 4+ hours with a release build, previous
 to that, it was a 2 hour game that I aborted.)
 If the release candidate falls on its face, then fine, maybe we should
 release another 2.1 version, but until that time, it would be quite
 unproductive to do that.

So you want to do release builds then as well? I thought it was no release
builds before release, not sure who said that though.

   OTOH announcing the 2.2 releases on more than just our website (Freshmeat 
  has
   only 2.1.3, the Linux Game Tome is at 2.1.1 for example) will reach a far 
  wider
   audience and thus potential testers.
 
 Last time I checked, I don't think we have ever done such a thing
 before, and I wouldn't really know how to 'announce' a new release to
 them.

We have a http://developer.wz2100.net/wiki/ReleaseChecklist where some sites
are listed. Most of those have a submit news link, though the freshmeat
project is owned by Karmazilla (but perhaps you can also submit updates as
logged in user), and for gamedev.net you probably also need an account.

Most of those list a 2.1 version as newest, linux-gamers even 2.0.7. So... why
don't people test 2.2? Because they like 2.1 better, or because they don't know
about it?

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Christian Ohm
On Sunday, 10 May 2009 at 15:41, Dennis Schridde wrote:
 Am Sonntag, 10. Mai 2009 14:10:00 schrieb Christian Ohm:
  So you want to do release builds then as well? I thought it was no release
  builds before release, not sure who said that though.
 In fact it is very different from that: Always release builds, never anything 
 else.
 I don't want to repeat the reasons again, so if you forgot, just ask me again.

Huh? Are we talking about the same thing? For 2.2 beta1/2 the available Windows
installers were only with debug builds, not release builds. While having
release builds as well might be an option (though not one I want to discuss
now), are you really arguing that not having debug builds is preferable? (And
if that's the case, yes, I'd like some link/explanation, as I really can't
remember any such discussion.)

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Stephen Swaney
On Sun, May 10, 2009 at 09:43:43AM -0500, Zarel wrote:

 Let's just release 2.1.4. Seriously, I've never heard of dropping
 support for the current release branch just because you want everyone
 else playing the development branch, no matter how close the
 development branch is to release (and, again, I don't consider three
 weeks if we're lucky and nothing goes wrong to be close).

Ah, the impatience of youth!  What fraction of our development cycle
is 3 weeks?

If you are going to release another 2.1, do it.  Just don't do it at
the same time as 2.2 RC1.  It's bad marketing.  It dilutes both the
impact and user testing of the RC release.

-- 
Stephen Swaney  
sswa...@centurytel.net


___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Dennis Schridde
Am Sonntag, 10. Mai 2009 16:43:43 schrieb Zarel:
 2009/5/9 Christian Ohm chr@gmx.net:
  I don't think releasing 2.1.4 will result in much less _useful_ testing
  for 2.2. Yes, you might get more people to try 2.2, but if they are
  tricked into using it they'll just see it's unstable yet, and go back
  to 2.1 (if 2.1 isn't too unstable for them either) instead of giving
  useful bugreports. But with a 2.1.4 release those who don't want to test
  2.2 get a better 2.1.

 What is it now?

 Devu, Christian, Kreuf, Per, me in support
 cybersphinx neutral
 Buggy, stiv against
And while all that talking was going on, one could have created a hundred 
builds...


signature.asc
Description: This is a digitally signed message part.
___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread Kreuvf
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Dennis Schridde wrote:
 Am Sonntag, 10. Mai 2009 16:43:43 schrieb Zarel:
 2009/5/9 Christian Ohm chr@gmx.net:
 I don't think releasing 2.1.4 will result in much less _useful_ testing
 for 2.2. Yes, you might get more people to try 2.2, but if they are
 tricked into using it they'll just see it's unstable yet, and go back
 to 2.1 (if 2.1 isn't too unstable for them either) instead of giving
 useful bugreports. But with a 2.1.4 release those who don't want to test
 2.2 get a better 2.1.
 What is it now?

 Devu, Christian, Kreuf, Per, me in support
 cybersphinx neutral
 Buggy, stiv against
 And while all that talking was going on, one could have created a hundred 
 builds...
After all team communication and coordination is essential.

- - Kreuvf
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFKBxSE4y86f1GXLDwRAqhWAJ9LO3b6eH5eQ2qOstVUS2F0U3ZwUQCgxIgn
CMHNYiStoSmde6InRP5eTBo=
=NzHm
-END PGP SIGNATURE-

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-10 Thread bugs buggy
On 5/10/09, Zarel zare...@gmail.com wrote:
  Let's just release 2.1.4. Seriously, I've never heard of dropping
  support for the current release branch just because you want everyone
  else playing the development branch, no matter how close the
  development branch is to release (and, again, I don't consider three
  weeks if we're lucky and nothing goes wrong to be close).

...
*Buginator's head explodes*

[speaker]Clean up on the warzone dev list![/speaker]


  2009/5/10 Dennis Schridde devuran...@gmx.net:
  Am Sonntag, 10. Mai 2009 14:10:00 schrieb Christian Ohm:
   So you want to do release builds then as well? I thought it was no 
 release
   builds before release, not sure who said that though.
   In fact it is very different from that: Always release builds, never 
 anything
   else.
   I don't want to repeat the reasons again, so if you forgot, just ask me 
 again.


 I dunno. The policy I've been recommending is (and what we've been
  doing since 2.1.2, iirc):

  Release builds only for stable, RC
  Debug builds only for beta, alpha, everything else
  Or maybe both for RC.

  Ensures testers have proper tools for testing and can provide good
  feedback, but stable is stable enough for widespread use.

  'Course, that's just my opinion - I have no idea what official policy
  is. And yes, I don't think I was here the last time you gave the
  reasons for always release builds, so I'd like to hear them again.

On hindsight, I think we should go back to all release builds (or
offer them both).
No matter how many times you try to explain the difference between
release  debug builds, the bottom line is people just *think* the
wrong thing/idea about when they play a debug build.

Debug builds are fine for those that understand what they are, and
more importantly, get the needed info from the assert, and the
callstack, but the vast majority of our userbase don't.  You can see
all the assert screenshots in the forums.

As I mentioned someplace, release builds aren't stripped, so that is a big plus.

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-09 Thread Kreuvf
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

bugs buggy wrote:
 On 5/8/09, Kreuvf kre...@warzone2100.de wrote:
  bugs buggy wrote:
   2.2 has been 'cooking' for a extremely long period of time.
   Pushing beta builds to a lackluster community response,  and, what is
   worse is, people are still creating new content for 2.1, which won't
   work with 2.2.

 Has there ever benn some kind of announcement/news/blog entry to make sure 
 that
  people _know_ about this? Something definite, something people can rely on 
 ;)
 
 Besides the Front page announcing the beta candidates, and the global
 announcement in the forums, and the talk on the ML, then, no.
And those announcements clearly state that the WRP recommends against creating
new content for 2.1 as there will be no compatibility to 2.2? And ML talk is not
visible (mind the quotation marks!) to the masses IMHO, forums is where all
the action takes place.

The global announcement (http://forums.wz2100.net/viewtopic.php?f=1t=2950) has
a misleading subject (misleading only if you want to tell people to not create
new 2.1 content). And the whole topic doesn't even mention any incompatibilities
with 2.1, 2.1 isn't even mentioned at all.

As I cover new versions on my site as well I read all the release announcements
and I cannot remember _anything_ about the attitude of the WRP towards 2.1 and
new content for it.

Perhaps it is a good idea to explicitly state what's up with 2.1 with the
release of the next upgrade for 2.1. Actually this is a good point for releasing
another upgrade for 2.1 ;)

Thanks for reading,
- - Kreuvf
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFKBVN94y86f1GXLDwRAh4xAKDFEe6ktgT2W0iJViwbHa7oRIyiQQCfSc3a
x8BAQh2qrrlVXtIskzAfhSM=
=EqZY
-END PGP SIGNATURE-

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-09 Thread Christian Ohm
On Friday,  8 May 2009 at  0:54, Zarel wrote:
 2009/5/8 bugs buggy buginato...@gmail.com:
  People are not going to test 2.2 if the 'most requested bugfixes' are
  fixed in 2.1.  They have no motivation to upgrade.
 I should also mention that I accept this as a valid objection to
 2.1.4, so there's little point in arguing further.

I don't think releasing 2.1.4 will result in much less _useful_ testing for
2.2. Yes, you might get more people to try 2.2, but if they are tricked into
using it they'll just see it's unstable yet, and go back to 2.1 (if 2.1 isn't
too unstable for them either) instead of giving useful bugreports. But with a
2.1.4 release those who don't want to test 2.2 get a better 2.1.

OTOH announcing the 2.2 releases on more than just our website (Freshmeat has
only 2.1.3, the Linux Game Tome is at 2.1.1 for example) will reach a far wider
audience and thus potential testers.

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-09 Thread bugs buggy
On 5/9/09, Christian Ohm chrx...@gmx.net wrote:
 On Friday,  8 May 2009 at  0:54, Zarel wrote:
   2009/5/8 bugs buggy buginax...@gmail.com:
People are not going to test 2.2 if the 'most requested bugfixes' are
fixed in 2.1.  They have no motivation to upgrade.
   I should also mention that I accept this as a valid objection to
   2.1.4, so there's little point in arguing further.


 I don't think releasing 2.1.4 will result in much less _useful_ testing for
  2.2. Yes, you might get more people to try 2.2, but if they are tricked 
 into
  using it they'll just see it's unstable yet, and go back to 2.1 (if 2.1 isn't
  too unstable for them either) instead of giving useful bugreports. But with a
  2.1.4 release those who don't want to test 2.2 get a better 2.1.

Yes, there are going to be issues with 2.2.  Just like there were
issues for 2.1, 2.0 and all the way down the line.

The whole purpose of doing a release candidate (as opposed to another
beta) is that to me, it is just as stable or more stable than 2.1 ever
was when that was released.
(Left a 4p AI game go on for 4+ hours with a release build, previous
to that, it was a 2 hour game that I aborted.)
If the release candidate falls on its face, then fine, maybe we should
release another 2.1 version, but until that time, it would be quite
unproductive to do that.


  OTOH announcing the 2.2 releases on more than just our website (Freshmeat has
  only 2.1.3, the Linux Game Tome is at 2.1.1 for example) will reach a far 
 wider
  audience and thus potential testers.

Last time I checked, I don't think we have ever done such a thing
before, and I wouldn't really know how to 'announce' a new release to
them.

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-08 Thread Kreuvf
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

bugs buggy wrote:
 2.2 has been 'cooking' for a extremely long period of time.
 Pushing beta builds to a lackluster community response,  and, what is
 worse is, people are still creating new content for 2.1, which won't
 work with 2.2.
Has there ever benn some kind of announcement/news/blog entry to make sure that
people _know_ about this? Something definite, something people can rely on ;)
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFKBITc4y86f1GXLDwRAkeDAKDLYm+nYjvC9d1PXT7j4AHBvlDsGgCgsKya
GFEsq+BksFGw43d/VZr+h/I=
=pstX
-END PGP SIGNATURE-

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-08 Thread Dennis Schridde
Am Freitag, 8. Mai 2009 07:13:48 schrieb Stephen Swaney:
 Actually, releasing two versions should be approximately twice as much
 work.
Build scripts are written and if no one broke them, then they will still work. 
Upload takes under an hour (depending on your connection).

Why did we backport bugfixes to 2.1 at all, if we are so very determined to 
never release them? In fact, why do we backport any bugfixes? If we are so 
much against maintaining older branches, why don't we just release a 2.2 
final, and continue with the 2.3 alphas/betas/rcs, to start 2.4 immediately 
after 2.3 final?

Actually not all of the above is pure sarcasm. It has some valid point, imo.
If attention is split so badly, and that is hurting us so much, we could as 
well not draw any attention from our latest efforts to something legacy.
On the other hand, if we want to support legacy versions, the statement that 
they prevent testing of the next generation looks somewhat against the policy.

--DevU


signature.asc
Description: This is a digitally signed message part.
___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-08 Thread bugs buggy
On 5/8/09, Kreuvf kre...@warzone2100.de wrote:
  bugs buggy wrote:
   2.2 has been 'cooking' for a extremely long period of time.
   Pushing beta builds to a lackluster community response,  and, what is
   worse is, people are still creating new content for 2.1, which won't
   work with 2.2.

 Has there ever benn some kind of announcement/news/blog entry to make sure 
 that
  people _know_ about this? Something definite, something people can rely on ;)

Besides the Front page announcing the beta candidates, and the global
announcement in the forums, and the talk on the ML, then, no.

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread bugs buggy
On 5/7/09, Zarel zarexxx...@gmail.com wrote:
 2009/5/7 bugs buggy bugixxx...@gmail.com:
  And I would like to remind everyone, that 2.2 is just as stable (if
   not more so) as 2.1 was at the time of its release.

 Are you sure of this? The forums are rife with people going back to
  2.1 because 2.2 is too unstable. Complain all you want about the
  vagueness of their bug reports, but the fact remains that the bugs are
  there.

  - Zarel

How about you compare apples to apples?
Those are 'debug' builds, as you are well aware of, and I am sure I
don't have to explain this to you, but since you brought it up...

The main reason why people find the current 2.2 beta 'unstable' is it
is a *debug* build.  Debug builds assert.  Release builds don't.  That
isn't exactly a news flash is it?
So of course a *debug* build will *seem* more 'unstable' than a
release build, it is *meant* to give us more information about
*possible* issues.

If we were to release a new 2.1 build and only do a *debug* build,
you will pretty much see the same issues  asserts, and, the forums
will be wondering how 2.1 became so 'unstable'.

If you look at trac, those (few) people that have reported issues,
they are all pretty much assert issues, and some issues that were
introduced a long time ago, but since we sit on builds for months at a
time without a release, things slip past us.

Perhaps to prove (or disprove) a point, we should have just a release
version build of this forthcoming Release Candidate.
The release versions should not have the .exe stripped, so it will
still have enough debug information to find the issues (which AFAIK,
is what we do now anyway).

And to tie this all up, we pretty much all agreed in the past, that we
should do ... I'd rather release early and often (with some bugs)
than less often with less frequent testing.

But, we are falling for the same mistakes we made in the past, and
things are getting in the code base, with little (or no) testing, and
even fewer people willing to test.  Then when we do finally release,
we then get the long needed/wanted bug reports.

There is only one solution for this, and that is to : Release early 
release often!

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread Zarel
2009/5/7 bugs buggy buginato...@gmail.com:
 How about you compare apples to apples?
 Those are 'debug' builds, as you are well aware of, and I am sure I
 don't have to explain this to you, but since you brought it up...

 The main reason why people find the current 2.2 beta 'unstable' is it
 is a *debug* build.  Debug builds assert.  Release builds don't.  That
 isn't exactly a news flash is it?
 So of course a *debug* build will *seem* more 'unstable' than a
 release build, it is *meant* to give us more information about
 *possible* issues.

 If we were to release a new 2.1 build and only do a *debug* build,
 you will pretty much see the same issues  asserts, and, the forums
 will be wondering how 2.1 became so 'unstable'.

See, the problem is that people are reporting that the game crashes
whenever people click Ignore on the assert. Since clicking Ignore
causes a crash for these asserts, I'd think that nearly every assert
corresponds to a crash in the release version.

- Zarel

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread Zarel
By the way, guys, is it okay if we release 2.1.4 concurrently with 2.2 RC1?

Buggy and stiv seem to be very against the idea, Cybersphinx sounds
neutral, Per sounds for it.

Here's why I think releasing 2.1.4 is a good idea:

- As mentioned earlier in the 2.2 RC1 thread, it's not as stable as 2.1 yet.
- 2.1.4 has one of the most requested bugfixes - namely, how the
interface hides itself whenever someone changes
color/team/position/etc in multiplayer game staging
- We're not going to be releasing a 2.2 stable for at least another
three weeks - people should get an interim stable release.
- It theoretically shouldn't be more work than just making a release.
I mean, it sounds like we're dropping support for 2.1 either way.

I can handle most of the work - I just need people to compile and
upload source tarballs, Windows binaries, and Mac binaries. Giel?
EvilGuru? If we do it at the same time as 2.2 RC2, it shouldn't be too
much more work, right?

- Zarel

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread bugs buggy
On 5/8/09, Zarel zare...@gmail.com wrote:
 By the way, guys, is it okay if we release 2.1.4 concurrently with 2.2 RC1?

  Buggy and stiv seem to be very against the idea, Cybersphinx sounds
  neutral, Per sounds for it.

  Here's why I think releasing 2.1.4 is a good idea:

  - As mentioned earlier in the 2.2 RC1 thread, it's not as stable as 2.1 yet.
  - 2.1.4 has one of the most requested bugfixes - namely, how the
  interface hides itself whenever someone changes
  color/team/position/etc in multiplayer game staging
  - We're not going to be releasing a 2.2 stable for at least another
  three weeks - people should get an interim stable release.
  - It theoretically shouldn't be more work than just making a release.
  I mean, it sounds like we're dropping support for 2.1 either way.

  I can handle most of the work - I just need people to compile and
  upload source tarballs, Windows binaries, and Mac binaries. Giel?
  EvilGuru? If we do it at the same time as 2.2 RC2, it shouldn't be too
  much more work, right?


  - Zarel


My biggest objection to having another 2.1 release is that it will
serve no real purpose, besides for us having to deal with another 2.1
release.  (More bug reports from 2.1.x, even though they most likely
have been fixed in 2.2, and whatever else... and I don't have my
epitaph ready either! :P)

People are not going to test 2.2 if the 'most requested bugfixes' are
fixed in 2.1.  They have no motivation to upgrade.

I rather have everyone hit 2.2 RC *hard*, and for the next 3 weeks,
file as many bug reports as possible, so we can release a better 2.2
release.
This will be much better for our project, in both the short and long run.

Hope that clarifys things.

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread Stephen Swaney
On Thu, May 07, 2009 at 11:30:47PM -0500, Zarel wrote:
 By the way, guys, is it okay if we release 2.1.4 concurrently with 2.2 RC1?
 
 Buggy and stiv seem to be very against the idea, Cybersphinx sounds
 neutral, Per sounds for it.

[16:28] per it is hardly effortless
[16:28] per and when looking more closely at the bugs fixed since 2.1.3, i'm 
not so sure anymore

 Here's why I think releasing 2.1.4 is a good idea:
 
 - As mentioned earlier in the 2.2 RC1 thread, it's not as stable as 2.1 yet.

Compile 2.2 RC1 with the asserts off.  That seemed to be the issue, judging
from the IRC conversation.

 - 2.1.4 has one of the most requested bugfixes - namely, how the
 interface hides itself whenever someone changes
 color/team/position/etc in multiplayer game staging

I concede to your claim that this is the most important bugfix
in the history of software development, but isn't it also in 2.2?

 - We're not going to be releasing a 2.2 stable for at least another
 three weeks - people should get an interim stable release.

3 Weeks - that's like 21 days?

 - It theoretically shouldn't be more work than just making a release.
 I mean, it sounds like we're dropping support for 2.1 either way.

Aren't we?  And if we are, why not get on with 2.2?

 I can handle most of the work - I just need people to compile and
 upload source tarballs, Windows binaries, and Mac binaries. Giel?
 EvilGuru? If we do it at the same time as 2.2 RC2, it shouldn't be too
 much more work, right?

Actually, releasing two versions should be approximately twice as much work.

-- 
Stephen Swaney  
sswa...@centurytel.net


___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev


Re: [Warzone-dev] 2.2 RC1 scheduled for release this weekend

2009-05-07 Thread Zarel
2009/5/8 Stephen Swaney sswa...@centurytel.net:
 [16:28] per it is hardly effortless
 [16:28] per and when looking more closely at the bugs fixed since 2.1.3, 
 i'm not so sure anymore

It seems a bit disingenuous to quote per from before I explained what
the bugs were. Here are some quotes from _after_ he said that:

[16:27] per and when looking more closely at the bugs fixed since
2.1.3, i'm not so sure anymore
[16:27] per which bug is it that warrants the release?
[16:28] Zarel * Fix: No longer shows map and hides interface when
others switch color/position/team/etc in multiplayer staging area.
(r7257)
[16:28] Zarel That's the most requested bugfix in #warzone2100-games

[16:30] per Zarel: if EvilGuru and Giel agree to spend the effort to
make builds, i see now reason not to do it

[16:31] Zarel Meh. What if we release it concurrently with 2.2 RC1?
[16:31] per well, that kinda has a precedent, doesn' it ;)
[16:31] Zarel ?
[16:31] per i mean, we did that for the previous release
[16:31] per so why not

 Compile 2.2 RC1 with the asserts off.  That seemed to be the issue, judging
 from the IRC conversation.

This is why not:

 See, the problem is that people are reporting that the game crashes
 whenever people click Ignore on the assert. Since clicking Ignore
 causes a crash for these asserts, I'd think that nearly every assert
 corresponds to a crash in the release version.

^^ which is the bigger issue.

 I concede to your claim that this is the most important bugfix
 in the history of software development, but isn't it also in 2.2?

True, but 2.2 is a significant time off.

 3 Weeks - that's like 21 days?

:P That's at least.

 Aren't we?  And if we are, why not get on with 2.2?

We don't really know for sure if 2.2 is actually going to be rushed
out that fast. While I currently have no objections, who knows what
could turn up? We're getting quite a few bug reports from the
allegedly small number of testers Buggy complains about.

 Actually, releasing two versions should be approximately twice as much work.

I'd assume that making a release is one of those things for which the
amount of time spent preparing for it is significantly greater than
the amount of time spent during it. It's kind of like writing an
e-mail. Writing a six-sentence message at once is a lot easier than
writing one sentence each day for six days, since most of the time is
spent in opening the e-mail client, going to Drafts, opening the
message, collecting your thoughts, etc, etc.

Heck, I wouldn't even need to say I'd assume - I know from
experience that two releases at once is easier than two releases
separately. The only part I don't know how to do is upload to Gna (or
build 2.1, but as long as it isn't significantly different from
building 2.2, my experience applies), and I've uploaded to other
places often enough to know it's easier to do concurrently than
separately.

2009/5/8 bugs buggy buginato...@gmail.com:
 People are not going to test 2.2 if the 'most requested bugfixes' are
 fixed in 2.1.  They have no motivation to upgrade.

I should also mention that I accept this as a valid objection to
2.1.4, so there's little point in arguing further.

-Zarel

___
Warzone-dev mailing list
Warzone-dev@gna.org
https://mail.gna.org/listinfo/warzone-dev