[#39432] require and case — Martin Stannard <martin@...>

17 messages 2002/05/02

[#39463] RE: require and case — "Berger, Daniel" <djberge@...>

> --- Erik Bagfors <erik@bagfors.nu> wrote:

32 messages 2002/05/02
[#39699] Re: require and case — "Baptiste Lepilleur" <gaiacrtn@...> 2002/05/06

I confirm this behavior. I stumbled on something similar on ruby 1.6.7

[#39700] Re: require and case — ts <decoux@...> 2002/05/06

>>>>> "B" == Baptiste Lepilleur <gaiacrtn@free.fr> writes:

[#39718] Re: require and case — Dave Thomas <Dave@...> 2002/05/06

ts <decoux@moulon.inra.fr> writes:

[#39735] Re: require and case — matz@... (Yukihiro Matsumoto) 2002/05/06

Hi,

[#39534] dynamically assigning instance variables — "Berger, Daniel" <djberge@...>

Hi all,

25 messages 2002/05/03
[#39547] Re: dynamically assigning instance variables — harryo@... (Harry Ohlsen) 2002/05/04

wconrad@yagni.com wrote in message news:<20020503220050.GA443@pluto>...

[#39565] Re: dynamically assigning instance variables — Jean-Hugues ROBERT <jean_hugues_robert@...> 2002/05/04

Hello,

[#39568] Class methods defined in a module, how-to ? — Jean-Hugues ROBERT <jean_hugues_robert@...> 2002/05/04

[#39585] Ruby vs. Java vs. Native trivia — ser@... (Sean Russell)

Hi,

18 messages 2002/05/04

[#39657] newbie Q: how to strip blank lines from file? — Stewart Midwinter <stewart@..._midwinter.ca>

Well, I've spent a few days reading about Ruby, and want to try my first

11 messages 2002/05/06

[#39723] WWW.RUBYCONF.ORG — Christine Hall <return@...>

15 messages 2002/05/06

[#39796] Result of I need your experience - classification and comparison of languages — yvan.radenac@... (Yvan Radenac)

Hi,

21 messages 2002/05/07

[#39809] 'Ultimate' FreeRIDE ? — "Euan Mee" <xlucid@...>

[Note: This is a post to the FreeRIDE developers mailing list, which I have

16 messages 2002/05/07
[#39871] Re: 'Ultimate' FreeRIDE ? — Lothar Scholz <llothar@...> 2002/05/08

On Tue, 07 May 2002 20:52:06 GMT, "Euan Mee"

[#39821] non-alphabetic character in symbol — kwatch@... (kwatch)

Hi,

14 messages 2002/05/08

[#39887] Thread#join doesn't accept a timeout? — Dossy <dossy@...>

Hi,

27 messages 2002/05/09
[#39961] RE: Thread#join doesn't accept a timeout? — "Nathaniel Talbott" <nathaniel@...> 2002/05/09

Dossy [mailto:dossy@panoptic.com] wrote:

[#39968] Re: Thread#join doesn't accept a timeout? — Dossy <dossy@...> 2002/05/09

On 2002.05.10, Nathaniel Talbott <nathaniel@talbott.ws> wrote:

[#39898] cgi params api — patrick-may@... (Patrick May)

One thing that I don't like is the way cgi forces you to treat every

93 messages 2002/05/09
[#40032] Re: cgi params api — patrick-may@... (Patrick May) 2002/05/10

patrick-may@monmouth.com (Patrick May) wrote in message news:<3b3ad3b4.0205091447.5b00ce98@posting.google.com>...

[#40037] Re: cgi params api — Jean-Hugues ROBERT <jean_hugues_robert@...> 2002/05/10

At 02:23 11/05/2002 +0900, you wrote:

[#40049] Re: cgi params api — Wakou Aoyama <wakou@...> 2002/05/10

On Sat, May 11, 2002 at 03:27:13AM +0900,

[#40052] Re: cgi params api — Jean-Hugues ROBERT <jean_hugues_robert@...> 2002/05/10

At 05:56 11/05/2002 +0900, you wrote:

[#40137] Re: cgi params api — patrick-may@... (Patrick May) 2002/05/12

Wakou Aoyama <wakou@fsinet.or.jp> wrote in message news:<20020511121152.GA29832%wakou@fsinet.or.jp>...

[#40139] Re: cgi params api — Sean Chittenden <sean@...> 2002/05/12

> > If you access reqs['key'], then you'll always get a non-array object.

[#40144] Re: cgi params api — Dossy <dossy@...> 2002/05/12

On 2002.05.12, Sean Chittenden <sean@chittenden.org> wrote:

[#40148] Re: cgi params api — David Alan Black <dblack@...> 2002/05/12

Hello --

[#40159] Re: cgi params api — Dossy <dossy@...> 2002/05/12

On 2002.05.12, David Alan Black <dblack@candle.superlink.net> wrote:

[#40162] Re: cgi params api — Wakou Aoyama <wakou@...> 2002/05/12

On Mon, May 13, 2002 at 01:37:40AM +0900,

[#40191] Re: cgi params api — Dossy <dossy@...> 2002/05/13

On 2002.05.13, Wakou Aoyama <wakou@fsinet.or.jp> wrote:

[#40194] Re: cgi params api — David Alan Black <dblack@...> 2002/05/13

Hello --

[#40197] Re: cgi params api — Dossy <dossy@...> 2002/05/13

On 2002.05.13, David Alan Black <dblack@candle.superlink.net> wrote:

[#40198] Re: cgi params api — David Alan Black <dblack@...> 2002/05/13

Hi --

[#40203] Re: cgi params api — Dossy <dossy@...> 2002/05/13

On 2002.05.13, David Alan Black <dblack@candle.superlink.net> wrote:

[#40209] Re: cgi params api — Dave Thomas <Dave@...> 2002/05/13

Dossy <dossy@panoptic.com> writes:

[#40222] Re: cgi params api — Dossy <dossy@...> 2002/05/13

On 2002.05.13, Dave Thomas <Dave@PragmaticProgrammer.com> wrote:

[#40228] Re: cgi params api — Wakou Aoyama <wakou@...> 2002/05/13

On Tue, May 14, 2002 at 12:59:37AM +0900,

[#40229] Re: cgi params api — Dossy <dossy@...> 2002/05/13

On 2002.05.14, Wakou Aoyama <wakou@fsinet.or.jp> wrote:

[#40250] Re: cgi params api — Wakou Aoyama <wakou@...> 2002/05/13

Hi,

[#40291] Re: cgi params api — Dossy <dossy@...> 2002/05/14

On 2002.05.14, Wakou Aoyama <wakou@fsinet.or.jp> wrote:

[#40322] Re: cgi params api — Wakou Aoyama <wakou@...> 2002/05/14

Hi,

[#40377] Re: cgi params api — patrick-may@... (Patrick May) 2002/05/15

Dossy <dossy@panoptic.com> wrote in message news:<20020514172939.GL14145@panoptic.com>...

[#40387] Re: cgi params api — Dossy <dossy@...> 2002/05/15

On 2002.05.15, Patrick May <patrick-may@monmouth.com> wrote:

[#40389] Re: cgi params api — David Alan Black <dblack@...> 2002/05/15

Hello --

[#40390] Re: cgi params api — Dossy <dossy@...> 2002/05/15

On 2002.05.15, David Alan Black <dblack@candle.superlink.net> wrote:

[#40394] Re: cgi params api — David Alan Black <dblack@...> 2002/05/15

Hello --

[#40004] Ruby Conference 2002: Call for Presenters — David Alan Black <dblack@...>

Ruby Conference 2002: Call for Presenters

25 messages 2002/05/10
[#40009] Re: Ruby Conference 2002: Call for Presenters — Jim Menard <jimm@...> 2002/05/10

David Alan Black <dblack@candle.superlink.net> writes:

[#40043] Premature end of script headers — ccos <ccos@...> 2002/05/10

hello,

[#40015] BUG! Platform independent? sockets and select — "Kontra, Gergely" <kgergely@...>

Hi!

12 messages 2002/05/10

[#40099] OT:is software eng an art? — ptkwt@...1.aracnet.com (Phil Tomson)

I signed up for a free seminar that's being held at a grad school nearby

34 messages 2002/05/11

[#40105] Re: OT:is software eng an art? — "Radu M. Obad磚 <whizkid@...>

Mind me... but I feel like stating my oppinions regarding this issue. I

30 messages 2002/05/11

[#40180] What is Ruby for? — "Steve Merrick" <Steve.Merrick@...>

Or even 'Why <insert scripting language of your choice>'? I know it's me

35 messages 2002/05/13

[#40346] ANN: REXML 2.3.3 — Sean Russell <ser@...>

Getting tired of the upgrades yet?

50 messages 2002/05/14
[#40407] Re: ANN: REXML 2.3.3 — Sean Russell <ser@...> 2002/05/15

<posted & mailed>

[#40412] RE: ANN: REXML 2.3.3 — "Rich Kilmer" <rich@...> 2002/05/15

I'll weigh in here...

[#40418] RE: ANN: REXML 2.3.3 — David Alan Black <dblack@...> 2002/05/15

Hi --

[#40426] Re: ANN: REXML 2.3.3 — Bob Hutchison <hutch@...> 2002/05/15

On 5/15/02 12:53 PM, "David Alan Black" <dblack@candle.superlink.net> wrote:

[#40440] Re: ANN: REXML 2.3.3 — Tobias Reif <tobiasreif@...> 2002/05/15

Bob Hutchison schrieb:

[#40397] ANN: Programmierung in Ruby — juergen.katins@... (Juergen Katins)

The translation of *Programming Ruby* by Dave Thomas and Andy Hunt is

31 messages 2002/05/15

[#40525] Metaclasses... — "Hal E. Fulton" <hal9000@...>

Hello, all...

15 messages 2002/05/16

[#40559] what's the calling method name? — Yohanes Santoso <ruby-talk@...>

Hello all,

27 messages 2002/05/17
[#40566] Re: what's the calling method name? — John Carter <john.carter@...> 2002/05/17

On Fri, 17 May 2002, Yohanes Santoso wrote:

[#40577] Re: what's the calling method name? — Yohanes Santoso <ruby-talk@...> 2002/05/17

John Carter <john.carter@tait.co.nz> writes:

[#40647] Re: what's the calling method name? — Yohanes Santoso <ruby-talk@...> 2002/05/18

Yohanes Santoso <ruby-talk@jenny-gnome.dyndns.org> writes:

[#40735] Re: what's the calling method name? — wconrad@... 2002/05/19

On Sat, May 18, 2002 at 06:16:29PM +0900, Yohanes Santoso wrote:

[#40782] Re: what's the calling method name? — nobu.nokada@... 2002/05/20

Hi,

[#40571] Shifting array element & regex on array element — Yohanes Santoso <freeride-devel@...>

In implementing a buffer gap mechanism, I was wondering if there is

10 messages 2002/05/17

[#40635] Ruby regex question — Dossy <dossy@...>

Maybe this is Perl envy, maybe I'm just doing something wrong.

28 messages 2002/05/18
[#40664] Re: Ruby regex question — Mike Stok <mike@...> 2002/05/18

In article <20020518152610.GJ9684@panoptic.com>, Dossy wrote:

[#40673] Re: Ruby regex question — Dossy <dossy@...> 2002/05/18

On 2002.05.19, Mike Stok <mike@stok.co.uk> wrote:

[#40687] RubyEclipse released — "Adam Williams" <awilliams@...>

First release of the highly anticipated (maybe just by me) Ruby IDE. Get it

32 messages 2002/05/19
[#40720] Re: [ANN] RubyEclipse released — Han Holl <han@...> 2002/05/19

Adam Williams wrote:

[#40741] RE: [ANN] RubyEclipse released — "Adam Williams" <awilliams@...> 2002/05/19

Okay. So I probably need to work on some documentation.

[#40723] are there unit tests for cgi.rb? — patrick-may@... (Patrick May)

if so, where could I find them?

27 messages 2002/05/19
[#40806] Re: are there unit tests for cgi.rb? — Tobias Reif <tobiasreif@...> 2002/05/20

Wakou Aoyama wrote:

[#40812] Re: are there unit tests for cgi.rb? — Wakou Aoyama <wakou@...> 2002/05/20

On Tue, May 21, 2002 at 12:27:46AM +0900,

[#40902] method_missing is cool (was: are there unit tests for cgi.rb?) — Ian Macdonald <ian@...> 2002/05/21

On Tue 21 May 2002 at 01:00:22 +0900, Wakou Aoyama wrote:

[#40778] automatic documentation: using tests in addition to / instead of comments — Tobias Reif <tobiasreif@...>

Hi,

40 messages 2002/05/20
[#40788] Re: [RDoc etc] automatic documentation: using tests in addition to / instead of comments — Dave Thomas <Dave@...> 2002/05/20

Tobias Reif <tobiasreif@pinkjuice.com> writes:

[#40826] DBI access to mysql auto_increment record numbers? — Brad Cox <bcox@...> 2002/05/20

How do you get the auto_increment record number for mysql in

[#40887] Re: [RDoc etc] automatic documentation: using tests in addition to / instead of comments — Sean Russell <ser@...> 2002/05/21

Man, how did I miss this thread?

[#40889] Re: [RDoc etc] automatic documentation: using tests in addition to / instead of comments — Paul Brannan <pbrannan@...> 2002/05/21

On Wed, May 22, 2002 at 12:14:05AM +0900, Sean Russell wrote:

[#40875] Memory consumption. — Erik Terpstra <erik@...>

Is there some method that tells me how much memory a certain object

14 messages 2002/05/21

[#40978] Re: Stymied by Ruby's garbage collector — Art Taylor <ataylor@...>

Is there a particular kind or kinds of object being created in huge numbers?

25 messages 2002/05/22
[#40986] Re: Stymied by Ruby's garbage collector — Bob Hutchison <hutch@...> 2002/05/22

On 5/22/02 4:07 PM, "Art Taylor" <ataylor@fortpoint.com> wrote:

[#41085] OS independent scripts, system calls on Windows and Linux — Tobias Reif <tobiasreif@...>

Hi,

19 messages 2002/05/24

[#41102] RE: OS independent scripts, system calls on Windows and Linux — "Morris, Chris" <chris.morris@...>

> All these incompatibility issues are why I want to talk about an RCR.

29 messages 2002/05/24
[#41105] Re: OS independent scripts, system calls on Windows and Linux — Tobias Reif <tobiasreif@...> 2002/05/24

Morris, Chris wrote:

[#41266] Re: OS independent scripts, system calls on Windows and Linux — Dennis Newbold <dennisn@...> 2002/05/28

On Fri, 24 May 2002, Tobias Reif wrote:

[#41352] Infinity (?!) — ptkwt@...1.aracnet.com (Phil Tomson)

I didn't know there was an Infinity value in Ruby, just found it today:

34 messages 2002/05/30

[#41434] Ruby jobs — Tobias Reif <tobiasreif@...>

Hi,

23 messages 2002/05/31
[#41475] RE: Ruby jobs — <james@...> 2002/05/31

>

[#41510] Finding all applications — Chris Gehlker <gehlker@...>

In trying to add drag and drop scripting to RubyStudio and the first task is

40 messages 2002/05/31
[#41561] Re: Finding all applications — Sean Russell <ser@...> 2002/06/01

Jim Menard wrote:

[#41564] Re: Finding all applications — nobu.nokada@... 2002/06/01

Hi,

[#41575] Re: Finding all applications — Chris Gehlker <gehlker@...> 2002/06/02

On 6/1/02 11:33 AM, "nobu.nokada@softhome.net" <nobu.nokada@softhome.net>

[#41576] RE: Finding all applications — "Mike Campbell" <michael_s_campbell@...> 2002/06/02

>> ruby -r find -e 'Find.find("/"){|f| puts f if f[/\.app$/]}'

[#41578] Re: Finding all applications — Chris Gehlker <gehlker@...> 2002/06/02

On 6/1/02 5:20 PM, "Mike Campbell" <michael_s_campbell@yahoo.com> wrote:

[#41579] Re: Finding all applications — Dossy <dossy@...> 2002/06/02

On 2002.06.02, Chris Gehlker <gehlker@fastq.com> wrote:

[#41586] Re: Finding all applications — Chris Gehlker <gehlker@...> 2002/06/02

On 6/1/02 7:44 PM, "Dossy" <dossy@panoptic.com> wrote:

Re: Stymied by Ruby's garbage collector

From: "Nat Pryce" <nat.pryce@...13media.com>
Date: 2002-05-22 15:28:16 UTC
List: ruby-talk #40952
I had the same problem writing a game in Eiffel, also using the SDL.  Here's
the steps I took:

1) Explicitly invoked the GC each frame.  That ensured that there was about
the same amount of delay for GC every frame, instead of sudden lurches every
few seconds.

2) Ran the animation loop as fast as possible (with a tiny delay to allow
the OS to collect events), rather than using a delay to force a fixed fps.
Fed the length of each frame into the simulation as the duration of the next
frame.

3) Smoothed the sampled frame duration by calculating the average frame
duration, starting with the first frame measured, up to about 1-2 seconds
worth of frames.  This avoided jitter due caused by GC and the OS process
scheduler.

4) Checked for frames with an unusual duration. When a frame was much too
long or too short the timing algorithm didn't use its duration, but instead
restarted the sampling of frame values for the average calculation.  This
avoided sudden jumps when one frame was too long because the user had put
the process to sleep or the game was loading large images between levels.

The final algorithm worked well:
* It stopped the GC making the animation jerky.
* It adapted gracefully to changing system load by dropping the frame rate.
* It smoothed out jittery frames.
* It could adapt to sudden events that caused drastic changes in frame rate.

The downside is that animation code becomes more a little more complex
because you are not sure how much time each frame is going to simulate, and
therefore how far each actor is going to move.  You also have to pass the
frame duration to the animation methods of all the actors in the game.

Cheers,
            Nat.

________________________________
Dr. Nathaniel Pryce
B13media Ltd.
Studio 3a, Aberdeen Business Centre, 22/24 Highbury Grove, London, N5 2EA
http://www.b13media.com

----- Original Message -----
From: "Matthew Bloch" <mattbee@soup-kitchen.net>
Newsgroups: comp.lang.ruby
To: "ruby-talk ML" <ruby-talk@ruby-lang.org>
Sent: Wednesday, May 22, 2002 4:00 PM
Subject: Stymied by Ruby's garbage collector


> Hello;
>
> This one is driving me crazy: I've got an about-to-be-deployed
> entertainment product written in Ruby using SDL for the graphics (through
> RUDL), and I've hit a brick wall with what I believe is the garbage
> collector.
>
> Basically, the drawing loop for a particular screen takes the same time
> every frame but occasionally the garbage collector kicks in and makes the
> whole game lurch for that frame: graphics jump to compensate etc. and this
> is a regular occurrence (once every 2/3 secs).  I've tried turning it off
> during critical sections of animation, but on most systems I've tried, it
> gobbles all the memory before Ruby can start garbage collection again.  I
> know the problem *is* down to the garbage collector's timing because I can
> see smooth animation for a few seconds after GC.disable (where there
wasn't
> before) before the inevitable seizure.
>
> Looking at Ruby's garbage collector (from 1.6.7), it seems an
> 'all-or-nothing' proposition.  That is, the whole algorithm is run at any
> point to free as much memory as possible, or it is not.  There's no
partial
> collection to satisfy what may be a small allocation request.
>
> Now my deadline is pretty tight on this, so I'm after some tips to solve
> this in the short term for now :-)  Various solutions present themselves,
> in order of simplicity:
>
> *) Upgrade the game's runtime to Ruby 1.7 -- does this have a less lumpy
gc
> algorithm?  Or is there an even more advanced version of Ruby around from
> which I could steal just a better gc?
>
> *) Redesign critical parts of the game to stop burning through so much
> damned memory-- but surely this would ruin the maintainability of the
code,
> to have to scope everything as widely as possible?  Or are there other
> techniques I can use for the same end?
>
> *) Hack the gc algorithm used by ruby_xmalloc etc. to stop recovering
> memory after the number of bytes needed is available, rather than running
> the whole algorithm.  I haven't studied the algorith in detail yet, so I
> have no idea whether this is viable;
>
> *) Find the correct places to GC.disable / GC.enable to smooth the more
> visible glitches over-- but I'd be very wary to deploy such a solution
> because it's unpredictable across different systems and in the worst case
> can seize the machine up totally.
>
> *) last resort: leave everything as it is, but deliberately slow animation
> loops to assume the worst case.
>
> Can anyone who's been in a similar situation comment?  I assume this kind
> of problem is endemic to games that rely on garbage collectors, so someone
> must have some opinions :-)
>
> thanks in advance,
>
> --
> Matthew
>

In This Thread