1. Contest details updated
I've updated the contest page with links to further details about
each submission.
I've still got submissions from the last 24-hours to process yet and
they are not on the contest page yet.
That's what I'm doing now....
--
Derek Parnell
Melbourne, Australia
2. Re: Contest details updated
Derek Parnell wrote:
>
> I've updated the contest page with links to further details about
> each submission.
If you get a spare moment, any chance of a hint as to why
pgrcom2 counts files 7 & 10 wrong?
Many thanks for organising the contest - I've enjoyed it immensely so far
(and most of the month still to go!). At the very least, it seems to have
stopped the various flame wars that were going on...
Regards,
Phil
3. Re: Contest details updated
Phil Russell wrote:
>
> Derek Parnell wrote:
> >
> > I've updated the contest page with links to further details about
> > each submission.
>
> If you get a spare moment, any chance of a hint as to why
> pgrcom2 counts files 7 & 10 wrong?
File 7 does not end with a line-end character. The very last character
in the file is part of a token. So if you program only gather's tokens
before EOF is found, it misses out on the last token.
File 10 is just random bytes. Many are greater than #7F. Your output
is ...
Total: 1, Unique: 1
01 I 1
01 1
Elapsed time: 0.000000
but the acceptable output is either a message stating that the file contains
invalid bytes, or a valid attempt to count tokens. ...
Total: 1563, Unique: 399
01 V 59
01 1152
02 304
03 73
04 21
05 12
06 1
Elapsed time: 0.640000
> Many thanks for organising the contest - I've enjoyed it immensely so far
> (and most of the month still to go!).
There is still some optimisation left. I have a submission from Jiri Babor
and it is very, very, slick.
>At the very least, it seems to have
> stopped the various flame wars that were going on...
And people call me mad 
It also gives a break to RDS - I'm anxious
for v2.5 and Robert doesn't need to many problems just now.
--
Derek Parnell
Melbourne, Australia
4. Re: Contest details updated
Derek,
you wrote: "File 10 is just random bytes. Many are greater than #7F..."
But the rule 13 states: "... The file will only contain bytes in the range #00 -
#7F..."
This is manifestly unfair!
jiri
5. Re: Contest details updated
jiri babor wrote:
>
> you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> But the rule 13 states: "... The file will only contain bytes in the range #00
> - #7F..."
>
> This is manifestly unfair!
jiri, he also wrote:
"If anyone believes that these rules and conditions
are unfair, then don't enter the contest."
LOL! What a predicament. :)
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
6. Re: Contest details updated
jiri babor wrote:
>
> Derek,
>
> you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> But the rule 13 states: "... The file will only contain bytes in the range #00
> - #7F..."
>
> This is manifestly unfair!
Probably, but I believe that the only place it will hurt you is in the
subjective style section, where one of the things he's looking for is
"Defensive coding that is tolerant of bad data." I don't think anyone
has gotten a 10 yet, and probably noone will...
Matt Lewis
7. Re: Contest details updated
- Posted by jiri babor <jbabor at paradise.net.nz>
Nov 04, 2004
-
Last edited Nov 05, 2004
Matt Lewis wrote:
>
> jiri babor wrote:
> >
> > Derek,
> >
> > you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> > But the rule 13 states: "... The file will only contain bytes in the range
> > #00 - #7F..."
> >
> > This is manifestly unfair!
>
> Probably, but I believe that the only place it will hurt you is in the
> subjective style section, where one of the things he's looking for is
> "Defensive coding that is tolerant of bad data." I don't think anyone
> has gotten a 10 yet, and probably noone will...
>
> Matt Lewis
>
Call me crazy, but I don't include an anti-radiation suit in my luggage every
time I go on holidays in Fiji...
jiri
8. Re: Contest details updated
- Posted by Patrick Barnes <mrtrick at gmail.com>
Nov 04, 2004
-
Last edited Nov 05, 2004
On Thu, 04 Nov 2004 07:46:18 -0800, jiri babor <guest at rapideuphoria.com>
wrote:
> you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> But the rule 13 states: "... The file will only contain bytes in the range #00
> - #7F..."
>
> This is manifestly unfair!
Files 6-10 are not scored, just used to find a style score (It's much
easier to just run some files with border conditions than look at the
code and try and figure whether it handles them)
It's alright, mine parses file 10 perfectly now... it's
thump-your-head-on-the-table-simple...
--
MrTrick
9. Re: Contest details updated
- Posted by Derek Parnell <ddparnell at bigpond.com>
Nov 04, 2004
-
Last edited Nov 05, 2004
jiri babor wrote:
>
>
> Derek,
>
> you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> But the rule 13 states: "... The file will only contain bytes in the range #00
> - #7F..."
>
> This is manifestly unfair!
Yep, absolutely unfair. No argument there.
However, its always a reasonable expectation that programs should be able to
handle
bad data. If a submission fails to deal with this type of bad file, it will
cost you at most one point out of 460. Not a big penalty. Certainly not
as big a penalty as one would get in real life coding.
--
Derek Parnell
Melbourne, Australia
10. Re: Contest details updated
- Posted by Derek Parnell <ddparnell at bigpond.com>
Nov 04, 2004
-
Last edited Nov 05, 2004
jiri babor wrote:
>
> Matt Lewis wrote:
> >
> > jiri babor wrote:
> > >
> > > Derek,
> > >
> > > you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> > > But the rule 13 states: "... The file will only contain bytes in the range
> > > #00 - #7F..."
> > >
> > > This is manifestly unfair!
> >
> > Probably, but I believe that the only place it will hurt you is in the
> > subjective style section, where one of the things he's looking for is
> > "Defensive coding that is tolerant of bad data." I don't think anyone
> > has gotten a 10 yet, and probably noone will...
> >
> > Matt Lewis
> >
>
> Call me crazy, but I don't include an anti-radiation suit in my luggage every
> time
> I go on holidays in Fiji...
Okay. You are crazy
--
Derek Parnell
Melbourne, Australia
11. Re: Contest details updated
On Thu, 04 Nov 2004 07:46:18 -0800, jiri babor
<guest at RapidEuphoria.com> wrote:
>Derek,
>
>you wrote: "File 10 is just random bytes. Many are greater than #7F..."
>But the rule 13 states: "... The file will only contain bytes in the range #00
>- #7F..."
>
>This is manifestly unfair!
He also just listed the files used in the tests, which he previously
promised would not be available.
By naming them, so that those who have, have, and those that do not,
do not, is far more unfair in my book.
I have no idea why Derek did this, on this I too cry foul.
Now that you have named them, it may be too late. Either make them
available for fairness sake, or change them.
Regards,
Pete
PS I'm currently ranked 2nd, behind Phil Russel, OK, but Ricardo Forno
beats both pants down, and should be 1st (making me 3rd).
12. Re: Contest details updated
Pete Lomax wrote:
>
> On Thu, 04 Nov 2004 07:46:18 -0800, jiri babor
> <guest at RapidEuphoria.com> wrote:
>
> >Derek,
> >
> >you wrote: "File 10 is just random bytes. Many are greater than #7F..."
> >But the rule 13 states: "... The file will only contain bytes in the range
> >#00 - #7F..."
> >
> >This is manifestly unfair!
>
> He also just listed the files used in the tests, which he previously
> promised would not be available.
I said that the FILES would not be available. I said nothing about the
NAMES of the files or what those files might contain.
> By naming them, so that those who have, have, and those that do not,
> do not, is far more unfair in my book.
??? What ??? I am the only person to have access to those files.
They are not on my web site. I've just named them and outlined the
sort of data that they might contain.
> I have no idea why Derek did this, on this I too cry foul.
I still can't see why giving people the names of these files is somehow
unfair. Everybody now has the names - so what?! No one has the files
themselves.
> Now that you have named them, it may be too late. Either make them
> available for fairness sake, or change them.
Make the FILES available? Is that what you're saying? Or are you saying
that I should rename the files THAT EXIST ON MY LOCAL MACHINE ONLY?
You cannot recreate the test files I used from external sources because
I've edited them for my own internal purposes. Simply put, the files
I'm using are unique and no one has access to them.
However, to sooth any ruffled feathers, I'll edit them further to absolutely
make sure that they can't be gotten from other sources, then re-run all the
submissions again.
> Regards,
> Pete
> PS I'm currently ranked 2nd, behind Phil Russel, OK, but Ricardo Forno
> beats both pants down, and should be 1st (making me 3rd).
I'm sure Ricardo will submit an amended version that addresses the 'style'
issue I raised. When he does, and if that doesn't torment his algorithm,
I can see him going to first place. However, I've got about 8 or 9 new
submissions that I'm putting on the results page today, so anything can
happen
I've automated the HTML generation now and now it's a *lot* faster to
get updates to you all.
--
Derek Parnell
Melbourne, Australia
13. Re: Contest details updated
Pete Lomax wrote:
>
>
> He also just listed the files used in the tests, which he previously
> promised would not be available.
>
> By naming them, so that those who have, have, and those that do not,
> do not, is far more unfair in my book.
>
> I have no idea why Derek did this, on this I too cry foul.
>
> Now that you have named them, it may be too late. Either make them
> available for fairness sake, or change them.
>
Methinks you are taking all this a wee bit too seriously... I'm sure no one can
recreate the files EXACTLY as he's got them, esp. "random bytes". I guess the
cat is out of the bag on the "empty file" though...
14. Re: Contest details updated
- Posted by Pete Lomax <petelomax at blueyonder.co.uk>
Nov 05, 2004
-
Last edited Nov 06, 2004
On Thu, 04 Nov 2004 17:25:27 -0800, Derek Parnell
<guest at RapidEuphoria.com> wrote:
>??? What ???
<snip> Sorry, I probably sounded a little harsh, or it was frustration
not common sense talking. Ignore it, it wasn't significant.
Pete
15. Re: Contest details updated
- Posted by Tone Škoda <tskoda at email.si>
Nov 05, 2004
-
Last edited Nov 06, 2004
isn't this a contest of who is the best expert in hash tables?
16. Re: Contest details updated
- Posted by Derek Parnell <ddparnell at bigpond.com>
Nov 05, 2004
-
Last edited Nov 06, 2004
Tone Škoda wrote:
>
> isn't this a contest of who is the best expert in hash tables?
Mostly, but even though all but one submission is using hash tables,
a number are getting the wrong results.
--
Derek Parnell
Melbourne, Australia
17. Re: Contest details updated
Derek Parnell wrote:
>
> Mostly, but even though all but one submission is using hash tables,
> a number are getting the wrong results.
Is mine using a hash table?
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
18. Re: Contest details updated
How often are you going to update the results? Also, have you received my
_v6 entry?
Just checkin'!
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
19. Re: Contest details updated
cklester wrote:
>
> Derek Parnell wrote:
> >
> > Mostly, but even though all but one submission is using hash tables,
> > a number are getting the wrong results.
>
> Is mine using a hash table?
>
> -=ck
> "Programming in a state of EUPHORIA."
> <a
> href="http://www.cklester.com/euphoria/">http://www.cklester.com/euphoria/</a>
>
What's a hash table?
Chris
http://members.aol.com/chriscrylex/euphoria.htm
http://uboard.proboards32.com/
20. Re: Contest details updated
Chris Burch wrote:
> cklester wrote:
> > Derek Parnell wrote:
> > > Mostly, but even though all but one submission is using hash tables,
> > > a number are getting the wrong results.
> > Is mine using a hash table?
> What's a hash table?
Okay, maybe it's not me. :D
http://ciips.ee.uwa.edu.au/~morris/Year2/PLDS210/hash_tables.html
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
21. Re: Contest details updated
On 5 Nov 2004, at 16:18, cklester wrote:
>
>
> posted by: cklester <cklester at yahoo.com>
>
> How often are you going to update the results? Also, have you received my
> _v6 entry?
Ah, a new version every day? But Derek gets no sleep, won't he be just as
likely to count that against you?
Kat,
preparing her _v13 submission too.
22. Re: Contest details updated
Kat wrote:
>
> On 5 Nov 2004, at 16:18, cklester wrote:
>
> >
> > posted by: cklester <cklester at yahoo.com>
> >
> > How often are you going to update the results? Also, have you received my
> > _v6 entry?
>
> Ah, a new version every day? But Derek gets no sleep, won't he be just as
> likely to count that against you?
>
> Kat,
> preparing her _v13 submission too.
>
>
Gee whiz, I only made one version. I've been tweaking it but I can only seem to
improve the speed 1% or so without rethinking the whole algorithm. The problem
with optimizations is that they tend to uglify the code and I would probably lose
some style points. I'm waiting to see how I stack up with v1.0 before I try a
totally new tack. I'm not even sure it works correctly...
23. Re: Contest details updated
Kat wrote:
> On 5 Nov 2004, at 16:18, cklester wrote:
> > posted by: cklester <cklester at yahoo.com>
> > How often are you going to update the results? Also, have you received my
> > _v6 entry?
> Ah, a new version every day? But Derek gets no sleep, won't he be just as
> likely to count that against you?
I would hope not, because his rule 17 clearly states:
"You are welcome to submit as many programs as you like."
-=ck
"Programming in a state of EUPHORIA."
http://www.cklester.com/euphoria/
24. Re: Contest details updated
On 5 Nov 2004, at 17:16, Chris Burch wrote:
<snip>
> What's a hash table?
Well, think of it this way: a large barn, with the tokens of the
contest written
seemingly randomly all over it. Take your farm breakfast outside, put
on a
blindfold, and toss an egg at the barn. Now what you don't know is some
aliens did some complicated math to determine the trajectory of each
throw,
while eating the hash browns, so after that first toss, every time you
want to
hit that same word, your toss will hit it. So scrape off that
splattered egg and
put it in a bucket (these are called buckets). All the tokens that egg
hit are
scraped off into the bucket too, they were painted on the barn with cheap
paint for this purpose. Put the bucket on a table, and peek into it.
If the
aliens do the math right, no token in that bucket will appear in any
other
bucket, and there's no fragments (altho a token fragment could be
there, it
likely isn't a fragement of a complete token in that bucket). Ideally,
and the
part NASA is peeved about, is you can't take the location on the barn
wall of
a particular token, and determine the math it will take to land it in
any one of
the buckets you'll end up with on the table! Only if you know the math
the
aliens used for your particular hash table can it be done, and if they
use big
enough numbers there is only one token per bucket, and if they use small
numbers in the math then many tokens will be hit per egg throw. And
this is
why it's called "hash", it's a slick greasy method to put eggs in
particular
buckets, and the how-to is about as clear as the hash. Feed to the pigs
when done, and wash that barn!
Kat
</joke>
25. Re: Contest details updated
On Fri, 05 Nov 2004 19:13:14 -0800, Andy Serpa
<guest at RapidEuphoria.com> wrote:
>I've been tweaking it but I can only seem to improve the speed 1%
> or so without rethinking the whole algorithm.
Ditto. I spent three days getting nowhere, then suddenly breakthrough
came. Sadly, it appears that meanwhile, the same or better
breakthrough occurred in Brazil as well )
Regards,
Pete
