[rc5] Stats solution? (long)

Eric Gindrup gindrup at okway.okstate.edu
Tue Nov 11 09:37:00 EST 1997


     Perhaps it was somewhat a troll.  But it was more a statement of 
     perception.  I'm not attacking anyone's personal life.  I'm attacking 
     their appearance of involvement, and I think I said so.
     
     You want a spec. ...
     Win16, identical interface with Win32 GUI, identical network 
     interfaces.  Preferably before the RC5-64 effort is a month old.  I 
     expect this will be pointless, because
     1) All the users who would have run it will be gone before it happens.
     2) The heavy duty perproxy network won't have any traffic to carry. 
     and
     3) The coders are busy developing clients that maybe a couple of dozen 
     users will run.
     And further, you suggest that I write it.  I don't see that I get 
     anything out of that as I can't attach it to your proprietary proxy 
     network.  I have made it perfectly clear that with some sort of proxy 
     spec I would have been *more* than willing to have gotten a couple of 
     clients out the door.
     The claim has long been: "Write the thing and then maybe we'll let you 
     into our circle."  But now, you say, the delay in v3 clients "is 
     proper planning before the first line of code is written."  Notice 
     that these two statements are strongly at odds.  You want us to write 
     spec-less code while waiting for you to write completely spec-ed code.
     
     How much time has been wasted on low-volume clients?  How many of 
     these users could have been kept with the much-maligned Java client?  
     Why is there no appearance of giving a damn?
     
     And to address your v3 claims:  v3 specs are 4 to 6 weeks away.  Code 
     won't happen until after that.  And, since you ask, I've been toodling 
     around with an RC5-56 Java client.  Since it's almost entirely network 
     and core code, it's as cross-platform as the JVM implementors will let 
     it be and the basic property of requesting Java cores from the server 
     makes it multifunctional.  It's not done.  Why?  No motivation.  
     Bovine is doing nothing for me but blowing smoke at me.
     
     <cough> <cough> moo.
     
        Now for your trolls:  I believe that Microsoft will innovate a 
     whole new slew of crippling bugs and trivial logic errors in Word98.
        Judging by the amount of "when we finish this" traffic, I can't 
     believe that the end wasn't in most people's minds when the key was 
     found.  At the point the key was found, there was a greater than 2% 
     chance per day that the key would be found.
        Let's assume that the key hadn't been found.  The keyspace was 
     estimated to be exhausted less than three weeks from now (because 
     participation had not topped out).  v3 specs aren't going to be done 
     by then.  Why would they have been if the key had not been found?
            -- Eric Gindrup ! gindrup at okway.okstate.edu


______________________________ Reply Separator _________________________________
Subject: Re: [rc5] Stats solution? (long) 
Author:  <rc5 at llamas.net > at SMTP
Date:    1997/11/11 06:39

     
was this an obvious troll?
     
You are pretty much incorrect in almost all of your assumptions.  Version 
3 clients are happening, but what slows the developement cycle is proper 
planning before the first line of code is written.  you develope a cross 
platform, multifunctional application right now... sit down, own vi and 
get cranking. see how far you get.
     
[snip]
     
The conversion from RC5-56 to R5-64 was a shoe-in.  We weren't ready for 
RC5-56 to come to an end.  There was a fair amount of stress over getting 
something together to keep people involved.  Just like you said, we 
tweaked up a little code, and bang! we got an rc5-64 client/proxy network 
going.  This doesn't mean we are lazy, it means we are mazimizing the 
resources at hand, on all levels.
     
1. at the client level we want to keep users involved.
2. at the proxy level we wanted the current core proxies to upgrade, and 
also recruit more.  we expect this to get much bigger than it is now.
3. at the coder level, we needed to maximize time and efficiency.
     
        Adam has let his personal life *suffer* to make this contest
happen.  *DO NOT* abuse him for what he has or hasn't done.  To make the 
most efficient time of software writers time, you must first know what you 
want him to write. How do you do that?  explain in detail, and on paper, 
everything you want the software to do from the inside out.  Document 
packet sizes, expected behaviors, modularity, etc.  Then after you have 
the utmost clear picture of what you want, you write it.
     
you can complain about the said 'Inner Circle' and thier failure in 
communication, but do not attack our whole hearted devotion to this 
project.
     
Chipper
Chris Chiapusio
     [snip]
 On Tue, 11 Nov 1997, Eric Gindrup wrote:
     [snip -- a whole lot]


----
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.



More information about the rc5 mailing list