Re: OODL: Goofball OT question

1999-07-23 Thread DeRobertis
At 6:10 PM +1100 on 7/22/99, spierings wrote: You wouldn't let the preprocessor create buggy C Code. If there was an error you would send it from the preprocessor. Chances are, you'd let one by every once and a while, unless you go through a full syntax parse and some other error-checking

Re: OODL: Any array classes

1999-07-23 Thread DeRobertis
At 3:14 PM +0200 on 7/22/99, M. Uli Kusterer wrote: Don't use GetInd calls. Use iterators. Why do the biggest problems always have the simplest solutions? Of course! And one more thing: Have to block file system assign ID's. Leaves it open to improvement.

OODL: Licenses - Fourth World Public License

1999-07-23 Thread Alain Farmer
Uli: Hi folks, just thought for completeness' sake I'll throw this in also. It sounds pretty close to what we want... Alain: Yet another licence! ;-) Uli: ...although I'm not sure we want to restrict distribution of an OpenCard clone that sets out to replace OpenCard. I think we should allow

Re: OODL: Any array classes

1999-07-23 Thread DeRobertis
At 2:18 PM +0200 on 7/23/99, M. Uli Kusterer wrote: Have you thought of using a linked list instead of an array. This way you could have arrays of variable sizes. Andre, I've been doing a variable-size array all the time (just using a malloc()ed pointer as an array). Basically, map and vector

Re: OODL: Re: Report on XBlockFile's Linux Compilation

1999-07-23 Thread DeRobertis
At 3:24 PM +0200 on 7/22/99, M. Uli Kusterer wrote: But, for the record, use 0x4d415020ul for 'MAP '. I've created the following macro for types: #define BLOCK_TYPE( a,b,c,d ) ((unsigned long)(d | (c 8) | (b 16) | (a 24))) Which should solve the problem. I use this whenever

Re: OODL: Any array classes

1999-07-23 Thread DeRobertis
At 3:30 PM +0200 on 7/22/99, M. Uli Kusterer wrote: I bet I can write a faster block file system :) Anthony, why don't you tell me what slow techniques I'm using and how to replace them with faster ones? This would maybe now take more time, but as I'd learn techniques to speed up my programs,

OODL: OC Licence - Copyright Holder

1999-07-23 Thread Alain Farmer
Alain: YES and no. There could be some bickering. A difference of opinion concerning the importance of the contribution made by someone that wishes to be cited as one of the authors. Anthony: Alain, in that event we would have the list decide on what to do. Alain: OK, but what kind of

OODL: XBF Linux, Take Two

1999-07-23 Thread Anthony DeRobertis
Bugs: o If you forget to call Open(), it segfaults o No data writeout until compact? o Inserting 100,000 blocks is horribly slow (I know, you're fixing it) -- it's still running, now at 17 minutes, and only 52,000 blocks...39:00 - 79,000 blocks...

Re: OODL: Any array classes

1999-07-23 Thread Yennie
Well, my first suggestion would be not to do a linear search for free space. Keep all the free blocks in a balanced tree, sorted by size. You can then quickly search for a block of the propper size. Just as a note for Uli if you do this, you can just use a map with the block size as the key.

Re: OODL: Any array classes

1999-07-23 Thread DeRobertis
At 10:06 PM +0200 on 7/23/99, M. Uli Kusterer wrote: And one more thing: Have to block file system assign ID's. Leaves it open to improvement. Anthony, English, please. What are you trying to say? The block file code should assign ID's -- it can no doubt most efficently alloocate them, and

Re: OODL: OC Licence - Copyright Holder (long)

1999-07-23 Thread DeRobertis
At 6:36 PM -0400 on 7/23/99, Alain Farmer wrote: Alain: YES and no. There could be some bickering. A difference of opinion concerning the importance of the contribution made by someone that wishes to be cited as one of the authors. Anthony: Alain, in that event we would have the list decide on

Re: OODL: Goofball OT question

1999-07-23 Thread DeRobertis
At 10:15 PM +0200 on 7/23/99, M. Uli Kusterer wrote: It will be very big -- something on the order of ten to twenty times faster, unless you've got some _great_ optimizer -- i.e., one far better than I've seen! Doesn't OTVar keep its values as some preferred value? Then there shouldn't be that

Re: OODL: Goofball OT question

1999-07-23 Thread M. Uli Kusterer
It will be very big -- something on the order of ten to twenty times faster, unless you've got some _great_ optimizer -- i.e., one far better than I've seen! Doesn't OTVar keep its values as some preferred value? Then there shouldn't be that much of a problem. Just a couple of branches more.

Re: OODL: Any array classes

1999-07-23 Thread M. Uli Kusterer
And one more thing: Have to block file system assign ID's. Leaves it open to improvement. Anthony, English, please. What are you trying to say? Cheers, -- M. Uli Kusterer http://www.weblayout.com/witness 'The

Re: OODL: Any array classes

1999-07-23 Thread M. Uli Kusterer
Have you thought of using a linked list instead of an array. This way you could have arrays of variable sizes. Andre, I've been doing a variable-size array all the time (just using a malloc()ed pointer as an array). Basically, map and vector work in a similar way as your list, they're just a