inprise_os_compared_pgsql

This is part of The Pile, a partial archive of some open source mailing lists and newsgroups.



Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source 
From: Joe Brenner <doom@kzsu.stanford.edu>
Date: Tue, 25 Jul 2000 15:40:30 -0700

Kaare Rasmussen <kar@webline.dk> wrote: 

> In case you didn't notice:
> 
> http://biz.yahoo.com/prnews/000725/ca_inprise_2.html

Yes, and now the story is up on slashdot: 

http://slashdot.org/comments.pl?sid=00%2F07%2F25%2F1439226&cid=&pid=0&startat=&threshold=-1&mode=nested&commentsort=3&op=Change

So if anyone feels like taking time out to do some
evangelizing, this might be a good moment to say a few words
about Postgresql vs.  "Inprise, The Open Source Database".  

In particular, this guy could use a detailed reply: 

> Interbase
> 
>  (Score:4, Interesting)
>  by jallen02 (:-( .) on Tuesday July 25, @09:06AM PST
>  (User #124384 Info) http://gdev.net/~jallen
>
>  I found myself wondering exactly what Interbase could do
>  for me.  So I dug through their site (not hard to find)
>  and found  this lil gem
>
>  Interbase Product Overview 
>
>  Interbase has some very awesome features. The overview
>  took the tone of a semi marketing type item yet it was
>  infomrative and if you read through some of the garbage
>  its rather clear to see as a programmer/developer what
>  Interbase offers.

>  Some of the features that stuck out in my mind from the
>  over view.
> 
>  -Small memory footprint
>  -Triggers 
>  -Stored Procedures
>  -User Definable Functions with some 'libraries' per say
>      already defined for math and string handling
>  -Alert events 
>
>  EX:A certain item goes below xyz dollars it can send an
>  alert using some sort of constant polling method. I am
>  not sure exactly what this one was.. but basically it
>  looks like whenever changes are done to the table if
>  certain criteria are met it can call up a stored
>  proc/UDF or something. This is a bit more powerful than
>  a trigger or a stored procedure since you do not have to
>  do any speical coding on a insert/update/delete.
>
>  Some other interesting things... There was a *LOAD* of
>  case studies on the interbase site.
>
>  Case Studies 
> 
>  I looked at some of these and they were real industry
>  proven case studies IMO. 
> 
>  Its Free.. and it has a good reputation 
> 
>  You can buy support for it 
> 
>  It appears to be VERY ANSI Compliant and supports all the
>  trappings of MS SQL Server.
> 
>  It also claimed to be self optimizing... anyways hope this
>  provided a little information. 

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open
From: Don Baccus <dhogaza@pacifier.com>
Date: Tue, 25 Jul 2000 15:54:03 -0700

At 03:40 PM 7/25/00 -0700, Joe Brenner wrote:
>
>Kaare Rasmussen <kar@webline.dk> wrote: 
>
>> In case you didn't notice:
>> 
>> http://biz.yahoo.com/prnews/000725/ca_inprise_2.html
>
>Yes, and now the story is up on slashdot: 
>
>http://slashdot.org/comments.pl?sid=00%2F07%2F25%2F1439226&cid=&pid=0&start
at=&threshold=-1&mode=nested&commentsort=3&op=Change
>
>So if anyone feels like taking time out to do some
>evangelizing, this might be a good moment to say a few words
>about Postgresql vs.  "Inprise, The Open Source Database".  
>
>In particular, this guy could use a detailed reply: 

Just be sure to avoid any talk of outer joins ...


===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source
From: Chris Bitmead <chrisb@nimrod.itg.telstra.com.au>
Date: Wed, 26 Jul 2000 14:30:33 +1000

$ find InterBase  -name \*.c |xargs cat |wc
  481977 1417203 11430946
$ find postgresql-7.0.2 -name \*.c |xargs cat |wc
  329582 1087860 8649018

$ wc InterBase/dsql/parse.y
    4217   13639  103059 InterBase/dsql/parse.y
$ wc postgresql-7.0.2/src/backend/parser/gram.y
    5858   20413  149104 postgresql-7.0.2/src/backend/parser/gram.y

I downloaded it, just to have a poke around. It doesn't build very
easily, if at all. The best way I can describe the source is that it is
dry reading. Not much interesting commentary. Very big source files,
with very long functions and huge case statements is how I would
characterise it. I suspect it's reasonably well written but not very
extensible in nature. I don't think this is going to set the world on
fire anytime soon.

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source
From: Thomas Lockhart <lockhart@alumni.caltech.edu>
Date: Wed, 26 Jul 2000 05:21:06 +0000

$ wc InterBase/dsql/parse.y
>     4217   13639  103059 InterBase/dsql/parse.y
> $ wc postgresql-7.0.2/src/backend/parser/gram.y
>     5858   20413  149104 postgresql-7.0.2/src/backend/parser/gram.y

Hmm. I suspect that I could shrink our gram.y by ~25% just by removing
comments and C support routines, and by consolidating some execution
blocks onto fewer lines. Does it look like their parse.y is more dense
than ours, do they do a lot of postprocessing to eliminate the yacc
rules, or have we missed the boat on writing the grammar in yacc?

Just curious; I probably won't look myself since I don't want to run the
risk of compromising our code and licensing. Or is that not an issue
with the Inprise license?

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source
From: Chris Bitmead <chrisb@nimrod.itg.telstra.com.au>
Date: Wed, 26 Jul 2000 15:56:12 +1000

Thomas Lockhart wrote:
> 
> > $ wc InterBase/dsql/parse.y
> >     4217   13639  103059 InterBase/dsql/parse.y
> > $ wc postgresql-7.0.2/src/backend/parser/gram.y
> >     5858   20413  149104 postgresql-7.0.2/src/backend/parser/gram.y
> 
> Hmm. I suspect that I could shrink our gram.y by ~25% just by removing
> comments and C support routines, and by consolidating some execution
> blocks onto fewer lines. Does it look like their parse.y is more dense
> than ours, do they do a lot of postprocessing to eliminate the yacc
> rules, or have we missed the boat on writing the grammar in yacc?
> 
> Just curious; I probably won't look myself since I don't want to run the
> risk of compromising our code and licensing. Or is that not an issue
> with the Inprise license?

I had a bit of a look. There's no obvious reason, just maybe postgres
has a few more comments and a bit more code inside the action blocks. No
obvious problem here.

It would be a pity if we can't look and learn from Interbase in this
instance, because this is one area where there is at least a possibility
of borrowing something useful.

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source 
From: Joe Brenner <doom@kzsu.stanford.edu>
Date: Thu, 27 Jul 2000 13:44:52 -0700

Chris Bitmead <chrisb@nimrod.itg.telstra.com.au> wrote: 

Thomas Lockhart <lockhart@alumni.caltech.edu> wrote: 

> > Just curious; I probably won't look myself since I don't want to run the
> > risk of compromising our code and licensing. Or is that not an issue
> > with the Inprise license?
> 
> I had a bit of a look. There's no obvious reason, just maybe postgres
> has a few more comments and a bit more code inside the action blocks. No
> obvious problem here.
> 
> It would be a pity if we can't look and learn from Interbase in this
> instance, because this is one area where there is at least a possibility
> of borrowing something useful.

Well, the license is just the Mozilla Public License with
the names changed.  I've just read through it several times,
and I think the main trouble with it is you probably really 
do need to have a lawyer look at it... but I think you could
go as far as to include some of the Inprise source files 
into postgresql:

http://www.inprise.com/IPL.html

   3.7. Larger Works. 
   
   You may create a Larger Work by combining Covered Code with
   other code not governed by the terms of this License and
   distribute the Larger Work as a single product. In such a
   case, You must make sure the requirements of this License
   are fulfilled for the Covered Code.

The requirements seem to be pretty commonsense things... 
If you use some source code from Inprise, you've got to 
keep track of where the source came from, label it with
their license, list any modifications you've made, always
provide the source with any executables, etc.

There's also a bunch of stuff about how this license doesn't
release you from any third party intellectual property
claims (duh! Legal docs always seem to state the obvious at
great length).  I might wonder what would happen if Borland
owned a software patent on some algorithm that's included in
this code...

But no, I *think* that's a non-issue: 

   The Initial Developer hereby grants You a world-wide,
   royalty-free, non-exclusive license, subject to third
   party intellectual property claims: 

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source
From: Bruce Momjian <pgman@candle.pha.pa.us>
Date: Fri, 28 Jul 2000 11:33:15 -0400 (EDT)

http://www.inprise.com/IPL.html
> 
>    3.7. Larger Works. 
>    
>    You may create a Larger Work by combining Covered Code with
>    other code not governed by the terms of this License and
>    distribute the Larger Work as a single product. In such a
>    case, You must make sure the requirements of this License
>    are fulfilled for the Covered Code.
> 
> The requirements seem to be pretty commonsense things... 
> If you use some source code from Inprise, you've got to 
> keep track of where the source came from, label it with
> their license, list any modifications you've made, always
> provide the source with any executables, etc.

But the BSD license doesn't require source for distributed binaries. 
Sounds like a GPL-style restriction.

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source
From: Kaare Rasmussen <kar@webline.dk>
Date: Sat, 29 Jul 2000 12:13:33 +0200

The requirements seem to be pretty commonsense things... 
> > If you use some source code from Inprise, you've got to 
> > keep track of where the source came from, label it with
> > their license, list any modifications you've made, always
> > provide the source with any executables, etc.
> 
> But the BSD license doesn't require source for distributed binaries. 
> Sounds like a GPL-style restriction.

What is more important to my mind is if the license permits
a developer to look at the code and get inspired, or if the
developer's mind will be "tainted" just by looking.

I hope someone can tell; I always wake up later with my head
on the keyboard when I try to read license stuff...

===

Subject: Re: [HACKERS] Inprise InterBase(R) 6.0 Now Free and Open Source 
From: Tom Lane <tgl@sss.pgh.pa.us>
Date: Sat, 29 Jul 2000 11:43:13 -0400

Kaare Rasmussen <kar@webline.dk> writes:
> What is more important to my mind is if the license permits a
> developer to look at the code and get inspired, or if the developer's
> mind will be "tainted" just by looking.

It is not possible to be "tainted" by looking.  There are only two kinds
of intellectual property rights (at least in the USA) and neither one
creates that risk:

1. Copyright means you can't take the code verbatim, just like you can't
plagiarize a novel.  You can use the same ideas (plot, characters, etc)
but you have to express 'em in your own words.  Structure the code
differently, use different names, write your own comments, etc, and
you're clear even if you lifted the algorithm lock stock & barrel.

2. Patent means you can't use the algorithm.  However, looking doesn't
create extra risk here, because you can't use a patented algorithm
(without paying) no matter how you learned of it --- not even if you
invented it independently.

As far as I've heard, Inprise isn't claiming any patent rights in
connection with the Interbase code anyway, but it might be a good idea
for someone to check before we all start reading their code...

===


the rest of The Pile (a partial mailing list archive)

doom@kzsu.stanford.edu