[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-ha-dev
Subject:    Re: [Linux-ha-dev] Re: Availability resource management for Linux
From:       Michael Moerz <e9625136 () student ! tuwien ! ac ! at>
Date:       2000-10-13 12:39:43
[Download RAW message or body]

Hi!

On Fri, 13 Oct 2000 TEREKHOV@de.ibm.com wrote:

> Michael,
> 
> > When I am now going to show you my own research to
> > what extend is my research then protected?
> 
> i am not a lawyer but i would expect that w/o patent
> your released research gets NO protection.
> 
I see. As it just was something I was thinking about I wanted to ask.

> > I am actually releasing all my stuff under GNU GPL.
>
That just means that my own writting is protected. The ideas behind
are, well, not protected by GNU GPL. So if someone takes  what I have
written and extends  it he  will have  to release it again under GNU GPL.
 
> which protects your CODE from being used in proprietary
> products. I am not so sure about METHODS/MODELS/..
> (subject of patenting).
> 
not only code can be protected by GPL, but written  docu too. I know  that
in fact cause I am related to GNU debian linux where currently a
discussion is going on about documentation that is not GPL'ed.            

> > Actually I don't even believe that it was so a good
 > > idea to post that stuff right here.
> 
> could you please explain why?
> 
First you caused a lot of traffic by posting such long things to this
list. Perhaps not everybody here was interested in that topic. I think a
better solution might have been to put it up somewhere in the web and post
the url (I still would have read it cause I was simply curious about it). 
That would have had the tremendous side-effect that people who weren't so
interested in the topic would simply be not faced with such huge mails.
Second is only a personal issue that I did state before and I don't think
that it so grave important as the first point. Actually I believe it's
importance to be near zero. I just have to think about licening issues;
thinking a bit more about it, I would say I would have had to think about
it anyway.
I am actually not sure to what degree such licensing issues apply in
Europe, but I think that I also have to cope with USA standards, cause I
want the people there to be able to use my products too (there will
be a software created based on what I am thinking all time).

> i have an impression that you are not happy with
> "disclamer that beheads the document suggest
> that everything is the property of IBM"
> 
> well, it is true that IBM have patented some
> things related to HA. Please understand that,
> being a developer I am not in position to
> comment on these issues.. But i would like
> to point out that IBM has a good track of
> records with releasing many technologies
> to OSS community (JFS, Java compiler, etc.)
> So, make your case and I promise that I'll
> forward your views to the management but I
> am not a right person to talk about various
> intellectual property issues, business
> models, etc.
> 
I understand this and your possition regarding the whole topic.
Now I am really curious about what patents IBM holds for HA.
Where can I get information about this and what should I do if
something of my design is related to a patent of IBM? 
Whom I may contact then?

> well, i'll try to give you an example.
> assume you have an application A (process)
> which needs about a minute to initialize
> itself (a lot of disk IO and network IO,
> e.g. registration/config-load over internet,
> etc.. ) it means that such application becomes
> up (AVAILABLE) with delay of 1 minute
> after it has been STARTED.
> 
> now, assume you have another application B
> (process) which as a part of its
> initialization checks whether process A
> has been STARTED and if not it just terminates
> (or tries to start another instance of A which
> is not what we want).
> 
> without STARTED status automation would
> have to wait until A becomes AVAILABLE
> (one minute more) before starting B.
> 
> Having STARTED status solves the problem
> of that delay. Which means less down time
> (for B) and MORE AVAILABILITY.
> 
> > I also think that this state cannot be
> > determined by just watching the resource
> > by a monitor-prog. That would mean that
> > the resource has to be adopted to use
> > some sort of API.
> 
> SOFTDOWN
>   STARTING
>     AVAILABLE
>   STOPPING
> SOFTDOWN
>
that's abstractly viewed similar to my own state model. Except that
I wasn't thinking about the issue that state starting resolves.
But thank you for pointing me out to this.
 
> well, i guess you misunderstood me. i am not
> talking about 'adoption' or instrumentation API.
> What I am talking about is an API which would
> enable sharing of resource monitor and control
> 'modules' (e.g. PID-monitor) between many HA
> products.
>
The question on that topic will be if everybody producing
HA-products would agree on a common API. I think that will be
shown to us by near/distant future.
 
--
kind regards,
Michael Moerz


_______________________________________________________
Linux-HA-Dev: Linux-HA-Dev@lists.tummy.com
http://lists.tummy.com/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/

[prev in list] [next in list] [prev in thread] [next in thread]