|
Message
From: Rudolf Usselmann <rudi@a...>
Date: Thu, 16 Oct 2003 10:41:26 +0700
Subject: Re: [oc] Erasing INTERNET (Maxim Vlasov)
On Wed, 2003-10-15 at 18:07, Graham Seaman wrote:
> Hi,
>
> Since I don't have any projects on opencores, I normally try just to lurk
> on this list.. but sometimes I get really tempted to reply. I hope what
> I've written below isn't out of line...
>
>
> On Wed, 15 Oct 2003 antti@c... wrote:
>
> > trogal. Dlja tebja spezialno: za zanovesom OC mozhno dengi delat, no k
> > etomu delu nado nemnoshko podrugomy poghodit. Esli hotches bolsche
> > uznat pishi mnje prjoma, posmotrju shto mozhno delat.
>
> Maksim ochen horosho pishet po-angliskii; pochemy nada pishat' po-russkii?
Thats what I was wondering about as well - sorry to lazy to
write in Russian on an English keyboard ... :*)
I think Antti was hoping that none-of use will understand. Hmm,
Should I translate it to English and post it ? :*) Hay hi to
Max when you do visit him, Antti ... Anyway ...
> This is exactly the problem: I think everyone would accept that people
> providing free services, whether cores or software, need to live - so
> trying to make money from them isn't a crime. All that's needed is
> openness, so others can see what's going on. It's hiding things that is
> problematic.
Exactly my point !
Something that Max could not understand - it's his deceptive
behavior that is unacceptable. We fight enough real bugs, we
don't need additional artificial bugs. If I remember correctly
he has also stated that his core did not meet some specification
the way it was, even though he did claim on the web site it did.
> Based on what I've seen on opencores, here are some ideas on what might be
> possible tactics; in each case there is a version which people may find
> acceptable, and a version they probably won't find acceptable.
>
> 1.
...
>
> 2.
...
>
> 3.
...
> The small core is so simple it is useless, so that it is ONLY an advertisement.
Yes, those are very good options.
...
> The solution is to create a culture where people are encouraged to give
> feedback: make it as easy as possible for people to comment on cores on
> the site; and in particular where people find bugs, make it easy and
> expected for them to at least give a bug report, and ideally provide a
> patch. Maybe a minimal start might be for each core to provide a
> publicly updateable page for each core with fields like:
>
> a.Found to compile/not compile successfully with: <name of program>
> b.Simulates correctly/does not simulate correctly with <name of program>
> and testbench <upload>
> c.Synthesized/failed to synthesize for : <ASIC/FPGA type>
> d.Clock rate achieved:
> e.No of blocks used:
> f.Comments:
Problem with these kind of comments is that you get guys who
are less experienced and start posting negative feedback that
is due to their lack of understanding how to drive a tool for
example (which can be a real chalange in hardware design).
Instead of a Rating system, what we need is a really good bug
tracking system. All of those things can be logged as bugs and
give me as a designer the opportunity to properly respond, and
either fix the bug or invalidate the claim.
...
> Also, maybe replace the 'done' tag on the projects with 'usable'.
> If it was software, you'd never say 'done' - it's always version x, which
> comes before version x+1... ;-). 'Done' encourages you to think - 'no
> feedback needed'.
Actually I disagree with you on this point.
What you say is true for software, BUT for an IP, it's
either done or still in development. If a IP is done it
is supposed to be 100% bug free and ready for silicon.
It it is not, than it is still in development.
...
> Graham
Regards,
rudi
========================================================
ASICS.ws ::: Solutions for your ASIC/FPGA needs :::
..............::: FPGAs * Full Custom ICs * IP Cores :::
FREE IP Cores -> http://www.asics.ws/ <- FREE EDA Tools
|
 |
Browse
