Re: [myhdl-list] Migen logic design toolbox - now with simulator

[Christopher F. <chr...@gm...>]

On 3/12/12 8:07 PM, Bob Cunningham wrote:
> On 03/12/2012 02:53 PM, Jan Decaluwe wrote:
>> On 03/12/2012 10:23 PM, Thomas Heller wrote:
>>> Am 12.03.2012 16:10, schrieb David Greenberg:
>>>> I like hearing about progress in the Python HDL front, whomever is
>>>> doing it. I don't think it makes sense to fracture these already tiny
>>>> communities.
>>> +1 from me too.
>>>
>>> Thomas
>>>
>> What do you both mean? Obviously Migen "fractures"
>> the community. Also, obviously they think they are making
>> progress. Do you want to hear about it or not?
>
>    From my perspective as an HDL newbie (but a 30+ year engineering veteran otherwise), major feature development in MyHDL appears to have stalled, and Migen provides important capabilities MyHDL either lacks or can implement only with cumbersome and fragile kluges or work-arounds.  (At least, that's what they look like to me, with my newbie eyes and brain.)
>

There are some reasons for this.  One, not many people are willing or 
able to contribute.  Two, things happen; people have babies (no sleep I 
get) and other events happen in peoples lives that can stall a project. 
  MyHDL has no intent of stopping (sorry if I am speaking for others).

I believe MyHDL would be more than happy to have people contribute. 
Sebastien's case is well documented in this mailing list.  I don't think 
it was an unfair handling.

It is very easy for outsiders to come and ask for new features (there is 
a similar thread on scipy-dev).  To use a project but not to give back 
and constantly ask for the latest buzz word feature, it can be taxing on 
the developers.  I am guilty of this as well, I have used MyHDL 
extensively and I haven't given back a whole lot.

I would like to discuss these "fragile kluges" or work arounds some 
more.  The best I recall (I haven't reread the mailing-list) there are 
enhancements people would like but no "fragile" kludge.

> MyHDL is more powerful overall, but Migen has attacked some specific problems is was designed to solve.  Problems, IIRC, that were prompted by perceived weaknesses in MyHDL.  As I understand it, Migen started as a rejected patch to MyHDL that was forced to find a life of its own.
>
The rejected patch has been discussed.  The idea wasn't rejected but the 
contempt of the process that had been established.  It isn't about a 
rejected patch it is about philosophical differences.

Constructive conversations about the process could be useful but no more 
about the rejected *broken* patch.  I think the process is fair, 
efficient, and required to make a successful project.

> I believe Sébastien continues to post here not to diminish MyHDL or its use in any way, but instead to show at least one way to more easily do things that are difficult in MyHDL.  If you look at his many non-Migen posts, you will see Sébastien has proven himself to be a very capable contributor to this list.
>

Sorry but I totally disagree.  Again, look back to the old posts.  He is 
not interested in the technical discussions as he tries to claim.  As 
you state, you have 30+ years as an Engineer.  Technical conversations 
can be long and in depth and don't end with one party exclaiming "you 
are just ...".

> My personal hope is that MyHDL and Migen will evolve toward one another, and eventually find ways to inter-operate, share features, or even merge capabilities.  I fully expect MyHDL may one day make Migen obsolete, but that day does not seem to be coming any time soon.
>

This is where I get confused.  Why is the burden on the MyHDL developers 
(Jan) to add the features and have the quality that people are use to 
with this project.  What is it about the development process that people 
object with that they would rather see a separate project created?

> Until the parties involved agree to work together toward shared goals, it may be best for Migen to start its own list.  When Sébastien gets his list going, he should be permitted to post sign-up instructions here.
>
> However, I hope a separate list won't be necessary:  When Sébastien has referred to his own Migen posts in other forums (such as MilkyMist), he has referred people to *this* *list*.  If anything, he has actively worked to increase MyHDL awareness, not to diminish it, or compete with it, or fracture it.

 From this side it just doesn't seem like that.  Some links to these 
conversations ... seeing is believing?

I don't see how responding to a question on the MyHDL mailing-list with 
"you should checkout migen" before any other responses is not 
deliberate: fracturing, diminishing, and competing.

>
> To clear the air, perhaps we could have a Migen-MyHDL shoot-out, where Sébastien posts some of what Migen does best, and the MyHDL pros can show us how best to do it in MyHDL.  The posts should be in the form of modules that can be integrated into other projects (MyHDL or Verilog or VHDL) for testing and comparison.  I know it would prove to be extremely educational for me, and may serve us all to better understand the underlying issues.
>

Who has the time.  And what does it prove, I simply have the opinion 
that it is rude to advertise and self promote on a project you branched 
from.

This is where this sucks, if Migen decided to branch for whatever 
reason, I don't care.  It is their right to try something different, do 
it their way.  Nothing wrong with branching and starting a new project. 
  I wish them the best, go conquer the world.

But I don't think the projects have similar goals.  And instead of 
making things personal and dragging this out it is best Migen goes on 
its own and stops exploiting this group.  It might not be fair for me to 
make such a statement, Jan has built this project/group and others have 
contributed as well, but that is my opinion.

> Sébastien/Migen has to be doing something right: It got Jan back on this list!  Yay!!!

You don't realize it but you are being a complete *ASS* with that statement.

>
>
> -BobC
>
>
> ------------------------------------------------------------------------------
> Keep Your Developer Skills Current with LearnDevNow!
> The most comprehensive online learning library for Microsoft developers
> is just $99.99! Visual Studio, SharePoint, SQL - plus HTML5, CSS3, MVC3,
> Metro Style Apps, more. Free future releases when you subscribe now!
> http://p.sf.net/sfu/learndevnow-d2d