Taken for Granted

There is an interesting debate and discussion going on at Glenn Perry’s blog on the definition of ESL – see “Keeping the “E” in ESL”.   I won’t try to hijack the debate here (a violation of “blogiquette” in my opinion) – so please follow the link to Glenn Perry’s blog and join in the discussion there, but I will add a couple of comments about where I see ESL.

Given our current state of understanding about System Level Design (SLD) and ESL, I remain in the “Big Tent” camp of ESL – that is, there are a number of design, verification and analysis activities that should be under the broad definition of the term ESL, and these are related to an even larger number of specification, analysis, partitioning and capture activities that are under the even broader definition of SLD.  Furthermore, the boundary lines between SLD and ESL are at the present time necessarily fuzzy and not completely determined.

As designers and developers continue to wrestle with these SLD and ESL activities and tasks,  and the tools, models and methodologies supporting them, eventually we may come to have a more consensus view on the “ESL Design Ideology” as complementing the current “RTL Design Ideology”  (see what I wrote in April:  “Do designers share a common ESL design ideology?”).   But it may take a while longer for this to emerge.   When it does, some of the ESL-ish activities currently in the broad tent may build tents of their own under new names, or even drop off in favour of other ones.

It is a little premature and dangerous to try to be overly definitive in establishing the bounds of a domain that is still quite new to many designers.  One piece of history to remember is the history of Computer Hardware Description Languages (CHDLs).   Between the 1960′s and the 1980′s, there were literally hundreds of them – and a long chain of CHDL conferences held to discuss them, held every second year from at least 1973 to 1997 (see the CHDL 1997 web page).    You probably can still find people who will argue the merits of Ella (Ella was to Occam as Verilog was to C and VHDL was to Ada) and n.  (as in N-dot).   And yet by the mid 1990s we were really down to two digital HDLs of importance – Verilog and VHDL (with of course AMS variants and other specialised languages like HVLs emerging later).   Letting the ESL domain sort itself out through natural usage, debate and the decisions of designers as to which tools, models and methods they end up calling ESL is probably a better path than to try to be overly prescriptive.

So please link to Glenn Perry’s blog and continue the debate on ESL there.  Hope you join in!