Should VHDL allow Unicode identifiers and comments

Thu Jul 28, 2011 7:13 pm   

Hi all,

I'm asking for a bit of input from the community...

As the title says, would you find it of use to allow Unicode identifiers
and comments in a future VHDL revision?

Would this be:
a) Something VHDL should not allow
b) Something that doesn't bother you either way
c) Something you'd find useful sometimes
d) Something you'd make use of all the time
e) Something that you'd switch away from SystemVerilog just to get at
(maybe I'm asking the wrong crowd for that :)

Thanks,
Martin

--
http://parallelpoints.com/

Thu Jul 28, 2011 9:10 pm   

On 7/28/2011 2:13 PM, Martin Thompson wrote:

b,

Regards,
Chris Fetlon

Thu Jul 28, 2011 9:19 pm   

On 7/28/2011 2:10 PM, Christopher Felton wrote:

Unless the introduction of said identifiers started breaking my existing
tools, in which case (a).

--
Rob Gaddi, Highland Technology
Email address is currently out of order

Thu Jul 28, 2011 11:45 pm   

On 7/28/2011 12:13 PM, Martin Thompson wrote:

c) Yes, will be helpful in the near future.
Otherwise everyone will have a different library for it.

Fri Jul 29, 2011 9:55 am   

Mike Treseler <mtreseler_at_gmail.com> writes:


I'm not sure I follow Mike - library for what? The original question
was about using Unicode within a VHDL source file (for example, variable
names and comments).

Or are you thinking of having a Unicode "string" replacement - which is
a whole different ballgame, but one we maybe ought to think of also!

Cheers,
Martin

Sat Jul 30, 2011 7:58 am   

On 7/29/2011 2:55 AM, Martin Thompson wrote:

OK. In that case probably (b) for English speakers.


Yes, I was thinking strings.
That seems safe and probably useful.
Programming languages without
Unicode strings built-in suffer as a result.


-- Mike Treseler

Sat Jul 30, 2011 2:04 pm   

On 28 Jul 2011 19:13:48 GMT, Martin Thompson
<martin_usenet_at_parallelpoints.com> wrote:


I'm not sure I see any use for it. What do you have in mind?

Unicode *strings* and file-IO might well be useful, but
I guess that's a very different story. A new type, either
built-in or in std.standard, for Unicode *characters* would
be a good start.
--
Jonathan Bromley

Mon Aug 01, 2011 8:58 am   

Jonathan Bromley <spam_at_oxfordbromley.plus.com> writes:


The original question was asked without much in mind beyond allowing you
to call a variable 'château' (to pull an example from the other end of
the scale spectrum to our usual fare here :)


From other comments, Unicode strings appear to be of much more value
than Unicode identifiers and comments. Although once you allow Unicode
strings in a source file, you've opened the "source-file encoding" can
of worms already, and then (I believe) allowing Unicode in comments
becomes easy. Unicode identifiers may have some negative impact of
parsing efficiency?

In which case, as you say, W_CHARACTER here we (might) come. However,
it also sounds like a large (huge?) amount of work which *may* be better
spent elsewhere.

Cheers,
Martin

--
martin.j.thompson_at_trw.com
TRW Conekt - Consultancy in Engineering, Knowledge and Technology
http://www.conekt.co.uk/capabilities/39-electronic-hardware

Mon Aug 01, 2011 10:03 am   

On Mon, 01 Aug 2011 09:58:21 +0100, Martin Thompson wrote:


C'est tout possible de faire son logiciel sans aucun accent :-)


I think I tend to agree. The EDA industry as a whole is
irremediably Anglophone, and muddles through pretty well
without internationalization.
--
Jonathan Bromley

Tue Aug 02, 2011 10:50 am   

Jonathan Bromley <spam_at_oxfordbromley.plus.com> writes:


I like that description :)


and likely will continue to do so!

Thanks,
Martin

--
martin.j.thompson_at_trw.com
TRW Conekt - Consultancy in Engineering, Knowledge and Technology
http://www.conekt.co.uk/capabilities/39-electronic-hardware

Mon Aug 15, 2011 1:20 pm   

Martin Thompson <martin_usenet_at_parallelpoints.com> writes:


I liked an example snippet in Python I saw some time ago. There's
another one at
http://programmers.stackexchange.com/questions/16010/is-it-bad-to-use-unicode-characters-in-variable-names

for example.

After all, if your angle is phi, then why bother writing it out when
you can just use 'φ' instead?

Sat Apr 28, 2012 5:23 pm   

I like that example. I prefer to write math equations using symbols too, rather than typing in 'English'. Sometimes, greek is better than english when it comes to math/physics... IMO.

I may not be answering the question, but I will vote:
f) something that the VHDL standard shouldn't be concerned about.

Let the tool vendors concern themselves on this when they have enough customer demand. I believe I won't face much problems when I'm using two different vendor tools (say for synthesis and simulation) who both claim to be Unicode-compliant.

Wed May 02, 2012 2:18 am   

On Thursday, July 28, 2011 12:13:48 PM UTC-7, Martin Thompson wrote:

Comments would be fine. I am more worried about the entire toolchain: netlist representation, p&r tools, graphical visualization.


- Daniel