Re: gEDA-user: âsymversionâ documentation update

[Kai-Martin Knaak <knaak@xxxxxxxxxxxxxxxxxxx>]

Colin D Bennett wrote:

> The âsymversionâ attribute's documentation says that this attribute
> should not be placed on new symbols, but only (1) on symbols âaccepted
> into the main gEDA symbol libraryâ, and even then (2) only when a
> change is made to the symbol.

I have asked about this before on the list. IIRC, nobody came up with 
a conclusive reason for (1) and (2). 


> I always add symversion=1.0 to my custom symbols because (1) 

Unfortunately, I did not since a few months ago. Now, that I started
using the symversion gschem throws unnecessary warnings at me. It seems
to interprete the appearance of a symversion symbol where there has none 
been before as a "major version change". Sorry, but it's not. Starting
a version series does not constitute a change of versions.


> it makes updating the version easier when the first change is made

ack.


> (2) version control system diffs show the version number change more
> readily and consistent with subsequent version updates when it's
> only a text change (e.g., â1.0â -> â1.1â) rather than an entirely new
> attribute after the first change.

ack.
 

> Anyway, what is the harm in adding symversion to brand new symbols?
> The wiki page says âNew symbols should not receive this attribute at
> all. Only when the symbol is change should this attribute be placed
> into the symbol file and maintained.â

The harm of this statement in the documenation is clear: It made me 
omit  symversion attributes in my symbol library. Now, I have to deal 
with the above update pains.


> It seems like the wiki has many unnecessary and unexplained mandates
> like this.

True.
In addition, the description should be less vague as to where the 
attribute is used and what it means to not have it set in a symbol.
Restrictions on attribute syntax are also badly missing. 
Some more attributes with missing/doubtful/oudated description are
   device --> what component uses this attribute for what purpose?
   pinseq --> restricted to integers? 
   numslots --> if there is no slotting gnetlist is fine without this
   

> How about replacing the symversion section with the following, updated
> text:
(...snip...)

+1 
The section on promotion and consequences of a change are particularily 
important.

---<)kaimartin(>---
-- 
Kai-Martin Knaak                                  tel: +49-511-762-2895
UniversitÃt Hannover, Inst. fÃr Quantenoptik      fax: +49-511-762-2211	
Welfengarten 1, 30167 Hannover           http://www.iqo.uni-hannover.de
GPG key:    http://pgp.mit.edu:11371/pks/lookup?search=Knaak+kmk&op=get



_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user