Saying "AJAX" or talking about any other implementation tech feels
to me like putting the cart before the horse.  I still don't know
what problems would expect be solved ... though I can think of three
right off the bat that *I* would expect to work:

 * Eclipse integration.  Probably covered for GDB support, but not
   necessarily for other tasks (and it's not like I can do a useful
   "apt-get install eclipse" on Ubuntu), like ...

 * Flash updating.  This being very board-specific.  "Write this
   version of U-Boot to that address" ... sometimes with a special
   header, or needing a "helper" boot loader (e.g. ROM loads helper
   from NAND to on-chip SRAM, which sets up DRAM and loads U-BOOT
   from a different part of NAND).

 * Tell me what's on that JTAG chain, and build an "openocd.cfg" to
   start out with.  Some basic textual or pictorial display; this
   would need ways to grow and manage a chip database.  Plus maybe
   ways to hint about special features.

Also, user interface design skills are rather distinct from hardware
interfacing skills.  Getting very deeply into that would probably
call for a new set of developers.  Especially if something like an
AJAX based web client were in the cards...

But I expect that some simple tools could be written just using TK/Wish.
That flash updating thing, for example ... have a look at the "SAM-BA"
tool from Atmel, which ISTR is written in TK and talks directly to the
ROM boot loader on newer AT91sam chips, using USB or a UART.  (Various
ARMs from TI could do the same thing; both DaVinci and OMAP3 parts, 
in particular.)

How would such a flash updating thing work?  Select interface dongle
from a menu; select board (or maybe just chip/target) ditto.  Then
the board setup file could say "provide these menu commands".  Doing
all that in TK/Wish *ought* to be really easy...

Right now very few boards have much in the way of script support.
I mention "write the boot loader" since it's a fairly basic thing
that is widely needed, and should thus be easy to approach ... but
there will likely be other tasks worth targeting.

- Dave

_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to