Hi Scott,
I tried to apply your patch v12 but got two rejects (using a fresh source
clone). See files attached.
Patch v11 applies correctly.
Regards,
Cedric
2014-12-30 1:25 GMT+01:00 Scott Talbert <s...@techie.net>:
> On Sat, 29 Nov 2014, Scott Talbert wrote:
>
> > On Thu, 27 Nov 2014, Phil Dibowitz wrote:
> >
> >>> - printf(" Serial Number: %s\n\t%s\n\t%s\n", get_serial(1),
> >>> - get_serial(2), get_serial(3));
> >>> + if (strlen(mh_get_serial()) != 0)
> >>> + printf(" Serial Number: %s\n", mh_get_serial());
> >>> + else
> >>> + printf(" Serial Number: %s\n\t%s\n\t%s\n", get_serial(1),
> >>> + get_serial(2), get_serial(3));
> >>
> >> How about:
> >>
> >> if (is_mh_remote()) {
> >>
> >> ?
> >
> > Can't, because not all MH remotes use this weird serial number format.
> > The earlier ones use the more traditional one.
> >
> >>> +int mh_read_file(const char *filename, uint8_t *buffer, const
> uint32_t buflen,
> >>> + uint32_t *data_read)
> >>
> >> Is that 2 tabs and a space? We just nuked all our tabs! I think that
> should be
> >> 17 spaces. :)
> >
> > Ooops. I don't know how that snuck in there. I'll fix in the next
> > version.
> >
> >>> int err;
> >>> - if ((err = usb_set_configuration(h_hid, 1))) {
> >>> - debug("Failed to set device configuration: %d (%s)", err,
> >>> - usb_strerror());
> >>> - return err;
> >>> - }
> >>
> >> Errrrr. I'm not sure 1 is always the default on every device. This
> change
> >> terrifies me.
> >
> > Heh. I figured you might ask about this. So, for one thing, this only
> > affects the libusbhid code (which the person who was testing the Harmony
> > Touch happened to be using). HIDAPI doesn't ever call Set Configuration
> > and we seem to be working fine there. If it would make you feel better,
> > maybe we can make a Get Configuration call first and then Set
> > Configuration if it isn't 1. Unfortunately, for the Touch (and probably
> > also at least the Link), calling Set Configuration when the configuration
> > is already set to 1 causes the device to reset.
> >
> >>> uint8_t msg_ack[MH_MAX_PACKET_SIZE] =
> >>> - { 0xFF, 0x03, get_seq(seq), 0x02, 0x01, param, 0x01,
> pkts_to_send };
> >>> + { 0xFF, 0x03, get_seq(seq), 0x02, 0x01, param, 0x01, 0x33 };
> >>> + if (pkts_to_send < 0x33)
> >>> + msg_ack[7] = pkts_to_send;
> >>
> >> I suppose it's the same, but this logic seems backwards to me. I would
> do
> >> something like:
> >>
> >> uint8_t msg_ack[MH_MAX_PACKET_SIZE] =
> >> { 0xFF, 0x03, get_seq(seq), 0x02, 0x01, param, 0x01,
> pkts_to_send };
> >> // cannot be > 0x33
> >> if (pkts_to_send > 0x33)
> >> msg_ack[7] = 0x33
> >>
> >> Oh, I see why you did it the way you did - because of the ACK logic
> below. OK,
> >> I don't actually care. In theory, if we are likely to send more than 50
> >> packets in the common case, you should keep your logic, and if that's
> the
> >> less-common scenario you should switch it - but again, I don't actually
> care,
> >> just thinking out-loud.
> >
> > Eh, it's hard to say whether we'll commonly send more or less than 50
> > packets. I think probably in this use case, we'll send less, but I was
> > thinking about another patch down the road to refactor the UpdateConfig
> > code to use this code, and I think in that case, we'd probably be sending
> >> 50 packets most of the time. So, eh, I think I'll just leave this
> > as-is.
> >
> >>> - * MH remotes do not support SetTime() operations, but we return
> >>> + * Some MH remotes do not support SetTime() operations, but we
> return
> >>> * success because some higher level operations (for example,
> update
> >>> * configuration) call SetTime() and thus the whole operation
> would be
> >>> * declared a failure, which we do not want.
> >>> */
> >>
> >> For this patch / now, I'm fine with this, but I wonder if we should
> instead
> >> expose a _is_set_time_supported function that upper-level functions can
> call
> >> to determine if this is a sensible thing to do.
> >>
> >> By doing that, when a user asks to set time we can actually say "Sorry,
> your
> >> device doesn't support that.
> >
> > Yeah, that's probably not a bad idea for the future.
>
> Ping! It's been a month...
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming! The Go Parallel Website,
> sponsored by Intel and developed in partnership with Slashdot Media, is
> your
> hub for all things parallel software development, from weekly thought
> leadership blogs to news, videos, case studies, tutorials and more. Take a
> look and join the conversation now. http://goparallel.sourceforge.net
> _______________________________________________
> concordance-devel mailing list
> concordance-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/concordance-devel
>
--- libconcord/remote_info.h
+++ libconcord/remote_info.h
@@ -129,7 +129,18 @@
{ MFG_UNK, "Unknown", NULL },
{ MFG_UNK, "Unknown", NULL },
// 90
- { MFG_UNK, "Unknown", NULL }
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_HAR, "Harmony Hub", NULL },
+ { MFG_UNK, "Unknown", NULL },
+ { MFG_HAR, "Harmony Touch", NULL },
+// 100
+ { MFG_HAR, "Harmony Ultimate", NULL }
};
static const unsigned int max_model=sizeof(ModelList)/sizeof(TModel)-1;
@@ -500,7 +511,7 @@
0, // eeprom_size
"", // usb
},
- /* arch 17: Link */
+ /* arch 17: Link/Touch */
{
0, // serial_location
0, // serial_address
--- libconcord/remote_mh.cpp
+++ libconcord/remote_mh.cpp
@@ -714,12 +744,32 @@
lc_callback cb, void *cb_arg, uint32_t cb_stage)
{
/*
- * MH remotes do not support SetTime() operations, but we return
+ * Some MH remotes do not support SetTime() operations, but we return
* success because some higher level operations (for example, update
* configuration) call SetTime() and thus the whole operation would be
* declared a failure, which we do not want.
*/
- return 0;
+ if (ri.architecture != 17) {
+ return 0;
+ } else {
+ /* Yes, the official sw seems to hard code the US-East TZ */
+ const char *tz_str = "EST5EDT,M3.2.0,M11.1.0";
+ size_t tz_str_len = strlen(tz_str);
+ const uint32_t tsv_len = 16 + tz_str_len;
+ uint8_t tsv[tsv_len];
+ tsv[0] = ht.year >> 8;
+ tsv[1] = ht.year;
+ tsv[2] = ht.month;
+ tsv[3] = ht.day;
+ tsv[4] = ht.hour;
+ tsv[5] = ht.minute;
+ tsv[6] = ht.second;
+ tsv[7] = ht.dow;
+ for (int i = 8; i < 16; i++)
+ tsv[i] = 0;
+ memcpy(&tsv[16], tz_str, tz_str_len);
+ return WriteFile("/sys/time", tsv, tsv_len);
+ }
}
int CRemoteMH::LearnIR(uint32_t *freq, uint32_t **ir_signal,
------------------------------------------------------------------------------
Dive into the World of Parallel Programming! The Go Parallel Website,
sponsored by Intel and developed in partnership with Slashdot Media, is your
hub for all things parallel software development, from weekly thought
leadership blogs to news, videos, case studies, tutorials and more. Take a
look and join the conversation now. http://goparallel.sourceforge.net
_______________________________________________
concordance-devel mailing list
concordance-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/concordance-devel
