Closed GoogleCodeExporter closed 9 years ago
Do you know if (3) is really still needed? Probably we should ask the people
who implemented the current tier-2 version? Who are they?
Phones which don't support UTF-8 listed here:
http://miniturbo.org/2006/12/29/034842
are all Docomo Mova phones which anyway no longer don't work. But I'm not sure
about the coverage of the list.
Original comment by ichik...@google.com
on 29 Mar 2012 at 5:16
Hi Rag, Masato-san,
Ichikawa-san and I are working on Person Finder on our 20%, and this is one of
the issues we are thinking to prioritize on in Q2.
Item (1) and (2) are straight-forward, but we don't have sufficient knowledge
in (3). Ichikawa-san suggested that we probably don't need to support
Shift_JIS. Do you have any thoughts on this?
Please let me know if you are interested in the context in which we're
integrating this into the PF mainline.
Also, if any of you are interested in taking some part in this migration,
you're super welcomed!
Original comment by r...@google.com
on 29 Mar 2012 at 8:48
Phones may support utf-8, but that doesn't mean they support it well ;)
In particular, what never worked with the default person finder utf-8 mobile
version is form submission, because the phones still prefer submitting their
data in shift_jis.
It's _easy_ to get tier-2 sites working if you force everything to shift_jis
and just don't worry about it - this is the approach we took. I think it's
_possible_ to get things working with utf-8, but it will be tricky and there
may be some difficult to find corner cases. Then again, we didn't have the time
to work through these corner cases after the earthquake, while now there might
be enough time.
Original comment by anur...@google.com
on 29 Mar 2012 at 8:55
The main reason why we created the separate PF for tier-2 was that there are
many tier-2 phones which ignore any http encoding format and assume that
everything is done in Shift_JIS even if you set charset tag or http response
header.
So if you'd support tier-2 phones in PF mainline, then you'd probably need a
feature to enforce on a specified encoding.
Original comment by tarui...@google.com
on 29 Mar 2012 at 8:59
I see. Then probably we should go with Shift_JIS. Thanks for the explanation.
Original comment by ichik...@google.com
on 29 Mar 2012 at 9:21
Original comment by r...@google.com
on 4 Aug 2012 at 3:43
Original comment by r...@google.com
on 4 Aug 2012 at 4:32
Renamed title per Kazawa-san's suggestion.
Original comment by ichik...@google.com
on 6 Aug 2012 at 7:05
Marking as fixed. Huge thanks for working on this, Ichikawa-san.
We need to add a link on the top left corner, that takes users to ui=light, but
I'll file a separate bug for that.
Original comment by r...@google.com
on 21 Oct 2013 at 1:29
Original issue reported on code.google.com by
r...@google.com
on 29 Mar 2012 at 3:38