15, 16, 115, 116, 215, 216, 270, 272, 275, 304, 309, 315, 316.
The testcase is http://bugzilla.mozilla.org/showattachment.cgi?attach_id=11073
from bug 44709.
Current status: x15, and x16 where x=0..3 are handled correctly. - kudos
The others [which are at least obscure :-) ] do not seem to be handled.
[For convenience and in the event the page changes] from the reference url:
The numbers 270, 272, 275, 304 and 309 are also numbers which are not written
in the normal pattern, because the normal pattern would spell a word with
negative connotations. 270 and 275 would normally be spelled raysh ayin and
raysh ayin hey. These spell ra and ra'ah, both of which mean "bad". 272 would
normally be spelled resh ayin bet which spells ra'av meaning "famine." 304
would normally be spelled shin dalet which spells shaid meaning "demon." 309
would normally be spelled shin tet which spells shat meaning "to go astray."
Since the normal way to write these numbers spells words with negative
connotations, these numbers are written in reverse order.
as of my 9/15-08 w32-talkback build we still don't default to including bidi
which is necessary for these bugs to be seen correctly (instead you see them
backwards which is I think the recommended output for these outliers), although
we could and should fix them now...
Re-assigning 7 bugs from Clayton's list to myself for triage...
Marking future and re-assigning to the layout team's manager, Chris Karnaze.
Reassinging to Buster.
Reversing the order of numbers with "negative connotations" is not standard
practice in Hebrew, and I have only seen it done in specifically religious
contexts. I think this should be set to INVALID.
There are some other problems with Hebrew numbering (especially on a platform
with Hebrew support) which will be fixed when the IBMBIDI code is checked in.
Reassigning QA Contact for all open and unverified bugs previously under Lorca's
care to Gerardo as per phone conversation this morning.
qa contact updated.
I made the origional list-style-type: hebrew algorithm. See the "The Hebrew
Numbering System" section for details.
reassign to firstname.lastname@example.org
Mass-move all BiDi Hebrew and Arabic qa to me, email@example.com.
Thank you Gilad for your service to this component, and best of luck to you
in the future.
afaik using hebrew numbers > 100 is hardly standard practice anywhere, which doesn't leave many consumers. in my amused opinion the major consumers would be either people like me (ie those seeking amusement) or those intent on using the numbers for something akin to religious purposes.
For comment 4 I based myself on some Hebrew books that used Hebrew digits for
the page numbers (the Koren edition of the Bible and (IIRC) the Shocken edition
of the complete works of Agnon). These both use the normal order for all numbers.
Where I have seen the reordering is in the paragraph numbers of various editions
of the Shulhan Aruch, Also in the names of years: e.g. 5698 (1937-8) was not
usually written as taf resh tzadi het, meaning "murder". However I hardly ever
saw reordering of 5744 (1983-4) from taf shin mem dalet, meaning "destroy".
So what do we want to do here? Is this INVALID? Should we even bother with the
reordering of 15 and 16? I'd love to know what we want to do here since whatever
we do is what I intend to put in the CSS3 lists module as the official way of
doing hebrew list numbering...
I have never seen in any place that uses Hebrew numbering yod he and yod vav
used, but always tet zayin and yod zain (both as stand alone and as part of a
So we should defininatly follow here.
Other "forbidden" combinations I have seen in use, albit rarely- so we do have
some slack here.
For the purposes of a spec I would say that x15 and x16 SHOULD be expressed as
x+9+6 and x+9+7 rather than x+10+5 and x+10+6; and other numbers MAY be
reordered if desired to avoid spelling out words with negative connotations.
For the purposes of Mozilla, if anyone wants to write code to implement the
reordering, they are welcome to, but I don't personally consider it a priority.
Ok, cool. Thanks for the help!
It's about time this one disappeared off our radar.
this bug isn't invalid. if you feel it worksfor you then resolve it as such.
for the record it seems like the old url with the relevant information is gone
as the domain is for sale. note to self: use the "Create a New Attachment" link :(