logs archiveBotHelp.net / Freenode / #ada / 2015 / July / 15 / 1
wtp_
wait a minute, is that the exact code for the Makers child package? Is so, then doesn't function Make have to be in the *private* section of the package spec?
Lucretia
why? It's a normal child package
with a normal function
wtp_
I thought a private with means the package content can only be used in the private section
Lucretia
and I have Surfaces doing a "private with SDL.C_Pointers;" as well
yeah, but Makers is not a private spec
C_Pointers is
wtp_
no but like all specs, they can have a private section
so what happens if you move function Make under a private section in the Makers spec
?
Lucretia
hmm
damn
hang
damn yer right
cheated by importing make into the windows package
wtp_
glad I could help
Lucretia
thanks
definitely had a cheat, but tbh, the make will only be used internally.
(Action) can sleep now
and come back tomorrow and do the same thing again, but for something else.
(Action) is re-working the internal data of the each of the major types.
zzzzzzzzzzzzzzz
sparre
How is it with Unicode? The basic unit is a "code point", and not a character, right?
Then; does the standard contain a table of mappings from (sequences of) code points to characters?
Natacha
I guess it depends unit of what
and what you mean by "character"
see for example "combining" code points, like t6h6a6t6
unicode standard associates code points with character-level semantics, and then the fonts associates the same semantics with "glyphs" you can see on screen
sparre
In my book ø and ö are characters, but I know that Unicode allows more than one representation of the same visual element.
Natacha
indeed, because they have a code point representing them directly, and combining stuff to build them from standard o
but some combinations can only be expressed that way, like o6
sparre
But then there is also Å (last letter in Danish alphabet) and Å (physical unit Ångström).
(looks wrong here, but the point is the se anyway)
Natacha
that's why I mentioned "semantics" above :-)
sparre
So there isn't a unique mapping :-(
Natacha
it would have been too easy otherwise
sparre
Making interaction with other character sets a mess...
AayJay
I think characters that can be represented directly are there because some previous character encodings includes them, and Unicode wants to have a 1-to-1 mapping from and to the encodings
Natacha
how can you justify the existence of an ISO comitee if you deal with a very simple thing?
sparre
AayJay: The separate ångstrøm character is a unicode invention, so I don't buy that claim.
Whatever. I think I've learnt/been reminded about enough of the mess Unicode is for today.
AayJay
Seems like the U+212B + angstrom sign is added because some Japanese encodings has it: https://books.google.com.hk/books?id=wn5sXG8bEAcC&pg=PA74&redir_esc=y#v=onepage&q&f=false
ya, Unicode is a mess. Backwards compatibility strikes again...
Natacha
I don't really see the problem in distinguishing programmatically between a unit symbol and a random Danish letter
(though I admit I don't really see the point of having angströms when there are nanometers and picometers)
alimiracle
hi ada users
ada is dad
*ded
Natacha
Indeed, and she has been so for more than 161 years
first time I try -pg, and all I get is immediate segfault :-(
and of course, on toy programs it works fine, but when I need it for real...
funny, it seems to be raising a Storage_Error exception while raising an exception, recursively until it's stopped by stack blow up
so I got calloc() calling __jemalloc_* stuff, pthread_mutex_lock(), and bam! signal handler called, resolving into recursive Storage_Error
so something is unprofilable somewhere in task creation
Lucretia
sparre: no such thing as a character in unicode
sparre: basically character is really a grapheme cluster which is 1 or more code points
nerdboy
you *are* a character...
« prev next »