runxiyu_ has quit [Quit: ZNC 1.8.2+deb3.1 - https://znc.in]
runxiyu_ has joined #commonlisp
awlygj has joined #commonlisp
kg7ski- has left #commonlisp [Leaving]
shawnw has quit [Ping timeout: 268 seconds]
Bubblegumdrop has quit [Read error: Connection reset by peer]
Bubblegumdrop_ has joined #commonlisp
rainthree has quit [Ping timeout: 268 seconds]
ronald has joined #commonlisp
istewart has joined #commonlisp
istewart has quit [Quit: Konversation terminated!]
robin_ has joined #commonlisp
robin has quit [Ping timeout: 240 seconds]
robin_ is now known as robin
wacki has joined #commonlisp
chiselfuse has quit [Remote host closed the connection]
chiselfuse has joined #commonlisp
bilegeek has quit [Quit: Leaving]
brokkoli_origin has quit [Ping timeout: 268 seconds]
brokkoli_origin has joined #commonlisp
shawnw has joined #commonlisp
shawnw has quit [Ping timeout: 256 seconds]
Spawns_Carpeting has quit [Ping timeout: 260 seconds]
shawnw has joined #commonlisp
Pixel_Outlaw has quit [Remote host closed the connection]
kenran has joined #commonlisp
skempf has quit [Quit: .]
skempf has joined #commonlisp
amb007 has quit [Ping timeout: 256 seconds]
amb007 has joined #commonlisp
rainthree has joined #commonlisp
Th30n has joined #commonlisp
danse-nr3 has joined #commonlisp
<rainthree>
beach: to serialize a hash-table with Clobber, I serialize it to a transaction form (a cons) that when read and executed gives the hash-table back. https://paste.rs/ciIHw . It would be good to find a way to use make-instance instead of alist-to-hash-table (which could of course be used behind the hood by make-instance), because I'd rather want all the databases to depend on the symbol 'make-instance rather than
<rainthree>
'clobber:alist-to-hash-table. (Unless of course, I am sure 'clobber:alist-to-hash-table is going to be used forever, I won't have to rename it, etc )
<rainthree>
using 'make-instance would be a cleaner way.
shka has joined #commonlisp
<beach>
It is probably not a good idea to have forms in the log.
<beach>
I also don't see how you can call MAKE-INSTANCE to create a hash table.
<beach>
Can't you just special-case the reader macro for #\[ so that when it sees HASH-TABLE instead of calling MAKE-INSTANCE it calls MAKE-HASH-TABLE?
<rainthree>
make-hash-table is limited and won't populate it with default data
<rainthree>
which the hash-table contains at the time of serialization
shawnw has quit [Ping timeout: 252 seconds]
<beach>
OK, so then call a function MAKE-A-HASH-TABLE that takes the elements as an argument as well.
<beach>
You can have the log contain [hash-table :element-type ... :elements ((key . value) ...)]
<beach>
Something like that.
<rainthree>
Great, let me try
danse-nr3 has quit [Read error: Connection reset by peer]
danse-nr3 has joined #commonlisp
random-nick has joined #commonlisp
danse-nr3 has quit [Ping timeout: 255 seconds]
danse-nr3 has joined #commonlisp
<beach>
I meant :test rather than :element-type. *sigh*
King_julian has joined #commonlisp
pve has joined #commonlisp
<rainthree>
it works fine. Before I push all the changes: how should we call the macro that ensures all the distinct (execute ...) forms in its body are committed only after all of them went through. clobber:with-atomic-logging ?
<beach>
Sure. Don't worry too much about it. A name can always be changed later.
<rainthree>
ok
rainthree_ has joined #commonlisp
rainthree has quit [Ping timeout: 256 seconds]
King_julian has quit [Ping timeout: 268 seconds]
rainthree_ is now known as rainthree
dino_tutter has joined #commonlisp
Gleefre has quit [Remote host closed the connection]
shawnw has joined #commonlisp
<younder>
beach: I have had trouble getting Munchnick’s book ‘Advanced compiler design and implementation’. When I ordered one from Amazon it somehow was sendt back. Do you happen to know where I can get a copy. I had temporary access to a online version, but I want a printed copy.
attila_lendvai has joined #commonlisp
King_julian has joined #commonlisp
<beach>
I am afraid I don't know. I also ordered from Amazon, but got a strange copy with Chinese characters on the cover.
<younder>
Well I have other books like 'Optimizing compilers for modern architectures' by Alan Kenedy. But that pertains mostly to parallel Fortran
<beach>
I haven't read them, but I have been told they are good.
danse-nr3 has quit [Remote host closed the connection]
danse-nr3 has joined #commonlisp
<beach>
I think it is one book written three times using three different programming languages, but I am not sure about that.
<pranav>
Compiling with Continuations?
<beach>
I was thinking of "Modern Compiler Implementation in <language>"
<pranav>
I see.
shawnw has quit [Ping timeout: 256 seconds]
King_julian has quit [Ping timeout: 264 seconds]
<younder>
Ok beach, Just bought modern compiler architecture in C by Andrew Apel
<beach>
Great.
King_julian has joined #commonlisp
Noisytoot has quit [Ping timeout: 246 seconds]
Gleefre has quit [Remote host closed the connection]
Gleefre has joined #commonlisp
Noisytoot has joined #commonlisp
shawnw has joined #commonlisp
Noisytoot has quit [Remote host closed the connection]
Noisytoot has joined #commonlisp
Guest5585 has quit [Ping timeout: 264 seconds]
Gleefre has quit [Remote host closed the connection]
Gleefre has joined #commonlisp
King_julian has quit [Ping timeout: 255 seconds]
danse-nr3 has quit [Read error: Connection reset by peer]
danse-nr3 has joined #commonlisp
varjag has joined #commonlisp
amb007 has quit [Ping timeout: 268 seconds]
amb007 has joined #commonlisp
shawnw has quit [Remote host closed the connection]
King_julian has joined #commonlisp
danse-nr3 has quit [Ping timeout: 246 seconds]
danse-nr3 has joined #commonlisp
awlygj has quit [Quit: leaving]
danse-nr3 has quit [Ping timeout: 256 seconds]
X-Scale has joined #commonlisp
King_julian has quit [Ping timeout: 264 seconds]
kevingal has joined #commonlisp
kevingal_ has joined #commonlisp
King_julian has joined #commonlisp
danse-nr3 has joined #commonlisp
donleo has joined #commonlisp
skeem has joined #commonlisp
yitzi has joined #commonlisp
danse-nr3 has quit [Ping timeout: 246 seconds]
<Gleefre>
I wonder if :NAME and ||:NAME should mean the same thing or not. It seems that different implementations answer this question differently.
<Gleefre>
Hm, is "" even a valid name for a package?..
<beach>
Sure. It is a valid string, so why not.
ronald has quit [Ping timeout: 246 seconds]
<Gleefre>
It seems that all implementations can create a package named ""; but Allegro CL specifically defines "" to be a global nickname for the KEYWORD package.
<beach>
That doesn't sound conforming.
<Gleefre>
I'm not sure. Implementations are allowed to have implementation-defined packages after all, having an implementation-defined nickname for the KEYWORD package might be fine as well.
prokhor has quit [Remote host closed the connection]
prokhor has joined #commonlisp
<Gleefre>
But it does mean that if you do (rename-package "KEYWORD" "KEYWORD" nil) on AllegroCL you will no longer be able to read keywords. (Unless at toplevel, but then they are instead "top-level commands" and not keywrods anyways)
X-Scale has quit [Ping timeout: 250 seconds]
danse-nr3 has joined #commonlisp
chomwitt has joined #commonlisp
<Gleefre>
Another thing is that you then are able to shadow this nickname with a local nickname "".
prokhor has quit [Remote host closed the connection]
prokhor has joined #commonlisp
<Gleefre>
About reading ||:NAME vs :NAME -- it seems that most implementations treat ||:NAME as :NAME; and only SBCL, CMUCL and CCL read it as a symbol in a package named "". That is, except for AllegroCL that reads :NAME as ||:NAME and looks it up in the package named "" which is a nickname for "KEYWORD".
<Gleefre>
Also ABCL just crashes when trying to read ||:NAME :/
<yitzi>
dbotten: You'd need to add a dispatch entry for defclass that is better than what SBCL does...
<dbotton>
know any examples of that?
hexa6 has quit [Ping timeout: 272 seconds]
<yitzi>
The output I gave is from Inravina's pprint. It does client dispatch so it is a little more involved. Give me a couple of minutes and I can come up with a simpler example if you want.
danse-nr3 has quit [Ping timeout: 256 seconds]
<dbotton>
I'll give it a try first, thanks!
jon_atack has quit [Read error: Connection reset by peer]
danse-nr3 has joined #commonlisp
X-Scale has joined #commonlisp
<rainthree>
https://github.com/robert-strandh/Clobber/pull/11 no idea why git hasn't been able to detect the DIFF between old and new versions of some files and thought they were written from scratch
<ixelp>
with-atomic-logging; serialization for generic-functions and hash-tables by rainthree · Pull Request #11 · robert-strand [...]
<ixelp>
A package named "", ||:symbol · Issue #63 · s-expressionists/wscl · GitHub
<beach>
Gleefre: Don't worry about it. It is going to be turned into something more elaborate at some point. So this is just like a reminder. Thanks!
<Gleefre>
FWIW this issue also affects package-local nicknames -- is it allowed to shadow ""? Does it mean that :NAME should be read not as a keyword but as a symbol in the package nicknamed by ""? e.t.c
<Gleefre>
(This was the original question of mine)
<bike>
says :foo is in the keyword package, irrespective of ""
<Gleefre>
That's true, however a lot of implementations treat ||:xxxx the same way, which I'm not sure is a correct behaviour; and AllegroCL treats :xxxx as ||:xxxx, so shadowing "" with a PLN results in :xxxx being read NOT as a keyword
kenran has quit [Remote host closed the connection]
<Gleefre>
I wonder if creating a local nickname of "" should affect :xxxx syntax. On the one hand, the way it is phrased in hyperspec; and just for consistency/readability it probably shouldn't. On the other hand, it is a fun feature, and it kind of simplifies the symbol token parsing if you were to simply globally nickname "KEYWORD" package as "" (the way Allegro CL does).
ronald has quit [Read error: Connection reset by peer]
mgl_ has joined #commonlisp
wacki has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
kg7ski- has left #commonlisp [Leaving]
wacki has joined #commonlisp
Lord_of_Life has quit [Ping timeout: 256 seconds]
kg7ski- has joined #commonlisp
<gilberth>
Does someone know without consulting the spec, if ANSI CL promises (eq #'foo #'foo) to be true?
Lord_of_Life has joined #commonlisp
akoana has joined #commonlisp
<thuna`>
I would assume that it does. Does it?
<gilberth>
I'm not so sure this is why I ask. It could be like with integers or floats. I writing a very, very simple minded compiler and was about to box a function pointer as needed. Like I would box a float.
<gilberth>
And I was hoping someone would know. Guess I would need to dive into the spec myself :/
<bike>
i don't believe it spells anything out explicitly
<thuna`>
Oh, I figured the spec didn't promise it since you said "without consulting the spec". I believe it should be EQ, since (function name) is "the functional value of name", of which there can only be one
<bike>
i can tell you that one time in cleavir I screwed up inlining in such a way that that would be false, and make-hash-table broke
<gilberth>
Yes, and like with numbers when you declare FTYPE there might be a different representation. Like pointing to another entry point skipping some type checks.
<gilberth>
Or skipping checking count of arguments. Whatever.
waleee has joined #commonlisp
<thuna`>
I would say that FUNCTION is clear enough that it returns the object held in the function slot of the function name passed to it, so if the function name you're passing to it is EQ then the result should be EQ as well.
<thuna`>
s/enough/in/
<gilberth>
What if there is no function cell? What if #'foo is a local function made with LABELS or FLET?
NotThatRPG has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
<thuna`>
It says "[it's] function value ... in the current lexical environment" where it later specifies FLET, LABELS, and MACROLET as forms establishing such lexical environments, so I would say that it doesn't matter if it was bound lexically or not.
<gilberth>
When I say (let ((x 1.0)) (eq x x)) might return false.
<thuna`>
That doesn't seem conforming
n00b1 has joined #commonlisp
<gilberth>
"An implementation is permitted to make “copies” of characters and numbers at any time. The effect is that Common Lisp makes no guarantee that eq is true even when both its arguments are “the same thing” if that thing is a character or number." <https://novaspec.org/cl/f_eq>
<ixelp>
eq | Common Lisp Nova Spec
<gilberth>
It doesn't mention functions though.
<thuna`>
Ahh, it's at the bottom. Hmm...
<gilberth>
So perhaps I make sure that (eq #'foo #'foo) even when I find a loophole allowing me to do otherwise. That's the safest bet.
<thuna`>
It's unfortunate that that's how that works. I would probably say that copying is allowed only for characters and numbers but I would have also said that (let ((x 1.0)) (eq x x)) be true so I guess checking (eq #'foo #'foo) is the safest way to go
<thuna`>
s/be/is/
<gilberth>
As ANSI-CL specfied it, it makes perfect sense for numbers.
<gilberth>
That 'x' might live in a floating point register and would need to boxed only when passed to someone not dealing with unboxed numbers. And so the compiler might box it only when needed.
<gilberth>
I was about to have unboxed function pointers.
<bike>
cool.
<bike>
i've thought about that a decent amount but haven't put them into practice yet. they're something i would definitely like to work on
<bike>
i think you're basically hitting an unfortunate fact about eq (that it exists)
<bike>
presumably you could implement eql on these boxes by comparing the underlying pointers
X-Scale74 has joined #commonlisp
<bike>
for your eq to be conforming, it probably needs to do the same thing, even though that's slower than the simple comparison
<gilberth>
There is a reason that EQL usually is the default. EQ exposes some implementation details. OTOH EQ is nice for objects having a true identity and is just your CMP machine instruction.
X-Scale has quit [Ping timeout: 250 seconds]
<gilberth>
I'll punt and forget about unboxed function pointers. I do not feel like compromising EQ being a single instruction. What I do is experiments only anyway.
n00b1 has quit [Ping timeout: 240 seconds]
<bike>
you could also have this as a compiler optimization and then just not apply it if a function is fed to something that doesn't just call it
<bike>
(that's what i was planning on)
hexa6`` has joined #commonlisp
<gilberth>
Heh. (let ((x (+ a b))) (declare (dynamic-extent x)) ...) ? Makes my head hurt.
hexa6` has quit [Ping timeout: 268 seconds]
acidbong has left #commonlisp [#commonlisp]
NotThatRPG has joined #commonlisp
Spawns_Carpeting has joined #commonlisp
chomwitt has quit [Quit: WeeChat 3.8]
mgl_ has quit [Ping timeout: 268 seconds]
cognemo has quit [Ping timeout: 256 seconds]
NotThatRPG has quit [Ping timeout: 268 seconds]
NotThatRPG has joined #commonlisp
varjagg has joined #commonlisp
X-Scale74 has quit [Quit: Client closed]
mwnaylor has joined #commonlisp
mwnaylor is now known as Guest6296
mwnaylor has joined #commonlisp
mwnaylor is now known as Guest3291
Guest6296 has quit [Client Quit]
Guest3291 has quit [Client Quit]
mwnaylor has joined #commonlisp
mwnaylor is now known as Guest5706
Guest5706 has quit [Remote host closed the connection]
mwnaylor has joined #commonlisp
zagura has quit [Quit: Reconnecting]
zagura has joined #commonlisp
zagura has quit [Changing host]
zagura has joined #commonlisp
zetef has joined #commonlisp
zetef has quit [Client Quit]
NotThatRPG has quit [Ping timeout: 268 seconds]
NotThatRPG has joined #commonlisp
X-Scale has joined #commonlisp
attila_lendvai has quit [Ping timeout: 256 seconds]
waleee has quit [Ping timeout: 256 seconds]
NotThatRPG has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
simendsjo has quit [Ping timeout: 252 seconds]
mgl_ has joined #commonlisp
istewart has quit [Quit: Konversation terminated!]
pyooque has joined #commonlisp
puke is now known as Guest1570
Guest1570 has quit [Killed (copper.libera.chat (Nickname regained by services))]
pyooque is now known as puke
Jach has joined #commonlisp
puke has quit [Max SendQ exceeded]
puke has joined #commonlisp
X-Scale has quit [Quit: Client closed]
tfeb has joined #commonlisp
reb`` has quit [Ping timeout: 252 seconds]
cage has quit [Remote host closed the connection]
pve has quit [Quit: leaving]
tfeb has quit [Client Quit]
akoana has quit [Ping timeout: 252 seconds]
wacki has quit [Quit: My iMac has gone to sleep. ZZZzzz…]
gdsg has joined #commonlisp
mwnaylor has quit [Ping timeout: 256 seconds]
bollu has quit [Ping timeout: 268 seconds]
bollu has joined #commonlisp
triffid has quit [Remote host closed the connection]
X-Scale has joined #commonlisp
mgl_ has quit [Ping timeout: 256 seconds]
varjagg has quit [Ping timeout: 246 seconds]
zagura has quit [Quit: Reconnecting]
zagura has joined #commonlisp
zagura has quit [Changing host]
zagura has joined #commonlisp
yitzi has joined #commonlisp
zagura has quit [Quit: Reconnecting]
zagura has joined #commonlisp
zagura has quit [Changing host]
zagura has joined #commonlisp
dino_tutter has quit [Ping timeout: 268 seconds]
mwnaylor has joined #commonlisp
reb`` has joined #commonlisp
shka has quit [Quit: Konversation terminated!]
zagura has quit [Quit: Reconnecting]
zagura has joined #commonlisp
amb007 has quit [Ping timeout: 246 seconds]
zagura has quit [Quit: Reconnecting]
zagura has joined #commonlisp
jweeks has quit [Remote host closed the connection]
zagura has quit [Client Quit]
jweeks has joined #commonlisp
zagura has joined #commonlisp
jweeks has quit [Changing host]
jweeks has joined #commonlisp
zagura has quit [Client Quit]
zagura has joined #commonlisp
jweeks has quit [Remote host closed the connection]
cognemo has joined #commonlisp
istewart has joined #commonlisp
cognemo has quit [Client Quit]
donleo has quit [Ping timeout: 256 seconds]
amb007 has joined #commonlisp
X-Scale83 has joined #commonlisp
X-Scale has quit [Ping timeout: 250 seconds]
kiyafirs has joined #commonlisp
amb007 has quit [Read error: Connection reset by peer]