<dnhester26>
Why does maphash not collect the results in a list by default? I had to debug to find out that was the case and replaced it with serapeum:maphash-return. I understand that not every function will have that, but in most cases it can just collect a list of nils. If the memory is a concern, it could have been an optional key argument to maphash whether to collect or not. Is collecting not the most common case?
<beach>
dnhester26: Because collecting requires memory allocation.
<beach>
dnhester26: Collecting is not necessarily the most common case. Operating on a subset of the entries with some function might be more common.
<dnhester26>
beach: wouldn't that be better handled by an optional key parameter allowing for `:allocate-p nil`?
<beach>
dnhester26: You have LOOP for that case.
<dnhester26>
or to at least allow it if it's not the most common case with a :allocate-p t
<beach>
Sure, we could imagine more keyword parameters to many functions in Common Lisp, but the language is already quite large as it is.
<dnhester26>
Right, I tried doing loop but the syntax was a bit hard for me to understand from the dictionary and didn't want to spend all that time searching the actual reference
<dnhester26>
what's the syntax for looping over a hash table with keys and values? `for-as-hash::=var [type-spec] being {each | the}` `{{hash-key | hash-keys} {in | of} hash-table` `[using (hash-value other-var)] |` `{hash-value | hash-values} {in | of} hash-table` `[using (hash-key other-var)]}` ?
<beach>
Yes.
<dnhester26>
what would it look like (loop for key as being hash-key of given-ht for value as being hash-value of given-ht do ...) ?
<dnhester26>
i take it this is faster than doing `for value = (gethash key given-ht) do ...`
<dnhester26>
to avoid the extra value lookup
<beach>
Maybe a bit.
<beach>
So, basically (loop for key being each hash-key of <table> using (hash-value value) ...)
<thuna`>
When I'm stuck trying to figure out how to write a LOOP, I usually either go to CLHS 6.1 or to PCL's LOOP chapter
<dnhester26>
thuna`: thanks, I read the dictionary entry but thought that reading the entry in 6.1 would take a very long time. I think an example is needed, which I'm adding now
King_julian has quit [Ping timeout: 244 seconds]
chomwitt has quit [Ping timeout: 264 seconds]
dnhester26 has quit [Remote host closed the connection]
mishoo has joined #commonlisp
dnhester26 has joined #commonlisp
eddof13 has joined #commonlisp
ingeniot has quit [Ping timeout: 256 seconds]
rainthree has joined #commonlisp
<aeth>
Yes, hash-table syntax in LOOP is one of the weaker parts.
<aeth>
In that it makes no sense.
cmack has joined #commonlisp
srji has quit [Ping timeout: 252 seconds]
<aeth>
I'd personally say that the hash-table syntax and the inability to bind multiple values without first consing up a small list are the two main flaws of LOOP.
<beach>
I find discussions about the shortcomings of the language quite useless. There is nothing that can be done about them, and in most cases, there would be no consensus anyway.
<aeth>
Generally, it means avoid LOOP in this situation (hash tables), but this is an edge case where you need it.
<aeth>
When the language is large, shortcomings mostly just means use one of the other 3 ways to do it
<aeth>
Unfortunately, it makes LOOP on hash tables particularly ugly when actually necessary because now one lacks the familiarity with it
spdegabrielle has joined #commonlisp
m5zs7k has quit [Read error: Connection reset by peer]
<ixelp>
loop | Common Lisp (New) Language Reference
<dnhester26>
with 3 examples
<dnhester26>
beach: I'm assuming that's the right syntax for the loop for value only, it worked on SBCL, but I'm not sure if that's necessarily according to the standard or more lenient...
<beach>
Looks good!
<beach>
It is the right syntax.
<dnhester26>
by the way, is that correct also according to the standard? I read somewhere where you wrote the loop in most implementations is an MIT loop and not ANSI CL loop
<dnhester26>
ok, thanks
<dnhester26>
not the* ANSI CL loop
<beach>
Well, MIT LOOP is technically an implementation of the ANSI Common Lisp LOOP macro, but it is not always correct, as I recall.
<dnhester26>
ah ok, my bad
wobbol has joined #commonlisp
<beach>
We believe Khazern is an implementation of the LOOP macro that is closer to the standard.
cdegroot has joined #commonlisp
notzmv has quit [Ping timeout: 276 seconds]
<dnhester26>
I think I may change the whole build process from the lisp-docs website and just do it with common lisp from scratch, that way people will be more willing to collaborate (I think)
<dnhester26>
Khazern is the project you wrote?
<dnhester26>
closer because you are not sure if there are any bugs? or are there deliberate changes from the standard?
<beach>
It is the LOOP implementation I initially wrote for SICL, but it is now maintained by yitzi.
<beach>
We are unaware of any bugs in Khazern, but one can never be too sure.
<dnhester26>
if all the implementations or most are using the MIT LOOP would that not be a breaking change if they actually conform with the standard and it breaks some people's code? or are the changes very minor?
<dnhester26>
ah ok
<beach>
There are very few uses of the LOOP macro that would work in MIT LOOP and not in Khazern and vice versa.
<dnhester26>
Has there been any interest from implementations to integrate it into their code bases? I imagine it'll be difficult according to what I understood (perhaps wrongly) from your ELS presentation on bootsraping and having a CL image available if they don't have CLOS available when they are compiling the loop though
notzmv has joined #commonlisp
<dnhester26>
So when you said closer what did you mean?
<beach>
There has been no such interest, for the reason you mentioned.
<beach>
I meant that we tried very hard to make it conform to the standard and nothing else, but we are obviously not sure whether there are any remaining defects.
<dnhester26>
Are there any plans for SICL to eventually be a usable implementation more than parts of the language? meaning, will we eventually be able to use it for development and production? if that's the case, would SICL also allow for the debugging capabilities you were discussing or would those be breaking from ANSI CL?
<dnhester26>
ah ok, so that's what you meant by closer; thanks for clarifying
<beach>
Some of our libraries have no such bootstrapping issues, like Constrictor, but I am pretty sure there won't be any interest for it either, other than perhaps for some new Common Lisp implementation. Then, the newest one (alisp) doesn't use Constrictor either.
cdegroot has quit [Ping timeout: 252 seconds]
wobbol has quit [Ping timeout: 252 seconds]
entro has quit [Ping timeout: 252 seconds]
cdegroot has joined #commonlisp
<beach>
dnhester26: I hope I have enough remaining life expectancy to finish SICL (the Common Lisp implementation) in some form (perhaps not super optimized), and it will then definitely include the debugging capabilities I wrote about.
<beach>
Allowing a debugger doesn't break ANSI conformance.
entro has joined #commonlisp
wobbol has joined #commonlisp
<dnhester26>
From what's available now, what's needed to make a full implementation? I think I remember form your presentation that only about 2000 lines of code were for actual binary code generation in the compiler. What's left?
<beach>
Finishing the new bootstrapping procedure, and generating native code from ASTs essentially.
<dnhester26>
I was thinking of maybe giving a try for generating a wasm implementation, though since it's just a couple of thousand lines of code, could it be like a protocol where we change the "compilation target" and generate wasm instead of regular assembly code?
<beach>
The current plan is to create Posterior which is a backend library, so code generation would go in that library.
<dnhester26>
that's the "generating native code from ASTs" part?
<beach>
Posterior would accept a control graph at a level roughly like MIR (Medium-level Intermediate Representation) and generate code.
<beach>
I still have no idea how to extract the AST-to-MIR part to a library.
<beach>
... so that's still part of SICL.
<beach>
The current plan for SICL is to reduce it to pretty much the bootstrapping procedure, and to allow people to plug in a few parameters, like object representation and calling conventions, and then create an executable from all the libraries, using the bootstrapping procedure. But we are not there yet.
<dnhester26>
does that mean sicl would be like an implementation generator based on those parameters more than a specific implementation itself?
<beach>
I hope so, but it will not be trivial.
pabs3 has quit [Ping timeout: 248 seconds]
<beach>
So first, we will create a specific implementation.
<beach>
I hope that will generate enough interest so that others will work on it.
<beach>
We have come quite far with the extracted libraries.
<dnhester26>
is object representation above referring to the binary file format specific to a target architecture and OS?
<beach>
No, it refers to how objects (especially standard objects) are represented in memory.
pabs3 has joined #commonlisp
<dnhester26>
calling conventions is about how to call a function? why does that change? time vs space efficiency?
<beach>
People have different preferences. Like register usage, stack usage, where to put multiple values, etc.
<dnhester26>
thanks for the explanations :)
zxcvz has joined #commonlisp
<beach>
Pleasure. I hope it made sense.
amb007 has quit [Ping timeout: 260 seconds]
zxcvz has quit [Client Quit]
amb007 has joined #commonlisp
<dnhester26>
yeah, it did, thanks
shawnw has joined #commonlisp
kurfen has quit [Ping timeout: 248 seconds]
wobbol has quit [Ping timeout: 246 seconds]
wobbol has joined #commonlisp
mishoo has quit [Ping timeout: 252 seconds]
chomwitt has joined #commonlisp
ski has joined #commonlisp
jonatack has quit [Ping timeout: 255 seconds]
mgl_ has joined #commonlisp
mishoo has joined #commonlisp
jonatack has joined #commonlisp
alcor has joined #commonlisp
varjag has quit [Quit: ERC 5.5.0.29.1 (IRC client for GNU Emacs 29.3)]
chomwitt has quit [Ping timeout: 248 seconds]
jon_atack has joined #commonlisp
jonatack has quit [Ping timeout: 255 seconds]
wacki has joined #commonlisp
dnhester26 has quit [Remote host closed the connection]
jon_atack has quit [Ping timeout: 255 seconds]
wobbol has quit [Ping timeout: 276 seconds]
eddof13 has quit [Quit: eddof13]
amb007 has quit [Read error: Connection reset by peer]
amb007 has joined #commonlisp
pabs3 has quit [Ping timeout: 264 seconds]
dnhester26 has joined #commonlisp
amb007 has quit [Ping timeout: 255 seconds]
amb007 has joined #commonlisp
dnhester26 has quit [Read error: Connection reset by peer]
dnhester26 has joined #commonlisp
Akbar-Birbal has left #commonlisp [#commonlisp]
eddof13 has joined #commonlisp
pabs3 has joined #commonlisp
jonatack has joined #commonlisp
wobbol has joined #commonlisp
dnhester26 has quit []
kevingal has joined #commonlisp
jonatack has quit [Ping timeout: 252 seconds]
semarie has quit [Quit: quit]
spdegabrielle has quit [Quit: Connection closed for inactivity]
jonatack has joined #commonlisp
semarie has joined #commonlisp
amb007 has quit [Ping timeout: 272 seconds]
amb007 has joined #commonlisp
chomwitt has joined #commonlisp
jon_atack has joined #commonlisp
jonatack has quit [Ping timeout: 252 seconds]
JuanDaugherty has quit [Quit: JuanDaugherty]
varjag has joined #commonlisp
shmorg83 has quit [Ping timeout: 248 seconds]
mgl_ has quit [Ping timeout: 265 seconds]
rakka has quit [Remote host closed the connection]
attila_lendvai_ has joined #commonlisp
rakka has joined #commonlisp
attila_lendvai has quit [Ping timeout: 264 seconds]
mgl_ has joined #commonlisp
cage has joined #commonlisp
cage has quit [Excess Flood]
cage has joined #commonlisp
kurfen has joined #commonlisp
notzmv has quit [Read error: Connection reset by peer]
rakka has quit [Remote host closed the connection]
rakka has joined #commonlisp
notzmv has joined #commonlisp
notzmv has quit [Remote host closed the connection]
pve has quit [Quit: leaving]
pestctrl` has joined #commonlisp
drbdf has joined #commonlisp
pestctrl` has quit [Client Quit]
pestctrl has joined #commonlisp
eddof13 has quit [Quit: eddof13]
cage has quit [Remote host closed the connection]
mishoo has quit [Ping timeout: 255 seconds]
mwnaylor has joined #commonlisp
srji has joined #commonlisp
mgl_ has quit [Ping timeout: 260 seconds]
srji has quit [Ping timeout: 248 seconds]
hayaliali has joined #commonlisp
triffid has quit [Ping timeout: 260 seconds]
alip has quit [Ping timeout: 276 seconds]
hayaliali is now known as alip
apac has joined #commonlisp
amb007 has quit [Read error: Connection reset by peer]
eddof13 has joined #commonlisp
triffid has joined #commonlisp
amb007 has joined #commonlisp
amb007 has quit [Ping timeout: 276 seconds]
wacki has quit [Quit: My Mac has gone to sleep. ZZZzzz…]
triffid has quit [Remote host closed the connection]
triffid has joined #commonlisp
Everything has joined #commonlisp
drbdf has quit [Ping timeout: 260 seconds]
chrcav has quit [Ping timeout: 276 seconds]
alcor has quit [Read error: Connection reset by peer]
chrcav has joined #commonlisp
anticomputer has quit [Remote host closed the connection]
anticomputer has joined #commonlisp
eddof13 has quit [Quit: eddof13]
livoreno has joined #commonlisp
livoreno has quit [Remote host closed the connection]
varjag has quit [Ping timeout: 252 seconds]
livoreno has joined #commonlisp
jon_atack has quit [Read error: Connection reset by peer]