LtU Forum

/join the #proglangdesign channel on Freenode

The #proglangdesign channel on Freenode hosts many discussions about programming language design, compilers, interesting algorithms, and anything else useful for language implementation. The channel is pretty active; say 'hello' if you join.

Many of the members are working on compilers, interpreters, whether they are just experimenting to learn, or advancing the state of the art.

Come join us!

More information can be found at our community website, which includes a list of our projects; submit a pull request to add your own.

Just entertainment: Click, click, click!

Just a silly website I thought you people might enjoy. Click, click, click!

Is Datalog negation(¬) similar to the built-in predicate (≠)?

I was reading "Principles of Database & Knowledge-Base Systems, Vol. 1" by Jeffrey D. Ullman. There is a chapter about Datalog negation and as I was seeing the problems of negation I kept thinking that using the predicate ≠ would solve those problems.


bachelor(X) :- male(X) & ¬married(X,Y).

would become:

bachelor(X) :- male(X) & married(Y,Z) & X ≠ Y.

but then I see the following:

p(X) :- r(X) & ¬q(X).
q(X) :- r(X) & ¬p(X).

The problem is this has 2 minimal models and if I'm not mistaken so does this:

p(X) :- r(X) & q(Y) & X ≠ Y.
q(X) :- r(X) & p(Y) & X ≠ Y.

Is there an equivalence between these 2 operators? If so, did I miss it or is it not mentioned that it's unsafe to use ≠ with recursion?

Please submit to LIVE! 2017 (SPLASH Vancouver)

LIVE 2017 aims to bring together people who are interested in live programming. Live programming systems abandon the traditional edit-compile-run cycle in favor of fluid user experiences that encourage powerful new ways of “thinking to code” and enable programmers to see and understand their program executions. Programming today requires much mental effort with broken stuttering feedback loops: programmers carefully plan their abstractions, simulating program execution in their heads; the computer is merely a receptacle for the resulting code with a means of executing that code. Live programming aims to create a tighter more fluid feedback loop between the programmer and computer, allowing the computer to augment more of the programming process by, for example, allowing programmers to progressively mine abstractions from concrete examples and providing continuous feedback about how their code will execute. Meanwhile, under the radar of the PL community at large, a nascent community has formed around the related idea of “live coding”—live audiovisual performances which use computers and algorithms as instruments and include live audiences in their programming experiences.

We encourage short research papers, position papers, web essays, tool demonstrations (as videos), and performance proposals in areas such as:

Recent work in REPLs, language environments , code playgrounds, and interactive notebooks.

  • Live visual programming.
  • Programming by example.
  • Programming tools for creative experiences and interactive audio visual performances.
  • Live programming as a learning aid.
  • Fluid debugging experiences
  • Language design in support of the above.

Submissions are due on August 1st and will go through HotCRP @ The workshop is open to various kinds of media: you can write a traditional short paper (PDF), a web essay with embedded videos, a narrated video, or whatever else you think can explain your work best! Content should be consumable in 30 minutes to an hour of casual reader’s time, which means around 5-10 pages for a paper, a video from 10-20 minutes (assuming the viewer would need to pause to contemplate), and essays of around a few pages length. Video and non-paper submissions can by listed as URLs (e.g. to a web page, file locker, or streaming site) in the submission’s abstract.

Any questions or trouble with submitting, please contact This CFP is hosted @

C runtime type info gimmick which supports scripting

I am experimenting with runtime type information in C, so that values in memory have types, rather like Lisp and Smalltalk values have types, and not just variables. This post is intended to be fun. (The idea is for you to enjoy this, and I don't want to argue. But I might respond to remarks with short dialogs in lieu of argument, since that would be fun.) The rest of this is explanation and some detail.

In an earlier post I described a kind of heap I call a vat. It has the effect of associating metainfo with every block of memory allocated, in a different position in memory. (There is a constant time arithmetic calculation that turns a block pointer into a metainfo pointer, and vice versa; so the main penalty in access is a cache line miss if this happens more often when lacking adjacency to its block.) A continuous memory block is a rod, while the fixed size metainfo describing it is a knob. A counted reference to rod (or any point within) is a hand. On a 64-bit platform, a knob is 128 bits, a hand is also 128 bits, and rods vary in size.

Some bits of a hand are a copy of bits in the knob, so a reference is only valid when it agrees with the current self description of an object, particularly the current 16-bit generation number, so dangling refs are seen. Of the bits copied, another 20 bits are the id of the object's plan, which describes the type, layout, and any dynamic behavior if you also supply methods for that type in hashmap entries. You can think of plan as morally identical to class in Smalltalk, but initially I was aiming for perfect format description on a field by field basis, with recursion grounded in primitive native types. I wanted to write centralized code to debug print anything based on the plan. But then I sort of noticed I would be able to write Smalltalk style code that works on the C objects, if you were willing to call a dynamic api to send messages (method selectors with parameters).

If you point at a subfield somewhere inside the middle of a rod, you can eventually figure out exactly what it is by drilling down in the rod's plan, after fetching the plan for the id in the rod's metainfo knob. But since this lookup is offset-based, it aliases anything with the same offset. That is, a struct named foobar that contains a foo and then a bar will have the same offset for foobar and foo. If you send a message, did you mean to foobar or to foo? I would start a search in method tables for the largest containing object, then work through nesting until I found a match.

At the moment I am writing metainfo by hand, but generating it from C declarations later is the idea. On 64-bit platforms a plan is 256 bits, and each field requires 128 bits, plus however long the string names get, and method tables if there are any. When a field is composed of bit fields, I use a plan that says all its fields are bitfields, and then each field describes bit sizes and shifts instead of byte sizes and offsets.

Even if most of an app is comprised of static code, I will want to write dynamic and interactive scripts to explore runtime state in tests, and to inspect things at runtime in ways I did not anticipate earlier. I would normally think of glomming on a scripting language to do this, on the side. But I can use C as the scripting language instead, if memory state is universally annotated this way with type metainfo.

What would be involved in moving logic beyond FOL?

Carl Hewitt has opined that first-order logic (FOL) should not be regarded as logic of choice for various people such as computer scientists (e.g., he says "First-order theories are entirely inadequate for Computer Science" in a reply to me in Mathematics self-proves its own Consistency (contra Gödel et. al.). He recommends instead his Inconsistency-Robust Logic for mathematicians, computer scientists and others.

What would it take to really change our perspective to be grounded in a logic other than FOL? I think that looking at John Corcoran's First days of a logic course provides the beginnings of an answer (the article treats Aristotelian logic rather than FOL): an alternative logic should provide a natural basis grounding all of the concepts that we expect a new student of logic to grasp, and a rival to the kind of traditional logic course based on first-order logic should be comparably good or better for building on.

What would an introductory logic course look like for, e.g., Inconsistency-Robust Logic? Would it start with Aristotelian logic or would it start with something else? What would having taken such a course be good for, as a prerequisite?

Restructor: Full Program Automatic Refactoring

Wrote up a description of some personal research I did 2005-2010:

This algorithm will take an arbitrary program (including side effects) and refactor (create and inline functions) until the code contains no more under-abstraction (copy paste code) or over-abstraction. It is an implementation of code-compression: it finds a set of functions representing your code such that the total AST nodes are minimized.

I had originally intended this as a new way of programming, where you could modify your code using copy paste to your hearts content, and leave the annoying work of abstraction to the computer. I now realize that is a bit naive, and note some reasons why this may not be ideal on the page above. Still, an interesting idea that I've not seen before, would love to hear y'alls expert opinions :)

For reference, I posted about this idea on LTU in 2004 under the topic "abstractionless programming".

Free links to all (or practically all) recent SIGPLAN papers


ACM OpenTOC is a unique service that enables Special Interest Groups to generate and post Tables of Contents for proceedings of their conferences enabling visitors to download the definitive version of the contents from the ACM Digital Library at no charge.

Downloads of these articles are captured in official ACM statistics, improving the accuracy of usage and impact measurements. Consistently linking to definitive versions of ACM articles should reduce user confusion over article versioning.

SW verification continues

There is new release of the Albatross compiler available.

The Albatross programming language shoots at making verified software construction available to everybody.

The language description has been completely updated. The document describes how to get the compiler.

The previous releases already contained induction and recursion and even inductive sets.

The most important new feature is abstract data types. Abstract data types are realized by abstract classes, abstract functions and abstract properties. By using the abstraction it is possible to verify a lot of properties which can be inherited by any type which satisfies the concept of the abstract data type.

The design of the language is still an ongoing activity any comment. Hints, issue reports etc. are welcome.


Finding Solutions vs. Verifying Solutions

[Edit] Due to a conversation below, it turned out that sets P and NP are commonly defined differently than I considered them in my initial report. Updated report with the same content, but without this inconsistency can be found here: Finding Solutions vs. Verifying Solutions. I also changed the blog title from "P is not equal to NP" to the above.

Below is the initial, but obsolete post kept for conversational reasons:

In this short report, I argue that P is not equal to NP.

Summary: an approach is made by analyzing functions and their inverses mappings between domain and codomain elements. Drawn conclusions state that verifying data is an inverse function of finding solutions to the same problem. Further observation implies that P is not equal to NP.

Any thoughts?

XML feed