Marcus Denker

Gradual typing for Smalltalk

Being able to combine static and dynamic typing within the same language has clear benefits in order to support the evolution of prototypes or scripts into mature robust programs. While being an emblematic dynamic object-oriented language, Smalltalk is lagging behind in this regard. We report on the design, implementation and application of Gradualtalk, a gradually-typed Smalltalk meant to enable incremental typing of existing programs. The main design goal of the type system is to support the features of the Smalltalk language, like metaclasses and blocks, live programming, and to accommodate the programming idioms used in practice. We studied a number of existing projects in order to determine the features to include in the type system. As a result, Gradualtalk is a practical approach to gradual types in Smalltalk, with a novel blend of type system features that accommodate most programming idioms. A practical gradual type system for Smalltalk that supports a smooth path from untyped to typed code.A novel combination of several typing features, with some interesting interactions.A discussion of some implementation tradeoffs and challenges for a gradual type system in a live programming environment like that of Smalltalk.An initial validation of the type system through typing several Smalltalk projects.

Read-only execution for dynamic languages

Supporting read-only and side effect free execution has been the focus of a large body of work in the area of statically typed programming languages. Read-onlyness in dynamically typed languages is difficult to achieve because of the absence of a type checking phase and the support of an open-world assumption in which code can be constantly added and modified. To address this issue, we propose Dynamic Read-Only references (DRO) that provide a view on an object where this object and its object graph are protected from modification. The read-only view dynamically propagates to aggregated objects, without changing the object graph itself; it acts as a read-only view of complex data structures, without making them read-only globally. We implement dynamic read-only references by using smart object proxies that lazily propagate the read-only view, following the object graph and driven by control flow and applied them to realize side-effect free assertions.

Language boxes: bending the host language with modular language changes

As domain-specific modeling begins to attract widespread acceptance, pressure is increasing for the development of new domain-specific languages. Unfortunately these DSLs typically conflict with the grammar of the host language, making it difficult to compose hybrid code except at the level of strings; few mechanisms (if any) exist to control the scope of usage of multiple DSLs; and, most seriously, existing host language tools are typically unaware of the DSL extensions, thus hampering the development process. Language boxes address these issues by offering a simple, modular mechanism to encapsulate (i) compositional changes to the host language, (ii) transformations to address various concerns such as compilation and syntax highlighting, and (iii) scoping rules to control visibility of fine-grained language extensions. We describe the design and implementation of language boxes, and show with the help of several examples how modular extensions can be introduced to a host language and environment.

Practical domain-specific debuggers using the Moldable Debugger framework

Understanding the run-time behavior of software systems can be a challenging activity. Debuggers are an essential category of tools used for this purpose as they give developers direct access to the running systems. Nevertheless, traditional debuggers rely on generic mechanisms to introspect and interact with the running systems, while developers reason about and formulate domain-specific questions using concepts and abstractions from their application domains. This mismatch creates an abstraction gap between the debugging needs and the debugging support leading to an inefficient and error-prone debugging effort, as developers need to recover concrete domain concepts using generic mechanisms. To reduce this gap, and increase the efficiency of the debugging process, we propose a framework for developing domain-specific debuggers, called the Moldable Debugger, that enables debugging at the level of the application domain. The Moldable Debuggeris adapted to a domain by creating and combining domain-specific debugging operations with domain-specific debugging views, and adapts itself to a domain by selecting, at run time, appropriate debugging operations and views. To ensure the proposed model has practical applicability (i.e., can be used in practice to build real debuggers), we discuss, from both a performance and usability point of view, three implementation strategies. We further motivate the need for domain-specific debugging, identify a set of key requirements and show how our approach improves debugging by adapting the debugger to several domains. HighlightsWe identify and discuss requirements for developing domain-specific debuggers.We present the Moldable Debugger, a model for developing domain-specific debuggers.We give real-world examples showing the usage of the Moldable Debugger model.We discuss a prototype implementation of the Moldable Debugger model.We discuss three different approaches for implementing debugging operations.

Performance evolution blueprint: Understanding the impact of software evolution on performance

Understanding the root of a performance drop or improvement requires analyzing different program executions at a fine grain level. Such an analysis involves dedicated profiling and representation techniques. JProfiler and YourKit, two recognized code profilers fail, on both providing adequate metrics and visual representations, conveying a false sense of the performance variation root. We propose performance evolution blueprint, a visual support to precisely compare multiple software executions. Our blueprint is offered by Rizel, a code profiler to efficiently explore performance of a set of benchmarks against multiple software revisions.

Modeling features at runtime

A feature represents a functional requirement fulfilled by a system. Since many maintenance tasks are expressed in terms of features, it is important to establish the correspondence between a feature and its implementation in source code. Traditional approaches to establish this correspondence exercise features to generate a trace of runtime events, which is then processed by post-mortem analysis. These approaches typically generate large amounts of data to analyze. Due to their static nature, these approaches do not support incremental and interactive analysis of features. We propose a radically different approach called live feature analysis, which provides a model at runtime of features. Our approach analyzes features on a running system and also makes it possible to grow feature representations by exercising different scenarios of the same feature, and identifies execution elements even to the sub-method level.n We describe how live feature analysis is implemented effectively by annotating structural representations of code based on abstract syntax trees. We illustrate our live analysis with a case study where we achieve a more complete feature representation by exercising and merging variants of feature behavior and demonstrate the efficiency or our technique with benchmarks.

Transactional contexts: harnessing the power of context-oriented reflection

The emerging field of context-oriented programming gives a predominant role to the execution context of applications, and advocates the use of dedicated mechanisms to allow the elegant expression of behavioural adaptations to such context. With suitable reflective facilities, language semantics can be adapted to context by reusing the same context-oriented mechanisms that allow base-level adaptability. This kind of meta-level adaptability, in which the computation model itself becomes adaptable to context, gives rise to context-oriented computational reflection. To explore this idea, we set out to implement a simple software transactional memory system that exploits meta-level adaptability by regarding transactions as contexts, and adapting fundamental system behaviour to such transactional contexts. The implementation is succinct and non-intrusive, giving us an indication of the power lying at the crossroads of context-oriented programming and computational reflection.

Evolving a reflective language: lessons learned from implementing traits

Traits are method groups that can be used to compose classes. They do not have a runtime existence and are conceptually folded into the classes that use them. Traits have been implemented in different languages. While implementing them in Smalltalk, our first reflex was to take advantage of the fact that traits are not run-time entities: we optimized the implementation for space and hence shared methods between traits and classes. However, by doing so we broke the introspective API of Smalltalk.n This paper illustrates a more general problem seen in all reflective systems: the implementation serves both as a model for execution and as the model that is exposed to the programmer. There is a conflict of interests between the information necessary for execution and the information the programmer is interested in. In addition, as soon as the implementation is exposed via reflection, we are not free to optimize. As the complete implementation is visible reflectively, there is no way to hide the optimizations.n Few papers report errors and this is one of them. We report our experience facing the initial API mismatch, which has a significant impact on the system because the language is reflective (i.e., written in itself and causally connected). We present the new introspective API we put in place.

Object spaces for safe image surgery

Long-lived systems rely on reflective self-modification to evolve. Unfortunately, since such a system is at both ends of a causal loop, this means modifications that impact the reflective layer itself can be overly difficult to apply.n This paper introduces ObjectSpaces, a reification of the familiar Smalltalk image as a first-class entity. By confining the system inside an ObjectSpace, we isolate the evolution tools from it, while still giving them reflective access to the confined system. We describe the ObjectSpaces idea, the interface to communicate, inspect, and debug objects contained inside and ObjectSpace, based on a prototype implementation in GNU Smalltalk.

Sista: Saving Optimized Code in Snapshots for Fast Start-Up

Modern virtual machines for object-oriented languages such as Java HotSpot, Javascript V8 or Python PyPy reach high performance through just-in-time compilation techniques, involving on-the-fly optimization and deoptimization of the executed code. These techniques require a warm-up time for the virtual machine to collect information about the code it executes to be able to generate highly optimized code. This warm-up time required before reaching peak performance can be considerable and problematic. In this paper, we propose an approach, Sista (Speculative Inlining SmallTalk Architecture) to persist optimized code in a platform-independent representation as part of a snapshot. After explaining the overall approach, we show on a large set of benchmarks that the Sista virtual machine can reach peak performance almost immediately after start-up when using a snapshot where optimized code was persisted.