SISC

Settings | Report Duplicate

6

I Use This!

Activity Not Available

Commits : Listings

Analyzed 4 months ago. based on code collected over 1 year ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Dec 01, 2021 — Dec 01, 2022 Showing page 10 of 107 Search / Filter on:
* use xsltproc to build manual - the Java XSLT processors seem incapable of handling the docbook stylesheets whereas xsltproc has consistently produced good results, so changing to xsltproc at least gives us a fighting chance of getting the manual to build - also change default location of docbook stylesheets to where they tend to live under debian	mradestock	More...				about 18 years ago
* removed duplicate export of \|stack-trace\| from debugging module	mradestock	More...				about 18 years ago
added java-io library, as per bug 1455041 "No conversion functions between ports and Java I/O types"	acolomba	More...				about 18 years ago
* return #f from continuation-nxp and continuation-stk when value is absent previously we returned '(), which is just weird	mradestock	More...				about 18 years ago
* fix error reporting in record module - this assumed a srfi-23 style error procedure rather than SISC's builtin - produce nicer error messages by showing the record type name instead of the raw record type	mradestock	More...				about 18 years ago
* cross-reference Type System docs from S2J docs	mradestock	More...				about 18 years ago
* clear the nxp last Thus any errors we encounter will still find the nxp and won't have to rely on the lxp. The only place where we still need the lxp is in NativeProcedure since calls to general primitives may modify the nxp and hence we can't null it after the call.	mradestock	More...				about 18 years ago
* don't synchronise on DynEnv when accessing Parameters there is no need to do so since DynEnvs are bound to threads	mradestock	More...				about 18 years ago
* force optimisation reversion on error this fixes bug 1453863 ("error-k too shallow due to optimisation")	mradestock	More...				about 18 years ago
* refactoring: simplified error-k creation in Interpreter.error() this elminates yet another place in the code with a dependency on the CallFrame structure	mradestock	More...				about 18 years ago
* refactoring: move vlr safe-making into its own Interpreter.makeSafe() method Currently there is only one place - Interpreter.setVLR() - where this is called, but that will change soon.	mradestock	More...				about 18 years ago
* refactoring: introduce Interpreter.setFailureContinuation() This allows changes to the CallFrame structure to be more contained and it also means we can make Interpreter.createFrame() private	mradestock	More...				about 18 years ago
* simplified FixedAppExp_0.getValue() the code was rather hard to understand, contained cruft and was slightly wrong	mradestock	More...				about 18 years ago
* moved sisc.modules.Debugging.SISCExpression to sisc.data.ExpressionValue - imho, that is a better place for it - also added valueHashCode, valueEquals, getAnnotations	mradestock	More...				about 18 years ago
* remove name propagation from non-free references Lexical and local references never have name annotations and adding them seriously bloats the heap. Hence there is no point having code that attempts to propagate these annotations.	mradestock	More...				about 18 years ago
avoid hash code collision	mradestock	More...				about 18 years ago
propagate annotations to during unoptimisation. This partially solves bug 1447904.	mradestock	More...				about 18 years ago
use the new Compiler.makeFillRib to produce FillRibExps during unoptimisation. This ensures that annotations are propagated correctly and partially solves bug 1447904.	mradestock	More...				about 18 years ago
added helper method to get all annotations as a list	mradestock	More...				about 18 years ago
- refactoring: moved FillRibExp creation into a function, so that both application() and compileLetrec() can use it - annotate fixedCall expressions, and annotate AppExp inside application() - this is so we when we call application() during unoptimisation all the annotation tracking is taken care of. There is some redundancy here when application() gets called by the compileApp(), since it adds some annotations anyway, but I can't think of a quick fix to that.	mradestock	More...				about 18 years ago
There is nothing optimistic about a FreeReferenceExp	mradestock	More...				about 18 years ago
forgot to commit that earlier on ... - rather than getting OptimisticHost.alter to call replaceWith.setHost, we do that in FixedAppExp_0.revert.	mradestock	More...				about 18 years ago
when altering an optimistic sub-expression, only cons up a new FillRibExp if the new rand is non-immediate	mradestock	More...				about 18 years ago
add assertion	mradestock	More...				about 18 years ago
in alter() - there is no need to call replaceWith.setHosts since that will have been done by the compiler when it produced replaceWith - there is no need to call replaceWith.setHost - we do that in FixedAppExp_0.revert	mradestock	More...				about 18 years ago
make LetrecEval an OptimisticHost. This is rarely ever of any use, but it's the only expression that can be an OptimisticHost and wasn't.	mradestock	More...				about 18 years ago
LambdaExp cannot be OptimisticHosts since their body always gets evaluated elsewhere, i.e. in the Closures.	mradestock	More...				about 18 years ago
there is no need to force PROGRAMs to be in REALTAIL position	mradestock	More...				about 18 years ago
some tidying up and minor refactoring	mradestock	More...				about 18 years ago
optimisation: 'if' branches are always allowed to be OptimisticExpression since they are guaranteed to be invoked from IfEval, which is an OptimisticHost.	mradestock	More...				about 18 years ago

←
1
2
…
6
7
8
9
10
11
12
13
14
…
106
107
→