1]<br>What is the meanning of warning 186 ?<br>For instance in <br>WARNING ! [186] conflict between set! @ TabuList and set! @ collection is not allowed for...
In the following trace, nth+(2,700) and l[3] := 23 should both return "out of scope" at least in interpreted mode and in safety 0. This lack of bound checking ...
Two sets s1 and s2 are equal iff they contain the same elements.<br><br>Is there a claire function testing pointers equality? This is because for some reason I...
Etienne found the solution:<br>Core/Address(any) (at the end of function.cl)<br><br>[d:\elab\mmethodes\wk]claire3 -n<br>-- CLAIRE run-time library v 3.2.2 [os:...
Where removing a value from an empty set, a stange behavior can occur:<br>[d:\elab\mmethodes\wk]claire3 -n<br>-- CLAIRE run-time library v 3.2.2 [os: ntv,...
Argl ! :-(<br><br>I offer my humblest apologies: I introduced a stupid bug while fixing the previous stupid bug. The point is that my testing environment is ...
1]<br>What is the meanning of warning 186 ?<br>For instance in <br>WARNING ! [186] conflict between set! @ TabuList and set! @ collection is not allowed for...
There is a method for this in CLAIRE !<br><br>identical?(s1,s2)<br><br>It turns out that there was a bug, which is fixed in v3.2.22. However, do not use...
many thanks for such a prompt reply !!<br>I will test fast-choco-claire3 this week end.<br><br>just to check: are you sure you uploaded that new version of ...
Well I got delayed when posting the v3.2.22 release with two questions (the zip filed was loaded exactly when Francois posted his comment :-) )<br><br>Here my...
>> (b) the characters < and > are separators in CLAIRE3. <br>>> Thus <> is not a valid name for a property (for instance). <br><br>Yes,...
Hi Yves,<br><br>I DO GET SOME FASTCALLS IN THE GENERATED CODE !!!<br><br>thank you for this wonderful new global variable (FASTCALLSTINKS): it is nice to get...
Hello,<br><br>I believe that once you have defined a method as interface(f) you don't need the abstract declaration. It is extensible as long as you respect...
Hello,<br><br>the abstract declaration seems to be needed:<br><br>in the following code, the call to f in g is a good fast dispatch call, while the call to f...
There is a new class "collection" in claire3:<br><br>* when we want to define a class representing a collection of objects that can be iterated (say a class ...
The following code cannot be executed in verbose 0 (~I in trace does not work). It works if the trace is replaced by a printf.<br><br>A:list<integer> := ...
Hello everyone,<br><br>A new release is there. Thanks to Arnaud, Francois and Thierry, a lot of bugs have been fixed:<br><br>(1) performance tuning for list...
~I is NOT allowed in trace. This is described in the doc. One could argue that the error message is not friendly ...<br><br>A happier note: I have created a...
Float printing does not work under the inspector:<br>-- CLAIRE run-time library v 3.2.24 [os: ntv, C++:MS VC++ ] --<br>claire> A <: object(f:float = ...
Yves, I am very sorry to post this message after all the hard work you have been doing: there are still a few cases in which fast dispatch does not seem to...
Hello everyone,<br><br>Francois should not feel very, very sorry ... his work is very,very useful, although I feel very,vey dumb :-). Anyway, another simple ...
It fixes the following problems:<br><br>- float printing (cf ThB's remarks + nicer<br> printing of 12.99)<br><br>- new(C) is now equivalent to C() as...
a small bug with the latest Claire3: when trying to compile the code below, I get in the generated C code:<br><br> v_list = GC_OBJECT(list,x->l2);<br> g0002...
For Claire 3.2.18 at least<br><br> You can use "-" in your module name (as it's not yet prohibe), but when you try to compile this module <br> 1/ the .h is...
In the following trace, we can't define an empty list of the desired type:<br><br>-- CLAIRE run-time library v 3.2.26 [os: ntv, C++:MS VC++ ] --<br>claire> ...