I have just released EiffelFox-0.7, which will work with the next release of SmallEiffel. Dominique sent me his current snapshot of SmallEiffel with some...
Darren Hiebert
darren@...
Jun 3, 2000 1:54 am
1758
... Hope that too ! ... I have downloaded both fox and EiffelFox. I have a problem during installation of EiffelFox, running .configure ... EiffelFox # make cd...
Dominique Colnet
Dominique.Colnet@...
Jun 3, 2000 7:54 am
1759
I would like to request that some option be added to the SmallEiffel compiler to force it to compile every feature of every class referenced by the root class,...
Darren Hiebert
darren@...
Jun 3, 2000 5:39 pm
1760
I have removed a few bugs in my flat-short form beautifier (Usage: short -html2 array | sfb > array.html): http://mageos.ifrance.com/bouchard/short/sfb If you...
Jacques Bouchard
bouchard@...
Jun 3, 2000 6:56 pm
1761
... Darren> [...] However, unless I were to write an application which Darren> calls every one of the more than 1200 routines, I cannot Darren> guarantee that...
jweirich@...
Jun 3, 2000 7:43 pm
1762
... The way I do this is to run short on every class. This checks all the class code. It doesn't do any more than check syntax and links to other classes,...
Oliver Elphick
olly@...
Jun 4, 2000 12:20 am
1763
The class comment of SmallEiffel/lib_std/numeric.e says: This class describes a ring (same definition as in OOSC p523 and in EiffelBase/kernel/numeric.e) So it...
Jacques Bouchard
bouchard@...
Jun 4, 2000 10:07 pm
1764
... Considering this is it is already weird to call the class "NUMERIC" since mathematematical Rings need not to be numbers (nor matrices) at all. NUMERIC also...
Pascal Poncin
Pascal.Poncin@...
Jun 5, 2000 6:42 am
1765
... How do you guarantee that routines that have been called 1 (or n) time(s) work properly? If the sun rises on 10000 mornings: what about the 10001st...
Pascal Poncin
Pascal.Poncin@...
Jun 5, 2000 6:44 am
1766
... Hmm... don't you have a test suite that calls every routine at least once? (Just kidding.) Seriously, it would certainly be helpful to have a coverage...
Durchholz, Joachim
Joachim.Durchholz@...
Jun 5, 2000 6:51 pm
1767
... I can imagine such a test suite (using either Jim's EiffelUnit or ISE EiffelTest) for batch libraries, but how do you write such a test suite for a GUI...
Eric Bezault
ericb@...
Jun 5, 2000 7:25 pm
1768
Hai Eric, ... <delphi mode on> Well, here's more or less how to do it with Delphi. 1. In Delphi any application, including dll's can create forms. So my test ...
Berend de Boer
berend@...
Jun 5, 2000 7:52 pm
1769
Hai SmallEiffel, Unless most other problems SIGCHLD is not ignored by default it seems. Is this by design? Isn't it better to leave it alone? There are more of...
Berend de Boer
berend@...
Jun 5, 2000 7:53 pm
1770
Class BASIC_TIME of SmallEiffel -0.76beta3 is bugged. Proof: program BASIC_TIME_BUG, compiled with -no_check, produces on Linux/Intel686: false false false 0 ...
Jacques Bouchard
bouchard@...
Jun 5, 2000 10:37 pm
1771
... 1. NUMERIC is in ELKS95 (without "<", ">", "sign") so it can't be renamed. 2. What ... is the definition of a "numeric"? Is it perhaps not so weird do...
Jacques Bouchard
bouchard@...
Jun 6, 2000 6:21 am
1772
... Two ideas: (1) There is some information about the GUI callbacks in the ACE/CECIL files. Maybe one could exploit this for testing. (2) The GUI classes...
Pascal Poncin
Pascal.Poncin@...
Jun 6, 2000 8:16 am
1773
... Yes, but how much bigger is the difference? Of course the simple number of tests won't improve anything (although there may be some cases - e.g. when no...
Pascal Poncin
Pascal.Poncin@...
Jun 6, 2000 8:24 am
1774
... Yes but that's a practical issue. That doesn't change the fact that "NUMERIC" is a badly chosen name when NUMERIC is intended to model rings. ... Don't ask...
Pascal Poncin
Pascal.Poncin@...
Jun 6, 2000 8:58 am
1775
I wanted to stay out of this one, but ... ... I guess it would be more satisfactory if we could write, say x + {DOUBLE}.zero and I would be happy if this kind...
Prager, Daniel (CMIS,...
Daniel.Prager@...
Jun 6, 2000 9:47 am
1776
Hi there. I've got a question concerning the GC (assuming that this is the reason for what I'm seeing). I've written a program which scans through a file (one...
Philip Malin
psm@...
Jun 7, 2000 1:47 am
1777
... I assume you don't need `list´ after the call to `output_data´ anymore? If so, have you tried to trigger a garbage collecting cycle manually after...
Andreas Leitner
nozone@...
Jun 7, 2000 2:16 am
1778
Hi there. ... Yup, that's correct. ... No, I haven't tried that yet. I'll give it a go and see what happens. Cheers. ... Phil Malin, Senior Systems Engineer...
Philip Malin
psm@...
Jun 7, 2000 2:46 am
1779
... Pascal> Of course the simple number of tests won't improve Pascal> anything (although there may be some cases - e.g. when no Pascal> garbage collection is...
jweirich@...
Jun 7, 2000 4:57 am
1780
Hey I'm totally newbie to smalleiffel and eiffel in general, but I really tried to figure this problem out without success. So I wonder if someone could...
Sami Lappeteläinen
sami@...
Jun 7, 2000 5:27 am
1781
... This may be solvable just you don't include the information one needs to help. Without the source how should I know where you missed an end. Remember, that...
Friedrich Dominicus
Friedrich.Dominicus@...
Jun 7, 2000 5:44 am
1782
... Totally true. I somehow just forgot that also the If-structure need ending. I really can't remember how I came up with the idea about non-ending...
Sami Lappeteläinen
sami@...
Jun 7, 2000 5:52 am
1783
You need to include the source code....
David Broadfoot
david@...
Jun 7, 2000 5:52 am
1784
I sent this message to Andreas and forgot to send it out to the list (no wonder I wasn't receiving it). Additional note (not in my reply to Andreas): I believe...
Philip Malin
psm@...
Jun 7, 2000 7:30 am
1785
... Even any software satisfies the specification "The system does something.". (considering NOPs as doing something) ... It would be interesting to know what...
Pascal Poncin
Pascal.Poncin@...
Jun 8, 2000 7:36 am
1786
... Of course I do believe that testing is a good thing. I simply wanted to point out that one should always ask what sort of testing gives what answers. It is...