I had the following memleak in mind: http://www.singular.uni-kl.de:8002/trac/ticket/183 is it fixed, too?
I officially declared Singular "bug free" (tongue in cheek, of course) > Despite its ironically, you've made my day! Since a while I'm dominating their bugtracker, reported tons of issues (about 40 were considered as bugs); I guess some Singular devs hate me already. At the same time we looked only at a handful of libraries we are using in our project, so there are much more bugs lurking in Singular. However, most of the bugs I discovered are minor (not all). One of my objections is, that they seem not actively look out for bugs from time to time, like "if I can't see the bug, it is not there"; using very often only just single examples as tests. Many by me observed bugs are about missing or incorrect handling of corner cases. I don't think Sage has a global "assertlevel", though. Ok, thanks! Jack Am Dienstag, 13. Mai 2014 23:26:36 UTC+2 schrieb Simon King: > > Hi, > > On 2014-05-13, kro...@uni-math.gwdg.de <javascript:> < > kro...@uni-math.gwdg.de <javascript:>> wrote: > > thanks for the info about checking the memory management, its quite > useful! > > It relates to the initial question about active testing, right? > > Maybe it will help me to pinpoint some memory leak bugs in Singular. > > This has already happened. It resulted in a patch that we applied to our > singular spkg, and is now used upstream too, if I recall correctly. And > when Sage's test suite finally passed with the debug version of Sage, I > officially declared Singular "bug free" (tongue in cheek, of course). > > > In Singular a similar concept is implemented for the new ASSUME keyword. > > So if I call in Singular "ASSUME(3, 1==2);" '1==2' is only checked if > the > > (package-local) *assumeLevel* is >=3. > > > > Since I'm new to sage, > > I'm questioning if sage has a similar framework ( if you are picky, the > > question belongs to sage-support or to ask.sage ) > > and if sage does not have such a framework, I suggest to think about it > and > > then its probably in the right place at sage-devel > > What often is done is to provide an optional argument "check" for a > constructor, and to skip tests if "check=False". However, this is (and > should be!) a case-by-case decision, since it is mainly used if a > constructor is called from a function that has already asserted that the > input for the constructor is valid, so that skipping the assertion in > the constructor is ok. > > I don't think Sage has a global "assertlevel", though. > > Best regards, > Simon > > > -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
