Page 8 of 8
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Mon Aug 29, 2016 5:42 pm
by Josef Templ
luowy wrote:Josef Templ wrote:nonsense
do you have the basic courtesy and mutual respect with each other?
your work is work, my work is work also.
you are not my boss, I'm not your employee!
why you say this kind words again and again like a king ?
are you tring to imitate Linus Torvalds ?
Nonsense, in this context, is not a bad word.
A bad word would be if I say "You are a moron", but I don't say that and I don't think that.
But this is Torwald's language. I am only talking about wrong approaches
not about persons.
What would be the most polite way to say that something is very wrong?
- Josef
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Mon Aug 29, 2016 9:46 pm
by cfbsoftware
What would be the most polite way to say that something is very wrong
Say that *you believe* that there are problems and explain what they are and why you believe they are problems. Unless of course the proposal is indisputably "nonsense" to any sane person.
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Wed Aug 31, 2016 12:55 am
by luowy
Bullshit is a one of synonyms for nonsence. It has more meanings.
No further explanation is needed.There are all kinds of bird in the wood.
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Mon Sep 12, 2016 8:30 am
by cfbsoftware
Robert wrote:The in-line BlackBox results are much better than the C# results.
What seems to be happening is that BlackBox is carrying all intermediate results to 80 bits, whereas C# is carrying them to 64.
(Sorry about the delayed reply but things have been hectic here recently)
I suspect the variables were declared as
double in your C# version.
double is a 64-bit data type and all intermediate results would have used 64 bits. However, C# has a built-in
decimal data type which is a 128-bit data type which you can use if you want greater precision. I tried your example with the variables declared as decimal and got the following result for both x and y:
0.3333333333333330
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Mon Mar 06, 2017 7:26 pm
by Robert
Since we decided to handle this issue by updating the PSI document, should this issue now have its status changed to "Closed", and this topic moved to "Rejected"?
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Mon Mar 06, 2017 7:59 pm
by Josef Templ
Robert wrote:Since we decided to handle this issue by updating the PSI document, should this issue now have its status changed to "Closed", and this topic moved to "Rejected"?
A clear Yes from my side. This issue has been discussed extensively and
the docu has been extended significantly.
- Josef
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Tue Mar 07, 2017 12:33 am
by Ivan Denisov
The status also should be "Rejected". Otherwise it will come to the "changeslist" file during our CI system settings.
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Tue Mar 07, 2017 7:49 pm
by Robert
Ivan Denisov wrote:The status also should be "Rejected". Otherwise it will come to the "changeslist" file during our CI system settings.
According to my drop-down the options are: "New", "In Progress", "Resolved", & "Closed", which is why I suggested "Closed" as the most suitable.
On a similar, but different, subject I think it would be useful to have an option like "Dormant". This could be applied to issues like #31 (Opened files damage), which we do not want to close, but there has been no activity for 6 months, and none is expected soon. This would help us manage the list of active topics.
Re: issue-#124 Range-check for SHORT(ENTIER(real))
Posted: Tue Mar 14, 2017 12:05 pm
by Ivan Denisov
Robert wrote:On a similar, but different, subject I think it would be useful to have an option like "Dormant". This could be applied to issues like #31 (Opened files damage), which we do not want to close, but there has been no activity for 6 months, and none is expected soon. This would help us manage the list of active topics.
The "Dormant" status was added. And also I found how to set full permissions for all developers to change statuses.