open-axiom port r2826

[Ralf Hemmecke]

Hello Raoul,

Thanks for your pull request and having an eye on the open-axiom tree.
I had to change your commit a little and added a ChangeLog entry.

https://github.com/hemmecke/fricas/commit/140fc7b3fbe1801469c5dc111fe0932a3c926329

In particular, I did not like this line.
https://github.com/raoulb/fricas/commit/92368df3327c436fd2138c35dc7fd0ead49536b9#L0L91
because it was indented too much.

Furthermore, the special implementation of opposite? appeared under
OrderedAbelianGroup, while I have moved it to AbelianGroup.

I'm not sure, however, whether we really want to take over the lines
https://github.com/hemmecke/fricas/commit/140fc7b3fbe1801469c5dc111fe0932a3c926329#L2L232
into integer.spad.pamphlet. They do not hurt, but they are redundant,
since the implementation is inherited from the category defaults.
Waldek?

Ralf

PS: There is still something wrong since I get the error

>> System error:
Couldn't load
"/home/hemmecke/g/fricas-bisect/b/target/x86_64-unknown-linux/algebra/RNG-.fasl":
file does not exist.

(1) -> Stage 3 object copy
cp: cannot stat `DFLOAT.NRLIB/DFLOAT.fasl': No such file or directory
make[3]: *** [stamp-oboo3] Error 1

during a parallel compilation. Probably a new dependency in an early stage.

[Gabriel Dos Reis]

Ralf Hemmecke <ra...@hemmecke.org> writes:

| Furthermore, the special implementation of opposite? appeared under
| OrderedAbelianGroup, while I have moved it to AbelianGroup.

I thought I put it in AbelianGroup. Thanks for the catch.

[Waldek Hebisch]

Ralf Hemmecke wrote:
>
> Hello Raoul,
>
> Thanks for your pull request and having an eye on the open-axiom tree.
> I had to change your commit a little and added a ChangeLog entry.
>
> https://github.com/hemmecke/fricas/commit/140fc7b3fbe1801469c5dc111fe0932a3c926329
>
> In particular, I did not like this line.
> https://github.com/raoulb/fricas/commit/92368df3327c436fd2138c35dc7fd0ead49536b9#L0L91
> because it was indented too much.
>
> Furthermore, the special implementation of opposite? appeared under
> OrderedAbelianGroup, while I have moved it to AbelianGroup.
>
> I'm not sure, however, whether we really want to take over the lines
> https://github.com/hemmecke/fricas/commit/140fc7b3fbe1801469c5dc111fe0932a3c926329#L2L232
> into integer.spad.pamphlet. They do not hurt, but they are redundant,
> since the implementation is inherited from the category defaults.
> Waldek?

Strictly speaking anything beyond defintions in Monoid and Rng is
an optimization. Given that both functions are unused this may
be premature optimization. OTOH, extra implementations are
tiny and look correct so the cost is quite small. So, when
I looked at the patch I decided that this is not worth
discussing.

One remark: I would like to adapt convention that there
is empty line before '@' -- this helps merging changes and
makes '@' better visible. AFAICS the commit tries to
remove some such empty lines.

> Ralf
>
> PS: There is still something wrong since I get the error
>
> >> System error:
> Couldn't load
> "/home/hemmecke/g/fricas-bisect/b/target/x86_64-unknown-linux/algebra/RNG-.fasl":
> file does not exist.
>
> (1) -> Stage 3 object copy
> cp: cannot stat `DFLOAT.NRLIB/DFLOAT.fasl': No such file or directory
> make[3]: *** [stamp-oboo3] Error 1
>
> during a parallel compilation. Probably a new dependency in an early stage.

AFAICS one has to add RNG to CATDOMS in Makefile.in

--
Waldek Hebisch
heb...@math.uni.wroc.pl

[someone]

Hi,

> In particular, I did not like this line.
> https://github.com/raoulb/fricas/commit/92368df3327c436fd2138c35dc7fd0ead49536b9#L0L91
> because it was indented too much.

Oh, I did not notice that.

> Furthermore, the special implementation of opposite? appeared under
> OrderedAbelianGroup, while I have moved it to AbelianGroup.

Ok.

> PS: There is still something wrong since I get the error
>

> cp: cannot stat `DFLOAT.NRLIB/DFLOAT.fasl': No such file or directory
> make[3]: *** [stamp-oboo3] Error 1

Oh, I got the very same error when compiling the changes, but then I compiled
the 'master' and retried and they were gone. This is of course not the solution
but the reason I did not mention the error anymore.


-- Raoul