also Maxima 5.19.0 doesn't build with newer Ecl --- CORRECTION

Sorry, below was a corruped build (built on top of
an old build): Actually building Maxima 5.19.0
with Ecl 9.8.2 SUCCEEDS (on a 64-bit machine),
and my tests are passed (mostly combinatorial),
however quite a few Maxima tests failed:

Error summary:
Errors found in Maxima/ecl/maxima-5.19.0/tests/rtest14.mac, problems:
(115 126 127 128 129 136 137 138 139 171 172 173 174 177 178 179 180 191 193
 195 200 201 202 203 206 207 208 209 218 219 220 221 222 223 224 229 230 231
 232 235 236 237 238 284 285 287 288 290 293)
Error found in Maxima/ecl/maxima-5.19.0/tests/rtest_gamma.mac, problem:
(543)

For the latter the error is


********************** Problem 543 ***************
Input:
closeto(erf_generalized(%i + 0.35, 1.25), erf(1.25) - erf(%i + 0.35), 1.e-15)


Result:
.3270998717435417

This differed from the expected result:
true



while for that list of errors, they all look like

********************** Problem 293 ***************
Input:
                                    1
test_wronskian(w_ik, lambda([n, z], -), 1.e-10, 5)
                                    z


Result:
[61420.8927537298, [[1, 5], 4.010388984809137, 0.2, 3.810388984809137], 
[[1, 4], 3.880513961318309, 0.25, 3.630513961318309], 
[[1, - 4], - 4209.361885857003 %i - 0.25, - 0.25, 4209.361885857003], 
[[1, - 5], - 39471.93681437964 %i - .2000000000000001, - 0.2, 
39471.93681437964], [[0, 5], 2.849212573397739, 0.2, 2.649212573397739], 
[[0, 4], 2.578278207140269, 0.25, 2.328278207140269], 
[[0, - 4], - 7407.735387770104 %i - 0.25, - 0.25, 7407.735387770104], 
[[0, - 5], - 61420.8927537298 %i - 0.2, - 0.2, 61420.8927537298], 
    1
[[- --, 5], .1458337635072795, 0.2, .05416623649272048], 
    10
    1
[[- --, 4], .1809333719309709, 0.25, .06906662806902908], 
    10
    1
[[- --, 3], .2380759725523081, .3333333333333333, 0.0952573607810252], 
    10
    1
[[- --, 2], 0.347384069528359, 0.5, 0.152615930471641], 
    10
    1
[[- --, 1], .6519492366476867, 1.0, .3480507633523133], 
    10
    1
[[- --, - 1], - .1476904975382651 %i - 0.910750020745156, - 1.0, 
    10
                        1
.1725631532514702], [[- --, - 2], - .08090423934761404 %i - .4227723458755075, 
                        10
                               1

... quite a few of such terms ...


- .07735763390674322 %i - .1901716268095477, - .3333333333333333, 
                        9
.1627251601272902], [[- --, - 4], - .05626137527332276 %i - .1463421824811206, 
                        10
                                9
- 0.25, .1179418733123353], [[- --, - 5], 
                                10
- .04414240778942258 %i - 0.118720709063723, - 0.2, .09249256889367727], 
[[- 1, 5], 2.849212573397739, 0.2, 2.649212573397739], 
[[- 1, 4], 2.578278207140269, 0.25, 2.328278207140269], 
[[- 1, - 4], - 2.578278207140269, - 0.25, 2.328278207140269], 
[[- 1, - 5], - 7.275957614183426e-12 %i - 2.849212573397739, - 0.2, 
2.649212573397739]]

This differed from the expected result:
[]



Oliver

On Wed, Aug 12, 2009 at 09:06:25PM +0100, Oliver Kullmann wrote:
> Hello,
> 
> I hoped somehow with 5.19.0 one
> could use newer versions of Ecl again,
> but also with Ecl version 9.8.2 building
> Maxima fails:
> 
> ;    - Compiling module "info"
> ;      - Loading binary file "binary-ecl/nregex.fas"
> ;;; Loading "/home/csoliver/SAT-Algorithmen/OKplatform/ExternalSources/builds/Maxima/ecl/maxima-5.19.0/src/binary-ecl/nregex.fas"
> 
> Internal or unrecoverable error in:
> Cannot grow stack size.
>   [2: No such file or directory]
> make[2]: *** [binary-ecl/maxima] Aborted
> 
> 
> ??
> 
> Oliver