sidhu1f

For the heavy lifting of the actual linear algebra computations, these tensor computation libraries typically use some variant of BLAS or eigen.

There was a time when GCC had to be installed by compiling from source. Doing so involved compiling the GCC source with the native compiler, and then using the resulting binary to compile the GCC source again.

I remember the install instructions advising "festina lente". The Latin phrase translates to "make haste slowly".

See https://www.planetcom.co.uk

Wouldn't GPL v3 be as effective as AGPL in this context?

My statement (as can be seen from its GP post) uses convenience to mean ease of use, not in the titles available sense that you (and the two following posts) do.

Is BitTorrent more convenient than Netflix?

Top of the line Arria-10 FPGAs have about 1500 floating point MACs [1]. Using such a device, Intel claims ~1 TFLOP sustained for GEMM, the standard matrix multiply operation [2].

[1] https://www.altera.com/content/dam/altera-www/global/en_US/p...

[2] https://www.altera.com/content/dam/altera-www/global/en_US/p...

Even calling them Hindu-Arabic would be preferable to the usual custom of labeling them simply Arabic (for e.g. in LaTeX).

Some of Mcron's innovations[1]:

> The new idea is to read the required command instructions, work out which command needs to be executed next, and then sleep until the inferred time has arrived. On waking the commands are run, and the time of the next command is computed.

> Each user looks after his own files in his own directory. He can use more than one to break up complicated cron specifications. Each user can run his own daemon. This removes the need for suid programs to manipulate the crontabs, and eliminates many security concerns that surround all existing cron programs.

> Vixie cron is implemented in 4500 lines of C code; mcron is 1500 lines of scheme, despite the fact that it offers many more features and much more flexibility, and complete compatibility with Vixie cron.

[1] https://www.gnu.org/software/mcron/design.html

Modern alternatives to BIND that I have had good (though limited) experience with:

- unbound (recursive resolver) https://www.unbound.net

- nsd (authoritative server) https://www.nlnetlabs.nl/projects/nsd

The GNU Artanis[1] web framework seems interesting...

[1] http://web-artanis.com

Not exactly, ME even without AMT is scary[1].

[1]https://libreboot.org/faq/#intelme

A majority (if not most) of the contributions to the Linux kernel come from developers paid to do so by the likes of Intel, Google, IBM etc. [1]. Do you have any reason to believe that said for-profit companies would have contributed to the same degree to a BSD/MIT licensed Linux kernel instead of focusing on a Linux kernel with proprietary extensions under a closed license?

[1] http://www.linuxfoundation.org/news-media/infographics/who-w...

What makes you so sure that, had the GNU stack been BSD/MIT licensed and as popular, that Microsoft etc. would not have come out with products featuring the same software but with proprietary extensions and closed licenses?

Well, as per the coreboot docs [1], the panic level is only 9000+...

[1] https://www.coreboot.org/Binary_situation#Intel

There are some sound technical reasons for choosing scheme: http://wingolog.org/archives/2011/08/30/the-gnu-extension-la...