Skip to content

Piero Bosio Social Web Site Personale Logo Fediverso

Social Forum federato con il resto del mondo. Non contano le istanze, contano le persone

Last week I had a chat with a colleague who is highly specialized in Microsoft solutions.

  • Last week I had a chat with a colleague who is highly specialized in Microsoft solutions. Young but not too young, smart, not very up to date simply because he has little time for anything else. His specialization depends entirely on where he works, not on personal interest. Lately he seemed a bit disillusioned with some choices made by "other operating systems", and he was starting to consider moving his personal projects toward Microsoft as well, since he already had the experience. Still, he said it with boredom. With the attitude of someone who is tired of wasting time.

    He had heard of the BSDs but had never tried installing them. He was convinced that there were no decent hypervisors outside the Linux world and that KVM belonged to Linux alone. I had the terrible idea of showing him the BSDs, how great bhyve is, and how nvmm on NetBSD uses qemu underneath, making it almost a replacement for KVM in many setups. He lit up with the look of someone waking up from a long sleep. I also had the terrible idea of showing him illumos and its distributions. He had no clue it existed and thought old, great Solaris had been dead for years thanks to Oracle.

    He called me a little while ago. He was furious. He spent the whole weekend doing tests and now he has no idea what to use among FreeBSD with bhyve, NetBSD with nvmm, and illumos with bhyve or kvm. He is slowly starting to explore jails and illumos zones. He was annoyed (in a positive way) because now he does not know what to pick since everything feels so different from what he was used to, and he found advantages in each option.

    I am obviously happy about it, but I also wonder: instead of reinventing the wheel every time, would it not sometimes be better to simply broaden our horizons?

  • Last week I had a chat with a colleague who is highly specialized in Microsoft solutions. Young but not too young, smart, not very up to date simply because he has little time for anything else. His specialization depends entirely on where he works, not on personal interest. Lately he seemed a bit disillusioned with some choices made by "other operating systems", and he was starting to consider moving his personal projects toward Microsoft as well, since he already had the experience. Still, he said it with boredom. With the attitude of someone who is tired of wasting time.

    He had heard of the BSDs but had never tried installing them. He was convinced that there were no decent hypervisors outside the Linux world and that KVM belonged to Linux alone. I had the terrible idea of showing him the BSDs, how great bhyve is, and how nvmm on NetBSD uses qemu underneath, making it almost a replacement for KVM in many setups. He lit up with the look of someone waking up from a long sleep. I also had the terrible idea of showing him illumos and its distributions. He had no clue it existed and thought old, great Solaris had been dead for years thanks to Oracle.

    He called me a little while ago. He was furious. He spent the whole weekend doing tests and now he has no idea what to use among FreeBSD with bhyve, NetBSD with nvmm, and illumos with bhyve or kvm. He is slowly starting to explore jails and illumos zones. He was annoyed (in a positive way) because now he does not know what to pick since everything feels so different from what he was used to, and he found advantages in each option.

    I am obviously happy about it, but I also wonder: instead of reinventing the wheel every time, would it not sometimes be better to simply broaden our horizons?

    @stefano what did you mean when you wrote that he was disillusioned with other operating systems? Did he mean Linux and what choices did he mean?

  • @stefano what did you mean when you wrote that he was disillusioned with other operating systems? Did he mean Linux and what choices did he mean?

    @christopher Linux and MacOS - for Linux, he says "it's moving to become a new Windows" and MacOS, according to him, is moving to a "mobile first OS".
    Not that I totally agree, but it's his opinion and I respect it.

  • @christopher Linux and MacOS - for Linux, he says "it's moving to become a new Windows" and MacOS, according to him, is moving to a "mobile first OS".
    Not that I totally agree, but it's his opinion and I respect it.

    @stefano I wonder what does it mean that Linux becomes a Windows...

  • @stefano I wonder what does it mean that Linux becomes a Windows...

    @christopher According to him, the initial concept of simplicity (Unix style) is being lost, replaced by complex solutions that do not solve old problems or just create some new ones. On this point, at least in part, I agree.

  • @christopher According to him, the initial concept of simplicity (Unix style) is being lost, replaced by complex solutions that do not solve old problems or just create some new ones. On this point, at least in part, I agree.

    @stefano @christopher I am not sure if I'd say is becoming like . I do recall similar statements made on the Debian-User mailing list on a previous release when xorg introduced autoconfiguration. A lot of people were pissed that it was making choices for you instead of manually configuring the xorg.conf file.

    Honestly, that was a good thing. Painful doesn't begin to describe it but users were unaware they could still hand-configure the file.

    There has been, however, more stuff added to Linux over the last several years. Call it bloat, call it whatever you want. OSes change. But it has been gradually moving away from simplicity.

    I miss the simplicity.

    However, to reply to your original post, coming from COTS solutions, sometimes the vast amount of choice can be overwhelming. For instance, when it comes to it used to just be jails. Now, it's thin, thick, classic, networking. I understand they have their places but it would be helpful to provide more detailed explanations, tutorials, or best practices for each. The FreeBSD Handbook is good but just scratches the surface but often leaves more questions. It would help with learning and in part...marketing.

    On a side note: The FreeBSD Handbook is a great resource but there are opportunities to improve it, like tailoring it to new users (better empathy), best practices, architectural examples, and links to additional resources and info.


Gli ultimi otto messaggi ricevuti dalla Federazione
Post suggeriti