On August 3, 2021 5:58:56 AM UTC, "mayur...@kathe.in" <mayur...@kathe.in> wrote: >On Monday, August 02, 2021 10:42 PM IST, Alan Coopersmith ><alan.coopersm...@oracle.com> wrote: > >> On 8/1/21 11:06 PM, mayur...@kathe.in wrote: >> >> > Where did IPS come from? >> >> Sun engineers who had worked with customers and release engineering >to deal with >> the many problems that people had with SVR4 packaging put together a >project >> proposal to replace it, started prototyping, and developed the >project under the >> OpenSolaris community. > >I have always had the belief that IPS was developed outside of Sun, and >looks like it has been validated. >Please read the comments by a certain "Shawn Walker-Salas" at the link; >https://unix.stackexchange.com/questions/288888/what-is-solaris-ips-how-does-it-work > >It would be interesting the understand what instigated "those" people >_outside_ Sun to write a package management suite and which in turn got >adopted by Sun. > >While I have been advised to avoid contacting the original team >directly to solicit comments about IPS, I'm getting this intolerable >itch to do exactly that. I don't know if I can restrain myself, so I >apologize upfront to all who might get offended by my activities >towards digging up details regarding the IPS. > >Cheers. > > >------------------------------------------ >illumos: illumos-discuss >Permalink: >https://illumos.topicbox.com/groups/discuss/Tf764b19e9f817c85-M36bbea08416f66c50ce27d4e >Delivery options: >https://illumos.topicbox.com/groups/discuss/subscription
I might relate to that "instigate" part: as a sysadmin in my early days, I regularly had some practical goal to fulfill, whether for employer or for fun (learning by tinkering). Back then error/exception handling was not all that great and documentation did not always help - even at Sun's scale of 5000 page tomes for one product of many bundled in an enterprise stack. That was my one large first benefit of open source: even if the final product was not FOSS, it often had sources available - either of an earlier version, or of mozilla etc. siblings, or eventually as part of OpenSolaris. That sufficed to understand many of the original developer's thinking and logic patterns, and often to track down error message text to code that handles the issues. Even as not a regular developer, I could quickly see if and what I misconfigured, or what They got wrong, rarely. And that leads to the second big benefit: ability and right to scratch your own itch. Some such scratchings are published and become more or less popular one-person-show products on their own, others find a way to some upstream and can live on, be supported, evolve and non-regress independently of you as author spending time on that. And often are rewritten beyond recognition by "real" developers who start not from scratch but from a proven viable and useful idea and logic. After all, people who have a real-world problem, who tried and did not find a way around it, are often ones best fit to identify it, not the ivory-tower project managers who just crunch numbers. May be not the best ones to solve correctly though :) At least, my route through some Solaris and Linux internals (and add-on scriptware), some service-wrapping units for Java appservers and Virtualboxes and ZFS pools as SMF units, some Jenkins plugins or NUT, and too many other projects to count that I used and usually had an itch, often had this pattern in common. Then I soothe it and go on just being a user. Until the next day ;) Jim -- Typos courtesy of K-9 Mail on my Android ------------------------------------------ illumos: illumos-discuss Permalink: https://illumos.topicbox.com/groups/discuss/Tf764b19e9f817c85-Ma0ece748d6613b4fbf8e60a7 Delivery options: https://illumos.topicbox.com/groups/discuss/subscription
