On 2020-06-09 at 06:51, Nicolas George wrote: > The Wanderer (12020-06-09): > >> (Please stop CCing me on replies - especially to messages which I >> did not actually send - unless you're specifically trying to draw >> my attention to a particular message and think I may not notice it >> without the CC. Not only am I subscribed, I am in fact reading this >> thread on a multiple-times-a-day basis, as my multiple replies to >> it to date may have indicated.) > > Instead of writing this periodically, you could include: > Reply-To: debian-user@lists.debian.org > in your headers just like I did. Having to add that by hand to every single reply I make would be much more of a hassle than taking the time to request this explicitly on the relatively few occasions when people send such CCs with enough frequency for it to be a bother to me. > Properly configured mailing-list software does it by default for > subscribed users, but Debian is an exception. It fixes the issue of > annoying ccs once and for all. I subscribe to probably dozens of mailing lists, and I don't know of any way to configure things to add that header with a proper value automatically on a per-mailing-list basis. Otherwise, I'd probably have done this years ago, unless other considerations (e.g., UI for when I want to do this vs. when I really do want to reply to the sender or to all recipients) took precedence. For myself, I use the "Reply to List" button in (a now-old version of) Thunderbird, and avoid the issue of Reply-To settings entirely unless I actually do want to reply to something other than just the list. >> FWIW, I have tried, at least in part. >> >> For the individual broken-out projects (which may or may not be >> rolled up into the larger "master" project, I can't easily tell), I >> succeeded with one, and failed with another, but suspect that I >> could succeed with the latter with more effort. >> >> For the apparent "master" project, I admit that I didn't bother to >> try, because of the exact "too many prebuilt apparent-dependency >> objects with no apparent way provided to rebuild them" issue; >> unless we can rebuild those objects, not only can we not be sure we >> have the source for them, we can't be sure that building with a >> different version of that object will even work. >> >> Even a successful build from a repository like that would not >> demonstrate that you can actually completely rebuild the project >> from scratch; you'd have to actually track down the source for all >> of those individual prebuilt objects, rebuild each one, and pull >> the result in to the build in a way which will get picked up, and >> that's more effort than I'm willing to put in for the sake of a >> mailing-list discussion like this one. > > Thank you for these clarifications. > >> I don't fault the developers too much for providing a version of >> the project tree with prebuilt dependencies like that; it's a >> useful convenience for those who just want to get it to work and >> for whom farting around with trying to find the right dependencies >> and get them into place would be too much of a hassle. But for (as >> far as I can tell) providing the tree in *only* that form, and not >> providing (as far as I've found) *any* documentation on what these >> prebuilt objects are and why they're needed and how to get them >> separately and build them and so forth, there I do fault them, and >> consider that a ding against proper Free status. > > I state it that way: the knowledge of how to obtain and build these > objects is part of the source code of the project, just as much as a > makefile or configure script. Unfortunately, that bit of source code > only resides in the head of the developers, it is not distributed. > > Consider a minified javascript program with a GPL license header > slapped on top of it: should we consider it Libre Software? Of course > not. The same happens here: out of negligence probably, the authors > keep part of the source code for themselves. It is not Libre > Software. While I wouldn't necessarily take the argument as far as you appear to, I am inclined to agree in principle. That said, while this is an important aspect of the situation, it's technically a tangent from the question of whether people other than the developers can build the program and have the result be usable. If we assume that the developers don't routinely update or replace these prebuilt objects, and don't hack these objects themselves as part of working on the project, then the tree we have is the tree the developers build from - and if we can build a working program from it, then that narrower question is answered "yes". I just don't care to bother with doing that myself at present. Which, to an extent, turns things back to Tomas' point. -- The Wanderer The reasonable man adapts himself to the world; the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man. -- George Bernard Shaw
Attachment:
signature.asc
Description: OpenPGP digital signature