View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0001592 | GNUnet | build process | public | 2010-07-28 00:26 | 2010-10-21 15:47 |
| Reporter | c89_nerd | Assigned To | Christian Grothoff | ||
| Priority | normal | Severity | minor | Reproducibility | always |
| Status | closed | Resolution | fixed | ||
| Summary | 0001592: Improper use of non-portable neninet/ip.h include file. | ||||
| Description | $ make install 2>&1|less [...] In file included from ../../src/include/platform.h:87, from bandwidth.c:26: /usr/include/netinet/ip.h:153: error: syntax error before "n_time" [... RRAAAAAAAAGH!!! ...] $ vi src/include/platform.h #include <netinet/in.h> +#include <netinet/in_systm.h> #include <netinet/ip.h> /*superset of previous */ $ make install 2>&1|less [... MUCH better ...] you should realize that netinet/ip.h is !!!NOT!!! portable. in a 0.8.1b version I had minor problems with your non-standard use of error_t... Please adhere to SUSV3: The Open Group Base Specifications Issue 6. What "portable source code" standard(s) do you adhere to? | ||||
| Tags | No tags attached. | ||||
|
|
Let me invert your question: what OS are *you* using!? |
|
|
OpenBSD, which admittedly is not complaint with the standards I mentioned (shm_open doesn't seem to exist on my system, but that is a simple enough function to implement manually for libextractor). I am still curious what value/relevance you place on portable source code standards. |
|
|
> you should realize that netinet/ip.h is !!!NOT!!! portable. yeah I guess that could be interpreted as screaming... it was supposed to be an expression of frustration and not hate. fyi |
|
|
Ok, thanks for your info. OpenBSD is on our list of platforms we'd like to support, but we don't have a buildbot for it (=> trouble ahead). For OSes that we do have a buildbot for, you can find our status report here: http://ng.gnunet.org:8010/waterfall (note that some tests will simply fail because we're currently working on that code and thus it may not always be a portability issue). If you're willing and able to setup an OpenBSD buildbot for us, you can find information about that in the FAQ. That would make it much easier for us to find & fix these issues. The priority order for platforms to support (in terms of how much time *we* invest) is roughly GNU/Linux, W32, OS X, FreeBSD, OpenBSD, NetBSD; for processors it is right now roughly AMD64, x86, ARM, PPC, Sparc, Mips, Cell. The lower you are on the list, the more you're likely to encounter frustrations. Note that neither order is really a judgement of the quality of the platform and more a combination of what we use (and have access to) and what the rest of the world uses. As far as standards are concerned, I'd say the answer is that we want to get it to work on most platforms, sometimes standards can help there, but generally that's simply not enough (and using the standard itself may work in some cases but not others). We do take patches... |
|
|
Should be fixed in SVN 12383. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2010-07-28 00:26 | c89_nerd | New Issue | |
| 2010-07-28 01:00 | Christian Grothoff | Note Added: 0004081 | |
| 2010-07-28 01:00 | Christian Grothoff | Status | new => feedback |
| 2010-07-28 03:20 | c89_nerd | Note Added: 0004083 | |
| 2010-07-28 03:26 | c89_nerd | Note Added: 0004084 | |
| 2010-07-28 21:57 | Christian Grothoff | Note Added: 0004087 | |
| 2010-07-28 22:01 | Christian Grothoff | Note Added: 0004088 | |
| 2010-07-28 22:01 | Christian Grothoff | Status | feedback => resolved |
| 2010-07-28 22:01 | Christian Grothoff | Resolution | open => fixed |
| 2010-07-28 22:01 | Christian Grothoff | Assigned To | => Christian Grothoff |
| 2010-10-21 15:47 | Christian Grothoff | Status | resolved => closed |
