| Age | Commit message (Collapse) | Author | Lines |
|---|
|
Until now, we defined `LIBRARY_PATH' to fix the problem of the `ld' linker
of Binutils needing several `*crt*.o' files to run. However, some software
(for example ImageMagick) over-write `LIBRARY_PATH', therefore there is no
other way than to put a link to these necessary files in our local build
directory.
With this commit, we fixed the problem by putting a link to the system's
relevant files in the local library directory. This fixed the problem with
ImageMagick. Later, when we build the GNU C Library in the project, we
should remove this step.
This bug reported by Raul Castellanos Sanchez.
|
|
Until now, when a Fortran compiler didn't exist on the host operating
system, the configure script would crash with a warning. But some projects
may not need Fortran, so this is just an extra/annoying crash!
With this commit, it will still print the warning, but instead of a crash,
it will just sleep for some seconds, then continue. Later, when if a
software needs Fortran, it's building will crash, but atleast the user was
warned.
In the future, we should add a step to check on the necessary software and
see if Fortran is necessary for the project or not. The project
configuration should indeed crash if Fortran is necessary, but we should
tell the user that software XXXX needs Fortran so we can't continue without
a Fortran compiler.
Also, a small sentence ("Project's configuration will continue in XXXX
seconds.") was added after all the warnings that won't cause a crash, so
user's don't think its a crash.
|
|
Until now, the main download script could only check one server for the
given URL. However, ultimately the actual server that a file is downloaded
from is irrelevant for this project: we actually check its
checksum. Especially in the case of software (which are distributed over
many servers), this can usually be very annoying: the servers may not
properly communicate with the running system and even the 10 trials won't
be enough.
With this commit, the download script
`reproduce/analysis/bash/download-multi-try' can take a new optional
argument (a 5th argument). It assumes this argument is a space-separated
list of server(s) to use as backup for the original URL. When downloading
from the original URL fails, it will look into this list and try
downloading the same file from each given server.
|
|
There were no conflicts in this merge.
|
|
Until now, when `./project make' was run after an insuccessful run of
`./project configure', it would just say to run `./project configure'. But
for a first time user, this could be confusing because when the
configuration is done in parallel, the error message can be very high on
the command-line outputs and not seen clearly.
With this commit, the error message is more complete and describes the
problem and what the users should do in which circumstance.
|
|
Until now, Make was just run ordinarily on the two Makefiles of the
software building phase. Therefore when there was a problem with one
software while building in parallel, Make would only complete the running
rules and stop afterwards. But when other rules don't depened on the
crashed rule, its a waste of time to stop the whole thing.
With this commit, both calls to Make in the `configure.sh' script are done
with the `-k' option (or `--keep-going' in GNU Make). With this option, if
a rule crashes, the other rules that don't depend on it will also be
run. Generally, anything that doesn't depend on the crashed rule will be
done. The `-k' option is a POSIX definition in Make, so it is present in
most implemenetations (for the call to `basic.mk').
|
|
Until now the shell scripts in the software building phase were in the
`reproduce/software/bash' directory. But given our recent change to a
POSIX-only start, the `configure.sh' shell script (which is the main
component of this directory) is no longer written with Bash.
With this commit, to fix that problem, that directory's name has been
changed to `reproduce/software/shell'.
|
|
Until now, the project would first ask for the basic directories, then it
would start testing the compiler. But that was problematic because the
build directory can come from a previous setting (with `./project configure
-e'). Also, it could confuse users to first ask for details, then suddently
tell them that you don't have a working C library! We also need to store
the CPATH variable in the `LOCAL.conf' because in some cases, the compiler
won't work without it.
With this commit, the compiler checking has been moved at the start of the
configure script. Instead of putting the test program in the build
directory, we now make a temporary hidden directory in the source directory
and delete that directory as soon as the tests are done.
In the process, I also noticed that the copyright year of the two hidden
files weren't updated and corrected them.
|
|
Rencely the building of GCC was allowed on Debian-based systems that have
their basic C library in architecture-specific directories, like
`/usr/lib/x86_64-linux-gnu'. However, these systems also have their headers
in non-standard locations, for example `/usr/include/x86_64-linux-gnu' and
this caused a crash on a new Ubuntu system.
/usr/include/stdio.h:27:10: fatal error: bits/libc-header-start.h: No
such file or directory
27 | #include <bits/libc-header-start.h>
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
compilation terminated.
The reason it didn't cause problems on other Ubuntus that we tested before
was historic: In the old days, we would ask Ubuntu systems to install
multilib features to have GCC. Because they had installed those features,
this problem didn't show up! But this wasn't mandatory!
With this commit, the `CPATH' environment variable is set (similar to how
`LIBRARY_PATH' was set) and this fixed the problem on a clean Debian
virtual machine.
This bug was reported by Sebastian Luna Valero.
|
|
Until this commit, the small section of `README-hacking.md' in which it
is explained how to do the first configuration of Git was at the
beginning of the section `First custom commit'. However, it is better to
have it just before the item `Your first commit' in that section. With
this commit, this change has been done. Now the reader has the necessary
steps for configuring Git just before it is needed for making the first
commit.
|
|
Until now, the initial project scripts were primarily tested with GNU
Bash. But Bash is not generally available on all systems (it has many
features beyond POSIX). Because of this, effectively we were imposing the
requirement on the user that they must have Bash installed. We recently
started this with setting the shebang of `project' and
`reproduce/software/bash/configure.sh' to `/bin/sh'. After doing so, Raul
and Gaspar reported an error on their systems.
To fix the problem, I installed Dash (a minimalist POSIX-compliant shell)
on my computer and temporarily set the shebangs to `/bin/dash', ran the
project configuration step and fixed all issues that came up. With this
commit, it can go all the way to building GCC on my system's Dash. After
this stage (when `high-level.mk' is called), there is no problem, because
we have our own version of GNU Bash and that installed version is used.
Probably some more issues still remain and will hopefully be found in the
future.
While doing this, I also noticed the following two minor issues:
- The `./project configure' option `--input-dir' was not recognized
because it was mistakenly checking `--inputdir'. It has been corrected.
- The test C programs now use the `<<EOF' method instead of `echo'.
- In `basic.mk', the extra space between `syspath' and `:=' was removed
(it was an ancient relic!).
|
|
Until now, the hashbang of these two shell scripts was set to `/bin/bash',
hence assuming that GNU Bash exists on the host system! But this is an
extra requirement on the host operating system and these two scripts should
be written such that they operate on a POSIX shell (the generic `/bin/sh'
which can point to any shell program).
With this commit this has been implemented! We may confront some errors as
the system is run on other systems, but we should fix such errors and work
hard to make these two scripts as POSIX-compatible as possible (runnable on
any shell, so as not to force users to install Bash before running the
project).
This completes Task #15525.
|
|
Until now, the main commands to run the project were these: `./project
configure' (to build the software), `./project prepare' (to possibly
arrange input datasets and build special configuration Makefiles) and
finally `./project make' to run the project.
The main logic behind the "prepare" phase `top-prepare.mk' is to build
configuration files that can be fed into the "make" step and optimize its
operation. For example when the total number of necessary inputs for the
majority of the analysis is not as large as the total number of
inputs. With "prepare" (when necessary), you go through the raw inputs,
select the ones that are necessary for the rest of the project. The output
of `top-prepare.mk' is a configuration file (a Make variable) that keeps
the IDs (numbers, names, etc). That configuration file would then be used
in the `top-make.mk' to identify the lower level targets and allow optimal
project organization and management.
But the last two are both part of the analysis, and while they indeed need
different calls to Make to be executed, many projects don't actually need a
preparation phase: ultimately, its an implementation choice by the project
developers and doesn't concern the project users (or the developers when
they are running it).
To avoid confusing the users, or simply annoying them when a projet doesn't
need it, with this commit, the top-level `top-prepare.mk' and `top-make.mk'
Makefiles are called with the single `./project make' command and
`./project prepare' has been dropped. I noticed this while writing the
paper on this system.
|
|
Since I (Raul) did some changes (and I hope to do more :-)) in the
`README-hacking.md', I am adding my information at the beginning of this
file.
|
|
Until this commit, we were asuming that Git was already properly
configured. However, in order to be as complete as possible, it would
be good if the basic commands to configure Git were in the
`README-hacking.md'. With this commit, a small paragraph has been added
in order to have the basic Git configuration commands (i.e. to configure
the name, email, and favorite text editor).
|
|
Until now, the explanation for a missing static C library didn't actually
guide the users to look above and see the error message! So with this
commit, I edited it a little to be more clear (and mention to look
above). Also, I noticed that on Amazon AWS systems, the static C library is
installed as a separate package, so to help the users, I added the
necessary command and some better explanation.
|
|
Until now Perl was built after Coreutils, but I recently noticed that
Coreutils actually uses Perl while creating its manpages. So it is now
built before Coreutils.
Also, while testing on an Amazon AWS EC2 server, we noticed that Coreutils
can't build its man page for `md5sum'. The problem was found to be due to
the fact that until now, we weren't actually setting LD_LIBRARY_PATH to our
installed library path in `basic.mk'. Therefore, it would crash because the
server had an older version of OpenSSL than the one that the template's
Coreutils was built with.
In the meantime (while addressing the issues above, because we only had one
thread on the AWS server) I also noticed a few programs that were using a
summarize compilation command (that just prints `CC xxx.c' instead of the
whole command) so I fixed them by adding `V=1'.
This bug was found by Idafen Santana Pérez.
|
|
Until now, the configuration Makefiles (in
`reproduce/software/config/installation' and `reproduce/analysis/config')
had a `.mk' suffix, similar to the workhorse Makefiles. Although they are
indeed Makefiles, but given their nature (to only keep configuration
parameters), it is confusing (especially to early users) for them to also
have a `.mk' (similar to the analysis or software building Makefiles).
To address this issue, with this commit, all the configuration Makefiles
(in those directories) are now given a `.conf' suffix. This is also assumed
for all the files that are loaded.
The configuration (software building) and running of the template have been
checked with this change from scratch, but please report any error that may
not have been noticed.
THIS IS AN IMPORTANT CHANGE AND WILL CAUSE CRASHES OR UNEXPECTED BEHAVIORS
FOR PROJECTS THAT HAVE BRANCHED FROM THIS TEMPLATE. PLEASE CORRECT THE
SUFFIX OF ALL YOUR PROJECT'S CONFIGURATION MAKEFILES (IN THE DIRECTORIES
ABOVE), OTHERWISE THEY AREN'T AUTOMATICALLY LOADED ANYMORE.
|
|
GNU Make 4.3 was just announced, so I have updated it here is well. This
was important because until now the installable version was in alpha-mode
(4.2.90), now its a stable version.
|
|
Until now, it was necessry to run a long `while true' loop to see what is
currently being built at configure time. So with this commit, a new
`--checkconfig' option has been added to `./project' that can be called to
run that loop and make it easier to check.
|
|
In the previous commmit, I had forgot to add a `\' after the newly added
`sys_library_path' variable to the `high-level.mk' call.
|
|
Until now, GCC wouldn't build properly on Debian-based operating systems
because `ld' needed to link with several necessary C library features like
`crti.o' and `crtn.o' (this is an `ld' issue, not GCC). The solution is to
add the directory containing them to `LIBRARY_PATH'. In the previous
commit, I actually searched for these files, but while testing on another
system, I noticed that it can be problematic (other architectures may
exist).
With this commit, we are actually finding the build architecture of the
running GCC (which is the same as the `ld') and using that to fix a fixed
directory to `LIBRARY_PATH'.
|
|
Until now, to see if a working static C library and `sys/cdefs.h' exist, we
were checking absolute locations like `/usr/include/sys/cdefs.h' or
`/usr/lib/libc.a' and `/usr/lib64/libc.a'. But this is not robust because
on different systems, they can be in different locations.
With this commit, we actually use `find' to find the location of `libc.a'
and use that to add elements to CPPFLAGS and LDFLAGS. This should fix the
problem on systems that have them on non-standard locations.
|
|
Until now, when find the versions of the TeXLive packages, we would assume
that `cat-date' is always present (because some packages don't have a
version!). However, apparently an update has been made in the TeXLive
Manager (`tlmgr') and `cat-date' is no longer present! As a result, none of
the TeXLive packages were being printed.
With this commit, it now assumes that `revision' is always present for
every package, but it also attempts to read `cat-date' (for backwards
compatability). When `cat-version' isn't present, it will try printing
`revision' and if that is also not present, it will print the date.
|
|
After a new rebuild of the project, I noticed that we now need to also
build the `mweights' package.
|
|
The checklist descriptions were slightly edited to be more clear. Also,
while following them, I noticed that while removing the "delete-me" parts
on `verify.mk', would cause an error: the `if [ $$m == delete-me ];'
statement we were saying to delete cause an error because `elif' was the
first statement Bash would see. So with this commit, the `download'
conditional (which isn't instructed to be deleted) was set to be the top
(with an `if') and the `delete-me' conditional now has an `elif'.
|
|
Until now, the small one-line script that lists programs was introduced in
the checklist after running `./project configure'. But people would mostly
miss it because they would wait until the configuration is complete.
With this commit, that point has been put above the `./project configure'
step. Readers are instructed to open a new terminal and run that script,
then go to the next step so they see the directories get filled
actively. It will also help them understand what is going on.
|
|
Until now the actual journal webpage was used for Raul's paper. However,
the journal webpage needs authorized access for people to read it,
therefore its will be inaccessible for many people. A better and more well
known place for the paper (atleast in astronomy) is the ADS link. In the
ADS link, if someone has access to the journal, they will get the journal's
version and if not, they will get the arXiv version. It also has a common
BibTeX export tool for all journals. We had also done this for the other
papers in that list.
With this commit, I thus changed the URL for the paper, and also removed
the "issue" number (4 in this case), since that is mostly irrelevant, only
the volume and page numbers are usually used for the other papers too.
|
|
The "SDSS extended PSFs" paper was already included as an example of
papers wich uses this template. However, the reference was the arXiv
one. With this commit, since the paper has been finally published, it
has been added the final reference to the journal.
|
|
Until this commit, the checking of X11 installation done to ensure that
it is already available in the host system was crashing in macOS
systems. The reason is that the place of the X11 libraries use to be
`/opt/X11/lib' in macOS systems. With this commit, this issue has been
fixed by adding this directory to the LDFLAGS.
|
|
Newer versions of Astropy package has been released. With this commit,
it has been updated. It has been increased from v3.2.1 to v4.0
|
|
In markdown, to specify that part of the text should be shown as code we
need to put it between two `. But in `reproduce/software/make/README.md',
the code parts started with ` and finished in ' (like the file name above).
With this commit, this has been corrected and it will display properly on
web browsers.
|
|
Until now, if the file to be verified didn't exist, a different checksum
would be generated, and it would stop, but it wasn't immediately clear if
the differing checksum is because the file doesn't exist at all!
With this commit, before calculating the checksum, we first make sure if
the file exists. If it doesn't exist an explicit error is printed and thus
will help the project editor to find the cause of the problem.
|
|
In the previous commit, I had forgot to update a small part in the
checklist (when modifying `top-make.mk') which is now corrected.
I also added a few sentences in the description of how to customize the
verification to make it easier to understand.
|
|
Until now, the only verification that the template provided was the
published PDF. Users had to manually compare the published and generated
PDFs (numbers, plots, tables) and see if they obtained the same
result. However, this type of manual verification is not good and is prone
to frustration and missing important differences.
With this commit, a new Makefile has been added in the analysis steps:
`verify.mk'. It provides facilities to easily verify the results that go
into the paper. For example tables that go into making the paper's plots,
or the LaTeX macros that blend into the text. See the updated parts in
`README-hacking.md` for a more complete explanation.
This completes task #15497.
|
|
In the previous commit, we added the files to ignore from the template
branch, but only the files that had been deleted. With this commit,
`paper.tex' is also added to the files that must be ignored from the
template branch (the file remains in the project, but in the template
branch, its contents are just dummy place-holders).
|
|
During the checklist we guide the user to delete the dummy `delete-me*'
files from their custom branch. Later, if the dummy files are updated in
the template's master branch, if the user merges with the template branch,
these files will be written back into their project! This is very annoying!
With this commit, a step was added in the `README-hacking.md' checklist,
just after deleting the dummy files to avoid this problem using the
`.gitattributes' file, telling Git to keep the changes as implemented in
the merging branch (`ours').
|
|
Now that its 2020, its necessary to include this year in the copyright
statements.
|
|
Until this commit, the number `2' was missing in the checksum variable
name of that library. It was `libxml-checksum' but it should be
`libxml2-checksum'. With this commit, this issue has been fixed.
|
|
An extra backslash in the prerequisites of Jeepney Python package was
causing a crash in the installation of that software. With this commit,
this problem has been fixed by removing the backslash.
|
|
OpenMP takes a LONG TIME to build, so to keep things reproducible we are
explicitly disabling OpenMP, if a user needs OpenMP, its trivial to just
add it as a prerequisite of R. The problem is that in some scenarios (based
on other dependencies and when they were built in the build directory),
OpenMP may be present when R is being installed and in other it may not. We
don't want the result to be different between the two builds.
|
|
With this commit, we now have the core R interpretter within the
template. We should later include instructions to install R packages
(possibly in a separate top-level Makefile like Python).
|
|
Recent builds of the template need these three packages to build the PDF.
This was reported by Hamed Altafi.
|
|
Until now we weren't including this package, but Alberto Madrigal recently
reported that his build of the template failed because it needed it.
With this commit, it is now being installed with the template.
|
|
Until now, while building Perl, the `-Dlddlflags="-shared $$LDFLAGS"'
option was being used for all systems. However, this caused `symbol(s) not
found for architecture x86_64' problems on macOS systems.
With this commit, this configure option is only passed for GNU/Linux
systems. With it, Perl can be built nicely on both GNU/Linux and macOS.
|
|
Newer versions of these packages have recently come out with major
improvements, so they have been updated in the template.
|
|
Until now, Ghostscript was using some host system's X11 libraries during
its build (and later at run-time). We should ideally install all these
necessary libraries within the template (Task #15481). But right now we are
too busy.
As a temporary work-around we try building a small dummy program that links
with some of those libraries before attempting to built Ghostscript. If it
fails, then a notice is printed with the cause and explaining a temporary
solution is suggested: how to install those libraries on the system when
you have root access.
|
|
Until now, when an input dataset already exists in `INDIR', the template
would just make a symbolic link to it in the build directory. However, in
many cases, the files in INDIR will actually be links to other locations on
the filesystem and some programs have problems following too many links.
With this commit, the template is now using the `readlink' program (part of
GNU Coreutils) to follow a possible link and point the link in the build
directory directly to an actual non-link file.
|
|
Until now, after removing all environment variables, we were just giving
Make the top Makefile to execute. By default, for every target, Make will
try many built-in rules (which is good when compiling programs, but
redundant in other cases). All these checkings also populate the debugging
output of Make (with `-d'). So its easier and slightly faster to just tell
Make to ignore builtin rules and variables.
With this commit, to address this issue, the `project' script runs
`.local/bin/make' with `--no-builtin-rules' and `--no-builtin-variables'.
|
|
Until now when building Matplotlib on macOS systems, we were using the
default C compiler. However, while Yahya Sefidbakht (previously) and
Mahdieh Nabavi (now) were trying to build the template, on their macOS
using the GNU Compiler Collection (GCC), we found that Matplotlib needs
special macOS headers that GCC doesn't recognize.
With this commit, when Matplotlib is being built on macOS systems, it uses
`clang' and this fixed the problem (so far checked on Mahdieh's machine).
|
