ANN: Pure 0.48

Albert Graef

unread,

Oct 18, 2011, 2:43:36 PM10/18/11

to pure...@googlegroups.com

At long last, Pure 0.48 has been released.

Download:
http://pure-lang.googlecode.com/files/pure-0.48.tar.gz

As usual, the complete online documentation is in a separate tarball:
http://pure-lang.googlecode.com/files/pure-docs-0.48.tar.gz

NOTE: A bunch of new addon module versions go along with this, see below.

Here's a brief rundown of the most important changes:

- LLVM 3.0 compatibility and experimental DragonEgg support
(http://dragonegg.llvm.org/).

- Minor changes in the Faust interface.

- Important bugfixes in the lexer and in the __locals__ builtin.

- Some new pragmas (--defined/--nodefined, --warn/--nowarn), builtins
(__func__, __namespace__) and library functions (lasterrpos; new and
much improved printf/scanf implementations), please see the ChangeLog
and the documentation for details.

- The regex and pointer arithmetic operations are now in separate regex
and pointers modules.

- Cosmetic changes and improvements in Emacs Pure mode (pure-mode.el).

Note that some of the added functionality in the runtime and the
standard library (specifically, support for printing GMP bigints and
MPFR multiprecision floats using the %Zd and %Rg conversions) now
require libmpfr to be installed to compile Pure (http://www.mpfr.org/).
The MPFR floats themselves aren't in the core package yet, however, they
are provided by a separate pure-mpfr package (TBA).

Addons
------

The following addon packages have been updated to be compatible with
this Pure release: faust2pd-2.5, pd-pure-0.14, pure-faust-0.5,
pure-fastcgi-0.4, pure-gen-0.14, pure-liblo-0.8, pure-sockets-0.6. You
*must* upgrade these to make them work with Pure 0.48.

In addition, there are also two bugfix releases: pure-midi-0.4 (compiler
warnings), pure-odbc (64 bit fixes). Older versions of these should
continue to work with Pure 0.48, but updating is still recommended.

Please check the Addons page on the wiki for a description of these
packages. Also note that in any case you'll have to recompile *all* your
addons; due to some ABI changes Pure 0.48 comes with a new version
(8.0.0) of the Pure runtime library.

Windows Packages
----------------

As usual, we also provide binary Windows msi packages to go along with
these. This is a *complete* update for all addon packages currently
available on Windows, as all the addon modules had to be recompiled.

Note that the pure-0.48.msi package already contains a bunch of "core"
addon packages (this now also includes pure-mpfr and pure-stlvec, to be
announced separately). Other addon modules (mostly graphics and
multimedia related stuff) are available as separate msi packages.

Downloads:
http://pure-lang.googlecode.com/files/pure-0.48.msi
http://pure-lang.googlecode.com/files/pd-pure-0.14.msi
http://pure-lang.googlecode.com/files/pure-audio-0.4.1.msi
http://pure-lang.googlecode.com/files/pure-gl-0.8.1.msi
http://pure-lang.googlecode.com/files/pure-gtk-0.10.msi
http://pure-lang.googlecode.com/files/pure-liblo-0.8.msi
http://pure-lang.googlecode.com/files/pure-midi-0.4.msi
http://pure-lang.googlecode.com/files/pure-tk-0.3.msi

Enjoy! :)
Albert

--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.G...@t-online.de, a...@muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikinformatik.uni-mainz.de/ag

Sidharth Kshatriya

unread,

Oct 19, 2011, 2:03:50 AM10/19/11

to pure...@googlegroups.com

Awesome! :-)

--
You received this message because you are subscribed to the Google Groups "pure-lang" group.
To post to this group, send email to pure...@googlegroups.com.
To unsubscribe from this group, send email to pure-lang+unsubscribe@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/pure-lang?hl=en.

--
Sidharth Kshatriya
www.sidk.info

manjula gandhi

unread,

Oct 19, 2011, 3:01:04 AM10/19/11

to pure...@googlegroups.com

Hello
I tried installing pure-0.48 in Windows system, I get the following
error message when I run pure "This application has failed to start
because libmpfr-1.dll was not found. Re-installing the application may
fix this problem."

I reinstalled it, again i get the same error.
Need your help. Have I missed something.
Regards
Manjula

On 10/19/11, Sidharth Kshatriya <sid.ks...@gmail.com> wrote:
> Awesome! :-)
>
> On Wed, Oct 19, 2011 at 12:13 AM, Albert Graef <Dr.G...@t-online.de> wrote:
>
>> At long last, Pure 0.48 has been released.
>>
>> Download:

>> http://pure-lang.googlecode.**com/files/pure-0.48.tar.gz<http://pure-lang.googlecode.com/files/pure-0.48.tar.gz>

>>
>> As usual, the complete online documentation is in a separate tarball:

>> http://pure-lang.googlecode.**com/files/pure-docs-0.48.tar.**gz<http://pure-lang.googlecode.com/files/pure-docs-0.48.tar.gz>

>>
>> Please check the Addons page on the wiki for a description of these
>> packages. Also note that in any case you'll have to recompile *all* your
>> addons; due to some ABI changes Pure 0.48 comes with a new version (8.0.0)
>> of the Pure runtime library.
>>
>>
>> Windows Packages
>> ----------------
>>
>> As usual, we also provide binary Windows msi packages to go along with
>> these. This is a *complete* update for all addon packages currently
>> available on Windows, as all the addon modules had to be recompiled.
>>
>> Note that the pure-0.48.msi package already contains a bunch of "core"
>> addon packages (this now also includes pure-mpfr and pure-stlvec, to be
>> announced separately). Other addon modules (mostly graphics and multimedia
>> related stuff) are available as separate msi packages.
>>
>> Downloads:

>>
>> Enjoy! :)
>> Albert
>>
>> --
>> Dr. Albert Gr"af
>> Dept. of Music-Informatics, University of Mainz, Germany
>> Email: Dr.G...@t-online.de,

>> a...@muwiinfa.geschichte.uni-**mainz.de<a...@muwiinfa.geschichte.uni-mainz.de>
>> WWW:
>> http://www.musikinformatik.**uni-mainz.de/ag<http://www.musikinformatik.uni-mainz.de/ag>

>>
>> --
>> You received this message because you are subscribed to the Google Groups
>> "pure-lang" group.
>> To post to this group, send email to pure...@googlegroups.com.

>> To unsubscribe from this group, send email to pure-lang+unsubscribe@**
>> googlegroups.com <pure-lang%2Bunsu...@googlegroups.com>.
>> For more options, visit this group at http://groups.google.com/**
>> group/pure-lang?hl=en <http://groups.google.com/group/pure-lang?hl=en>.
>>
>>
>
>
> --
> Sidharth Kshatriya
> www.sidk.info

>
> --
> You received this message because you are subscribed to the Google Groups
> "pure-lang" group.
> To post to this group, send email to pure...@googlegroups.com.
> To unsubscribe from this group, send email to

> pure-lang+...@googlegroups.com.

Albert Graef

unread,

Oct 19, 2011, 4:38:10 AM10/19/11

to pure...@googlegroups.com

On 10/19/2011 09:01 AM, manjula gandhi wrote:
> I tried installing pure-0.48 in Windows system, I get the following
> error message when I run pure "This application has failed to start
> because libmpfr-1.dll was not found. Re-installing the application may
> fix this problem."

Drat, I forgot to include that dll in the package. Will upload a new msi
asap.

Albert

--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany

Email: Dr.G...@t-online.de, a...@muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikinformatik.uni-mainz.de/ag

Albert Graef

unread,

Oct 19, 2011, 4:50:20 AM10/19/11

to pure...@googlegroups.com

On 10/19/2011 10:38 AM, Albert Graef wrote:
> On 10/19/2011 09:01 AM, manjula gandhi wrote:
>> I tried installing pure-0.48 in Windows system, I get the following
>> error message when I run pure "This application has failed to start
>> because libmpfr-1.dll was not found. Re-installing the application may
>> fix this problem."
>
> Drat, I forgot to include that dll in the package. Will upload a new msi
> asap.

Here's a new, hopefully fixed version of the msi package, please try and
let me know if it works now.

http://pure-lang.googlecode.com/files/pure-0.48.msi

Thanks,

manjula gandhi

unread,

Oct 19, 2011, 6:13:31 AM10/19/11

to pure...@googlegroups.com

Still I get the the same error.

Albert Graef

unread,

Oct 19, 2011, 6:59:28 AM10/19/11

to pure...@googlegroups.com

On 10/19/2011 12:13 PM, manjula gandhi wrote:
> Still I get the the same error.

Can you please check whether your "Program Files\Pure" directory
contains the libmpfr-1.dll file?

Jiri Spitz

unread,

Oct 19, 2011, 7:04:47 AM10/19/11

to pure...@googlegroups.com

Albert Graef wrote:
> On 10/19/2011 12:13 PM, manjula gandhi wrote:
>> Still I get the the same error.
>
> Can you please check whether your "Program Files\Pure" directory
> contains the libmpfr-1.dll file?
>

As I remember, the used installer requires either to uninstall Pure
before upgrading to a new version or to be run two times.

Jiri

Albert Graef

unread,

Oct 19, 2011, 7:34:18 AM10/19/11

to pure...@googlegroups.com

On 10/19/2011 01:04 PM, Jiri Spitz wrote:
> As I remember, the used installer requires either to uninstall Pure
> before upgrading to a new version or to be run two times.

Or just make sure that you uninstall any old Pure package before
installing the new one.

I thought that the AdvancedInstaller guys had fixed that bug in their
free version by now (I did increment the internal version number and
generate a new key for the updated package). But yes, that's what I
suspect, too (either that or a stale download).

I double-checked the msi download from GoogleCode and it does contain
the libmpfr-1.dll file. I can also see that it's being installed in my
virtual Windows box.

Albert

P.S.: Jiri, nice to hear from you again! Have you taken the new version
for a test drive yet?

Eddie Rucker

unread,

Oct 19, 2011, 8:14:08 AM10/19/11

to <pure-lang@googlegroups.com>

I uninstalled Pure, then installed Pure 0.48 and it works on the Win 7 box in my office.

Eddie Sent from my iPhone

Jiri Spitz

unread,

Oct 19, 2011, 8:18:25 AM10/19/11

to pure...@googlegroups.com

Albert Graef wrote:
>
> P.S.: Jiri, nice to hear from you again! Have you taken the new version
> for a test drive yet?
>

I was doing a big renovation of my flat and I was without access to my
home computer for several months. I was online in my office but I could
not hack on Pure during my working hours, of course.

The version 0.48 works fine for me but I use my own compilation and not
the MSI package.

The only thing I did in the meantime is that I found Pure can be
compiled with i686-w64-mingw32. It only requires inclusion/exclusion of
some header files and declarations. When time permits I'll try to
compile for 64 bit Windows as well.

Jiri

Albert Graef

unread,

Oct 19, 2011, 9:22:42 AM10/19/11

to pure...@googlegroups.com

On 10/19/2011 02:18 PM, Jiri Spitz wrote:
> I was doing a big renovation of my flat and I was without access to my
> home computer for several months. I was online in my office but I could
> not hack on Pure during my working hours, of course.

Ah ok, I was a bit worried that you might be ill or lost interest in
programming.

> The version 0.48 works fine for me but I use my own compilation and not
> the MSI package.

Yes, I know. But did you try pure-odbc lately? In the latest version I
did some 64 bit fixes which were needed for Ubuntu compatibility. It
works fine for me on 32 bit Windows (using the sqlite3-odbc driver), but
I'd appreciate your input on that. (Or, for that matter, anyone else's
who uses pure-odbc on Windows.)

Concerning 64 bit Ubuntu, unfortunately the sqlite3-odbc driver seems to
be broken there, now that they changed SQLLEN from int to long in the
unixodbc package. I posted about that a while ago in the "pure-odbc 64
bit issues/fixes" thread
[http://groups.google.com/group/pure-lang/browse_thread/thread/498ef6e30c508c4a].
If anyone knows more about this, please help! Of course there's Peter's
pure-sql3 module, but having sqlite3 work transparently through the ODBC
interface is still useful as well.

> The only thing I did in the meantime is that I found Pure can be
> compiled with i686-w64-mingw32. It only requires inclusion/exclusion of
> some header files and declarations. When time permits I'll try to
> compile for 64 bit Windows as well.

Very nice! Patches are gladly accepted. :)

Cheers,
Albert

Jiri Spitz

unread,

Oct 19, 2011, 10:41:59 AM10/19/11

to pure...@googlegroups.com

Albert Graef wrote:
>
> Yes, I know. But did you try pure-odbc lately? In the latest version I
> did some 64 bit fixes which were needed for Ubuntu compatibility. It
> works fine for me on 32 bit Windows (using the sqlite3-odbc driver), but
> I'd appreciate your input on that. (Or, for that matter, anyone else's
> who uses pure-odbc on Windows.)
>

Pure-odbc works for me on 32 bit Windows with MSAccess (real
stresstest), MySQL 5.1 and PostgreSQL 8.1.

Jiri

Albert Graef

unread,

Oct 19, 2011, 7:17:37 PM10/19/11

to pure...@googlegroups.com

On 10/19/2011 04:41 PM, Jiri Spitz wrote:
> Pure-odbc works for me on 32 bit Windows with MSAccess (real
> stresstest), MySQL 5.1 and PostgreSQL 8.1.

Thanks, that's good to know.

manjula gandhi

unread,

Oct 20, 2011, 12:02:21 AM10/20/11

to pure...@googlegroups.com

It works fine for me now. Thanks a lot.
Manjula

Ryan Schmidt

unread,

Oct 20, 2011, 2:44:49 AM10/20/11

to pure...@googlegroups.com

I'm working on getting the pure 0.48 update into MacPorts.

On Oct 18, 2011, at 13:43, Albert Graef wrote:

> Also note that in any case you'll have to recompile *all* your addons; due to some ABI changes Pure 0.48 comes with a new version (8.0.0) of the Pure runtime library.

Not all of the add-ons link with the pure runtime library, including pure-gtk and pure-doc.
And some don't install any compiled files at all, like pure-g2.
I don't need to make our users reinstall those, right?

Albert Graef

unread,

Oct 20, 2011, 5:31:14 AM10/20/11

to pure...@googlegroups.com

On 10/20/2011 08:44 AM, Ryan Schmidt wrote:
> I'm working on getting the pure 0.48 update into MacPorts.

Thanks!

A quick heads up (also for the other package maintainers). I'm currently
rolling new tarballs with fixed up READMEs (i.e., expanded @version@ and
|today| tags) for *all* packages. This is only for the latest release of
each package, and serves purely cosmetic purposes. You might just ignore
these as there are no code changes whatsoever, but for those of you who
are building source packages (source debs, rpms, whatever) which contain
the upstream sources, it may be advisable to just wait with rolling and
uploading your packages until I'm finished with this (shouldn't take
more than an hour or so).

Ryan, this probably won't affect you, unless MacPorts also keeps their
own copies of the upstream tarballs somewhere.

> Not all of the add-ons link with the pure runtime library, including pure-gtk and pure-doc.
> And some don't install any compiled files at all, like pure-g2.
> I don't need to make our users reinstall those, right?

That's correct. I was just a bit lazy (or on the safe side, if you prefer).

Albert Graef

unread,

Oct 20, 2011, 8:02:51 AM10/20/11

to pure...@googlegroups.com

On 10/20/2011 11:31 AM, Albert Graef wrote:
> A quick heads up (also for the other package maintainers). I'm currently
> rolling new tarballs with fixed up READMEs (i.e., expanded @version@ and
> |today| tags) for *all* packages. This is only for the latest release of
> each package, and serves purely cosmetic purposes. You might just ignore
> these as there are no code changes whatsoever, but for those of you who
> are building source packages (source debs, rpms, whatever) which contain
> the upstream sources, it may be advisable to just wait with rolling and
> uploading your packages until I'm finished with this (shouldn't take
> more than an hour or so).

Ok, the new tarballs should all have landed by now. I did a scripted
upload, I'm still crossing my fingers that everything went right. :) So
please let me know if you notice that anything's gone missing, or if
some download seems to be corrupted or has the wrong description and/or
tags.

The READMEs in all packages should be fixed now. I looked at a few
samples and they all seemed fine. If you notice any remaining unexpanded
@version@ or |today| tags or other apparent weirdness in the README
files of the most recent packages, please let me know.

Thanks for your patience, and sorry for the hassle,

Ryan Schmidt

unread,

Oct 20, 2011, 4:47:26 PM10/20/11

to pure...@googlegroups.com

On Oct 20, 2011, at 04:31, Albert Graef wrote:
> On 10/20/2011 08:44 AM, Ryan Schmidt wrote:
>> I'm working on getting the pure 0.48 update into MacPorts.
>
> Thanks!
>
> A quick heads up (also for the other package maintainers). I'm currently rolling new tarballs with fixed up READMEs (i.e., expanded @version@ and |today| tags) for *all* packages. This is only for the latest release of each package, and serves purely cosmetic purposes. You might just ignore these as there are no code changes whatsoever, but for those of you who are building source packages (source debs, rpms, whatever) which contain the upstream sources, it may be advisable to just wait with rolling and uploading your packages until I'm finished with this (shouldn't take more than an hour or so).
>
> Ryan, this probably won't affect you, unless MacPorts also keeps their own copies of the upstream tarballs somewhere.

It does affect MacPorts, both in that we do have a network of servers [1] that automatically mirror tarballs when we commit updated portfiles, and in that we record checksums of each tarball in each portfile, and if upstream changes it, we'll have to update and re-test the portfiles. We call this a "stealth update" [2]. It's not a problem in this particular instance since I haven't committed the 0.48 update yet but in general we would love it if developers wouldn't do stealth updates. :)

[1] https://trac.macports.org/wiki/Mirrors#Distfiles

[2] https://trac.macports.org/wiki/PortfileRecipes#stealth-updates

Albert Graef

unread,

Oct 20, 2011, 6:08:59 PM10/20/11

to pure...@googlegroups.com

On 10/20/2011 10:47 PM, Ryan Schmidt wrote:
> It does affect MacPorts, both in that we do have a network of servers [1] that automatically mirror tarballs when we commit updated portfiles, and in that we record checksums of each tarball in each portfile, and if upstream changes it, we'll have to update and re-test the portfiles. We call this a "stealth update" [2]. It's not a problem in this particular instance since I haven't committed the 0.48 update yet but in general we would love it if developers wouldn't do stealth updates. :)

I understand that, and I'm sorry about this. I hope that I'll never have
to do this again. But the only way to resolve this specific issue was to
update all the tarballs for the current releases at some time. Otherwise
it might have taken months (at least) during the "normal" release
cycles. So I decided that it's better to just go ahead and get this
fixed *now* and be done with it.

In any case, I promise that I'll do all I can to resolve any technical
issues due to this update asap. I really appreciate your work to port
Pure to OSX.

Albert Graef

unread,

Oct 21, 2011, 1:05:17 PM10/21/11

to pure...@googlegroups.com

On 10/21/2011 12:08 AM, Albert Graef wrote:
> On 10/20/2011 10:47 PM, Ryan Schmidt wrote:
>> It does affect MacPorts, both in that we do have a network of servers
>> [1] that automatically mirror tarballs when we commit updated
>> portfiles, and in that we record checksums of each tarball in each
>> portfile, and if upstream changes it, we'll have to update and re-test
>> the portfiles. We call this a "stealth update" [2]. It's not a problem
>> in this particular instance since I haven't committed the 0.48 update
>> yet but in general we would love it if developers wouldn't do stealth
>> updates. :)
>
> I understand that, and I'm sorry about this.

Well, I really goofed on that one. :( The maintainer of the FreeBSD Pure
packages has complained, too. I didn't understand that this would cause
so much fallout, neither was I aware that there are already some systems
officially shipping a fairly complete set of Pure packages (other than
MacPorts). I should have known better. It won't happen again. Next time
an issue like this arises, I will just bump the version numbers.

Again, my apologies to all Pure package maintainers affected by this. I
really appreciate your work, and it wasn't my intention to make it any
harder than necessary. Well, I guess everybody does a silly blunder like
this every once in a while.

Reply all

Reply to author

Forward