Scribit Chris Laux dies 25/09/2007 hora 19:33:
> > Could you give a small testcase that demonstrates this?
> [...] I don't think a small test case would produce this, but that it
> is produced by a fairly large database (GBs) with >>1000 BTrees that
> was queried by >10 parallel threads.
Could you
Scribit Chris Laux dies 13/09/2007 hora 15:26:
> while using Elephant I've noticed that garbage collections now take up
> to 10 minutes with the (current model) CPU at 100% load!?
Could you give a small testcase that demonstrates this?
Curiously,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
Scribit Ian Eslick dies 03/09/2007 hora 23:38:
> However I haven't looked at the local tree recently so I'm not sure
> what shape it was in so would hesitate to make it available without
> that review.
You can make it available as a branch separate from the trunk, and
making such local development
Scribit Ian Eslick dies 30/08/2007 hora 21:45:
> There is some early prototype functionality that is partially in my
> local development tree and partially in the current development tree.
> Unfortunately I did not finish the query system this summer and so its
> release has been indefinitely delay
Scribit Robert L. Read dies 05/07/2007 hora 10:20:
> However, it uses tables more effectively, and uses stored procedures
> which are very Postgres-specific and not portable to a different
> database.
Well, I think that constitutes an answer to my earlier question about
the performance gains of st
Scribit Robert L. Read dies 04/07/2007 hora 17:49:
> The Postmodern backend is not quite twice as fast on our testsuite as
> the CL-SQL backend.
Great. I found the CL-SQL backend with PostgreSQL a bit slow for my
specific use case. Do you know why and how this backend is so much
faster?
> Postmod
Scribit Mariano Montone dies 27/06/2007 hora 13:12:
> I've been developing web applications so far, so I'll be describing a
> web UI.
I've been making web apps too, mostly. I don't think it's web specific,
but it's probably worse in web UIs, because they have a much worse
latency than GUIs running
Scribit Robert L. Read dies 31/05/2007 hora 22:32:
> This sounds very nice --- good luck!
Well, the paper has been accepted!
Gladly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature.asc
Description: Digital signature
___
elephant-devel site l
Scribit Pierre THIERRY dies 25/05/2007 hora 20:39:
> I will co-author the paper
I just submitted the paper. Here is the final abstract:
The data model of an application, the nature and format of data stored
across executions, is typically a very rigid part of its early
specification, e
Hi all,
I mentioned the Lisp workshop in ECOOP 2007, and my intention to try and
write a paper for it. Sadly, I did find time to do so before the may 13
deadline. I was very disappointed, so when they extended the deadline to
may 31, I took my bibliography and began to write.
I will co-author the
Scribit Ian Eslick dies 27/04/2007 hora 19:21:
> Actually at this point you shouldn't need to put locks around slot
> access.
It was a patch against 0.6.0, when I ran into trouble with a
multithreaded Web application.
> > > you need to create storage for at least the oid and
> > > store-controlle
Scribit Ian Eslick dies 27/04/2007 hora 15:37:
> The persistent slots of that class have the allocation parameter
> :database instead of :instance to capture these semantics.
Should'nt that be orthogonal? Currently, that means persistent class
allocated slots are not supported. I never had to use
Scribit Ian Eslick dies 23/04/2007 hora 18:55:
> I'm not sure that deserializing a persistent object without actually
> creating it would make much sense. What would you return from the
> deserializer? What operations would it need to support?
I tend to think that deserialization should be somet
Scribit Ian Eslick dies 22/04/2007 hora 10:39:
> However, if the objects are not already in-memory, the deserializer
> will call make-instance on that instance using the OID and class
> stored in the DB.
>
> This does not call initforms or use initargs but will trigger any
> initialize-instance
Scribit Ian Eslick dies 12/04/2007 hora 20:43:
> Tell us more!
I was thinking of generating a accessor/setf couple of functions that
would wrap access to a built-in type in a persistent class. When you wan
a persistent list, for example, it would use a class similar to one
defined by:
(defclass p
Just a thought about 1.0:
one thing that is absolutely needed is a comprehensive documentation of
the persistent metaclass. Some of its behaviour can be quite unexpected,
like the fact that an object goes through its initialization each time
it is deserialized.
When initialization of an object ha
Scribit Ian Eslick dies 12/04/2007 hora 13:17:
> I added a simple persistent set class, pset, to Elephant.
It make me think that I wanted to write a generic persistent wrapper for
built-in types. Do you think it should be included in Elephant?
Curiously,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D
Scribit Ian Eslick dies 11/04/2007 hora 13:44:
> This sounds intriguing! I'm not sure what you mean by data model,
> however.
I did that in my very first application with Elephant. It was a real
estate catalog. Several times, I had to add slots to the objects
describing the products.
All I had t
Scribit Robert L. Read dies 09/04/2007 hora 18:43:
> However, if someone else can go, I am also happy to help or to
> co-author a paper with them.
It occurred to me that the combination of persistence and the MOP goes
far beyond anything else I know in terms of flexibility WRT to the data
model.
Scribit Robert L. Read dies 11/04/2007 hora 09:01:
> The real problem here is that Elephant is so different, and so much
> more powerful and convenient than a relational system, that [...] if
> we use a relational benchmark, we are tying our hands behind our back.
We could also look at OO specific
Persistence is one of the desired topics, so maybe someone may be
willing to talk about Elephant:
http://lisp-ecoop07.bknr.net/home
Quickly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature.asc
Description: Digital signature
___
elephant-deve
Scribit Ian Eslick dies 03/04/2007 hora 14:42:
> However none of [the portability ports] change the user experience or
> significantly enhance the current operational stability of the
> project. I'd be willing to look at this happening on a separate
> branch post-1.0, but I think it would be a dist
Scribit Robert L. Read dies 03/04/2007 hora 11:08:
> Stored procedures tend to not be very portable; therefore to put them
> in the current "postgres" backend, which should really be called a
> "clsql" backend, would make it less likely to work with MySQL.
I was thinking at having some PostgreSQL-
Scribit Robert L. Read dies 03/04/2007 hora 11:30:
> I'm pleased that you think Elephant is ready for a 1.0 release, but
> unfortunately Ian and I disagree. Ian and I have a number of small
> features that we think are needed to be "1.0 complete", not lease of
> which is an off-line garbage collec
Scribit Ian Eslick dies 03/04/2007 hora 09:28:
> moving towards a 1.0 release in the near future. Therefore the
> current CVS (0.6.1 beta) will shortly become 0.9rc1.
Why use such a confusing numbering? There's no need for a 0.9 before
1.0, so you could as well have some "road-to-1.0" branch rec
Recently, a discussion started in the Torque project, an ODB for Java
that also provides persistence, about the possible use of stored
procedures instead of building SQL requests, because for frequent cheap
operations, the overhead of parsing and executing SQL can be
non-trivial.
Do you think such
Scribit Robert L. Read dies 02/04/2007 hora 08:17:
> We have a chapter called "Scenarios", in which we would like to put
> hypothetical, or even better, real, usages of Elephant.
I've only prototyped a real estate catalog for now.
BTW, is there anyone working on the tutorial? I'm wanting to write
Scribit Ian Eslick dies 30/03/2007 hora 08:50:
> I just took a peek at Mercurial - it's basically Darcs written in
> Python although with some C extensions.
Well, not quite. As some Debian developper said, as he switeched from
Darcs to Mercurial, they don't have the same philsophy. Darcs will avoi
Scribit Ian Eslick dies 30/03/2007 hora 13:22:
> Is there a way to implement class keyword arguments in a metaclass?
What do you mean?
> ensure-class-using-class will fail if it gets a keyword argument it
> doesn't recognize
Strange, because its signature includes &allow-other-keys...
Curiously
Scribit Erik Garrison dies 30/03/2007 hora 12:23:
> I'm trying to keep the myriad dependencies of my project within the
> subset of tools that can be asdf-installed. Otherwise the benefits of
> working in Lisp are quickly nullified.
If you're working with Linux, maybe your distro already has some
Scribit Robert L. Read dies 25/03/2007 hora 13:27:
> Obviously, if darcs doesn't work well with Windows, that is not a
> point in its favor.
As I said, if eventually you don't decide to use darcs, but wish to use
a DCVS, I strongly recommend Mercurial. It seems to work nicely on
Windows, has a cle
With latest HEAD, both BDB and SQLite3 passed the whole test suite
including migration on both i386 and amd64 under Debian GNU/Linux, with
SBCL 1.0.0.0.
Which is great news, because this release will come just as I need
Elephant in a project with similar constraint as the previous one, where
I had
Scribit Ian Eslick dies 24/03/2007 hora 12:19:
> Also, the new manual will be available online in texinfo-tree form, a
> single HTML file and in PDF. I'm happy to take your suggestion and
> supply a pre-build version of all these files in the distribution and
> source control.
FWIW, I think the l
Scribit Pierre THIERRY dies 24/03/2007 hora 05:52:
> BDB still gets stuck at:
>
> Migrating class indexes for: IDX-UNBOUND-DEL
I tried again, after running delscript.sh and setting rt::*catch-errors*
to nil, and got in the debugger
Scribit Ian Eslick dies 23/03/2007 hora 12:33:
> I think I've fixed this.
I tested again with latest HEAD, and end up with this for SQLite:
2 out of 122 total tests failed: MIGRATE-BASIC, MIGRATE-IPCLASS.
WARNING:
Unable to clear class index caches Object's store controller was lost
BDB still
With:
- Debian GNU/Linux 2.6.18 amd64
- SBCL 1.0.0.0
- BDB backend
Compilation failed with:
Berkeley DB error: Invalid argument
Here is the compiler output for the relevant compilation unit:
; file: /home/pierre/Lisp/Elephant/elephant-CVS/src/db-bdb/bdb-transactions.lisp
; in: DEFMETHOD
Scribit Ian Eslick dies 21/03/2007 hora 11:42:
> We have tagged CVS with the tag ELEPHANT-0-6-1-BETA.
With:
- Debian GNU/Linux 2.6.18 amd64
- SBCL 1.0.0.0
- SQLite3 backend
3 of 122 tests failed:
- MIGRATE-BASIC
- MIGRATE-IDX-BTREE
- MIGRATE-IPCLASS
I still have difficulties loading CL-SQL, exce
Scribit Ian Eslick dies 16/03/2007 hora 14:59:
> That would be a good ticket to add the Trac database
BTW, I can submit tickets but not edit the wiki, as I have no c-l.net
account. I emailed Erik Enge 9 days ago asking to get one, without
answer at this day.
Curiously,
Pierre
--
[EMAIL PROTECTE
Scribit Ian Eslick dies 16/03/2007 hora 14:59:
> As you say, I wasn't going to add a new version of the database
> format just to read alpha release databases, and so my question was
> to see if it was a big problem for someone.
It's an alpha release. Depending on it's behaviour is a mistake alrea
Scribit Ian Eslick dies 16/03/2007 hora 10:53:
> For various reasons I need to renumber the type tags which means that
> existing 0.6.1 databases will not be properly deserialized.
Coudln't the database format be versioned? We could then keep apart the
code that work on a given version, to alway
Scribit Aycan iRiCAN dies 10/03/2007 hora 14:03:
> Darcs has some performance penalties, git is another alternative.
As far as lispers are concerned, darcs is much more used, and known. But
if an alternate choice were to be made, I wouldn't go to git, which is
quite OS sensitive. At least it won't
Scribit Ian Eslick dies 25/01/2007 hora 19:59:
> We need a good set of unicode tests! I redesigned unicode support in
> 0.6.1 so I'm hoping these issues will go away, but I'd like to
> understand it.
I'll make some tests of my current application under 0.6.1 to see.
> What lisp are you using and
Scribit Pierre THIERRY dies 26/01/2007 hora 00:40:
> - I have "Idéal"
> - I want "Idéal"
Oh dear. My mail was sent in UTF-8 as Latin-1... I *hate* encoding
issues!
Quickly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature.asc
Desc
I'm working on a web application that uses 0.6.0, and I may have hit a
bug in Elephant.
I have a fairly reproducible bug, when storing a string. I sometimes
have to decode a badly-read string. E.g.:
- I have "Idéal"
- I want "Idéal"
For this I use a function that if the à character is found
Scribit Ian Eslick dies 22/01/2007 hora 18:37:
> - Most changes I just made are subsumed by testing the existing
> regressions on different platforms, such as running the regression on
> 64-bit machines with a 64-bit lisp.
I'll run the tests under Debian Etch on an amd64. The same box is
running i
Scribit Ian Eslick dies 22/01/2007 hora 16:09:
> If you want to help with the switch to darcs I'd prefer to use it too
> - anyone else want to weigh in on this?
I've now come to prefer DVCS like Darcs or Mercurial over classical like
CVS or Subversion, mostly because of the freedom they offer. I'l
Scribit Ian Eslick dies 21/01/2007 hora 17:21:
> How often do users of Elephant rely on BTrees where the symbols are
> ordered according to the symbol's name?
That's strange enough for me that I can't think of any use case at first
try...
> Most symbol serialization is highly redundant and can b
Scribit Ian Eslick dies 25/11/2006 hora 14:44:
> Trac and SVN
About the migration to trac: maybe everyone doesn't know, but I
discovered that it can be used with pretty any VCS widely used. I
personnally decided myself to switch to Mercurial and it was trivially
easy to add Mercurial support to it
Scribit Ian Eslick dies 20/01/2007 hora 16:54:
> Does anyone have a better suggestion here? For example is there a
> portability layer that can detect the current thread ID and use that
> to index the default global values?
http://trac.common-lisp.net/bordeaux-threads/wiki/ApiDocumentation#curren
Scribit Robert L. Read dies 20/01/2007 hora 14:06:
> I personally solve problem by using SBCL mutexes around the
> serialize/deserialize code, but I have not committed that code, since
> it is highly SBCL-specific, and in fact I don't know of a good way to
> do locking portably.
I used the bordeux
Scribit Gábor Melis dies 20/01/2007 hora 19:55:
> * guards against loading the pthreads library on threaded sbcl where
> the right version is already loaded and it crashes sbcl
I always used a threaded SBCL and did not notice any crash, are you sure
it's the problem?
Curiously,
Pierre
--
[EMAIL
Would you mind if a provide a patch that only add at least some generic
functions, with the only goal to remove the associated style-warnings?
Quickly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature.asc
Description: Digital signature
___
ele
Scribit Ian Eslick dies 25/11/2006 hora 13:37:
> Install 0.6.0, upgrade 0.5.0 -> 0.6.0 then install 0.6.1 and upgrade
> 0.6.0->0.6.1
Thansk, I'll try and see if there's anything I can do about it.
BTW, is there a way to help you migrate to svn/trac, or any other
combination of VCS/BTS?
Quickly,
Scribit Ian Eslick dies 24/11/2006 hora 17:32:
> I want to remove support for 0.5.0->0.6.0 upgrades. This makes the
> code cleaner as I found a problem with DB version tagging support I
> put in for 0.6.0 - I won't need to do this in the future.
Could you describe the problem you found? We may tr
Scribit Ian Eslick dies 21/11/2006 hora 01:30:
> It turns out that most implementations have a non-standard call
> 'without-interrupts' that allow you, on a single CPU architecture, to
> suspend interrupts and create a low-cost critical section.
Well, after some reading on it, I understand that
Scribit Alain Picard dies 17/11/2006 hora 21:18:
> This little snippet fixes this problem:
> (change from the 0.6 distribution)
Could you just provide the patch, that would be more easy both to read
and apply? (for me, at least)
Quickly,
Pierre
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
signature
If I ain't mistaken, currently Elephant only has a CVS, and no bug
tracker of any sort.
Before or after 0.6.1, could you consider switching to subversion and
trac, as they are provided by common-list.net? It could help possible
contributors to see what remains to be done (comments and attached
pat
Scribit Pierre THIERRY dies 10/11/2006 hora 03:00:
> > Could be the new SBCL not playing nice with some assumption left in
> > 0.6.0.
> I'll install a sarge chroot and test this within it.
Well, in a sarge chroot, I could run the test suite for the BDB backend
with SBCL. I ev
Scribit Daniel Salama dies 13/11/2006 hora 21:00:
> So, if I make my class persistent, it's persistent... period! I don't
> need to add it to a collection or to the root.
Well, I did not know that before that discussion. I think it's not
clearly indicated in the docs...
And then I wonder: if I us
Scribit Daniel Salama dies 12/11/2006 hora 10:28:
> I guess we would have to sequentially navigate thru the results in
> order to "manually" select each record based on all the other
> possible search arguments. I suppose, in a way, this can be done
> relatively painless by using macros
I d
Scribit Ian Eslick dies 10/11/2006 hora 23:15:
> Do we have a windows user who would be willing to test and tweak a new
> feature in HEAD (originally contributed by [EMAIL PROTECTED]) to
> build the libraries from asdf rather than by calling out to an
> external Makefile?
I have a Windows XP box o
Scribit Ian Eslick dies 09/11/2006 hora 20:24:
> Could be the new SBCL not playing nice with some assumption left in
> 0.6.0.
I'll install a sarge chroot and test this within it.
> Also, if you are going to be editing the MOP - better you do it off of
> the CVS HEAD rather than 0.6.0. So you kno
Scribit Ian Eslick dies 09/11/2006 hora 16:30:
> Did you recompile libsleepycat.c between tests? It depends on the
> db.h that you've set in the Makefile. That has to be set to the
> BDB-4.3 directory for 0.6.0
I did not set the directory in the Makefile, but always did the test
with a freshly c
Scribit Ian Eslick dies 09/11/2006 hora 17:27:
> PS - I'd vote to keep the current defaults but allow :persist
> :inherited-slots to override that. This avoids breaking backward
> compatibility for the existing user base.
That seems a sensible way of doing this, yes.
Compatibly,
Nowhere man
--
[
Scribit Ian Eslick dies 09/11/2006 hora 17:26:
> If the base class is persistent, then it's transient/persistent slot
> values will be propagated to the subclass. If it's a normal class,
> then all its slots become transient - their normal properties if in
> the baseclass.
I understand this point
Scribit Ian Eslick dies 11/10/2006 hora 11:29:
> If you have a non-persistent base class, should the semantics of the
> slot storage change based on whether it's included in a persistent
> subclass?
When you inherit from a class, you expect the sub-class to have all the
properties that the super-c
Scribit Pierre THIERRY dies 06/11/2006 hora 14:40:
> FWIW, since this patch [...]
And since I'm not that much proud of the patch, I forgot to attach it!
Quickly,
Nowhere man
--
[EMAIL PROTECTED]
OpenPGP 0xD9D50D8A
Index: elephant/elep
> > 2) Depending on the demands of your application, you might be able
> > to put in a smaller number of higher-level mutexes
> I may try that first.
Well. I just did the opposite. ;-)
As all my use of Elephant was through classes with the persistent
metaclass, I tried to protect all methods of
Scribit Ian Eslick dies 04/11/2006 hora 10:19:
> If you run db_deadlock on the open BDB database then all the
> transactions through BDB are thread safe.
I tried BDB, but without success. When doing (do-backend-tests) with the
default BDB spec, it stops when compiling the wrap-errno macro. I tried
Scribit Robert L. Read dies 31/10/2006 hora 08:54:
> 1) The serializer is definitely not thread-safe. [...]
>
> 2) CLSQL is not threadsafe unless you use a separate connection in
> each thread. [...]
Well, I was not aware of those problems. That is a quite huge step
forward.
> 1) If your acces
I may have a problem with concurrency and elephant. I'm developing a web
application using SBCL and Araneida (with it's threaded listener).
Whereas araneida by irself doesn't seem to have problems dealing with
many requests at a time (I tested it with apache's benckmark tool, with
thousands of requ
I'm happily using Elephant in a Web-based project right now, but I'm
getting annoyed by the behoviour of Elephant WRT inheritance. The
problem is that only direct slots of the class with the persistent
metaclass as metaclass are made persistent.
If I have the following classes:
(defclass foo ()
Hi,
I'm just about to try to package some Lisp systems I use for Debian, and
I noted some minor problems with the Elephant's tarballs:
- it contains an 'elephant' dir, instead of 'elephant-0.6.0' (if you
extract in a dir also containing the CVS dir, it will be a problem)
- it contains the CVS dir
73 matches
Mail list logo
