Hey everyone,
Due to a prior commitment, I have to bow out of the hack night
tomorrow. My slot opened for anybody who wants it.
Cheers
Brett
David Lee wrote:
thanks Mr X,
If I tried hacking on the node AMQP lib at the same time I think my
head would asplode, but thanks for the heads up -
I like having an empty database for each test too, but I'd be more
than willing to trade that for faster tests.
If it's not solved before then, anyone fancy spending some time at
Railscamp working on making this happen?
- Pete
--
You received this message because you are subscribed to the Googl
Has anyone in this thread tried using an in-memory database rather than looking
at using SSD or re-inventing machinist?
Both sqlite and mysql have in-memory databases or tables. I'm sure that other
rdbms implementations would also.
I've used sqlite :memory: in the past when using datamapper and
Last time i benchmarked it vs postgresql tuned for testing (no fsync, etc)
PG came out on top on the suite i was running by a reasonable margin. If
memory serves ... i think it was partly because sqlite is so crippled (no
subselects, etc) it had to run (sometimes vastly) more queries to do the
equi
Hi there, my first post.
The Mysql in-memory database is pretty basic; sort of like MyISAM, but
with even fewer features. To me it sort of defeats most of the
purposes of testing to test against a completely different database
engine (that doesn't even try to support the same features). I think
Other option may be to still use the same database engine but have a
transparent caching layer in between.
I really wanted to get people to step back away from the problem with a
solution and back to the problem itself.
On 23/03/2010, at 10:48 AM, Simon Russell wrote:
> Hi there, my first po
If you're hand-writing complex queries that for one reason or another
rely on features that are only available to one SGDB in particular,
then here's part of the cost for doing so.
I wouldn't say it means the idea of in-memory databases is a bad one
unless you're leaning a lot of your logic agains
Sounds like argument type checking would be a useful extension to my
chattr gem...
which never made it off RubyForge. If folk think it's useful, I'll
drag it to github.
At the least, there's some code that could be shared.
Clifford Heath.
On 13/03/2010, at 12:01 PM, Xavier Shay wrote:
hell
OK kids,
One case of beer (24 bottles) has been purchased. That's enough for
everyone to have one beer, and for most to have two. If you'd like
more beer, please feel free to stop by the BWS around the corner from
the office (in the Forum, next to the Coles Express).
See ya anytime after 6pm!
-
I'll probably be buying soft drink (non-caffeinated). Does anyone else
have any orders?
Andrew
Who downloaded a copy of the yeast genome last night. I've also got a
copy of the human genome.
On Tue, Mar 23, 2010 at 1:40 PM, Gabe Hollombe wrote:
> OK kids,
>
> One case of beer (24 bottles) has be
Another possibility would be to create a RAM disk setup for testing -
that way you get proper database features but nice and fast access and
tests. I am not sure how straightforward this is, but I'm confident
that it's doable.
On 23 March 2010 10:48, Simon Russell wrote:
> Hi there, my first pos
Doable, but not really practical for a development environment.
Most of the laptops used by the ruby community will be SSD in a year, right?
On 23/03/2010, at 2:15 PM, James Sadler wrote:
> Another possibility would be to create a RAM disk setup for testing -
> that way you get proper database
Anything scriptable should be practical, assuming it doesn't take too
much time to run the script. But you should only have to run the
script when you boot up your machine, which is infrequent. It's just
some config that says "put your tables in this directory (on the RAM
disk) instead of the usua
Guys, as noted earlier in this conversation a couple of us have done this
and it made no appreciable difference. The databases engines just weren't
dumb enough to write to the disk when the transactions are all being rolled
backed at the end of each test anyway.
Cheers,
Anthony
On Tue, Mar 23,
On 23/03/2010, at 1:52 PM, Andrew Grimm wrote:
Who['s] also got a copy of the human genome.
I have trillions of copies. None machine-readable unfortunately,
except on one old machine I've had, like, forever - and I haven't
figured out how to bootload any new code onto that one yet. Let
me know
On Tue, Mar 23, 2010 at 3:30 PM, Clifford Heath
wrote:
> On 23/03/2010, at 1:52 PM, Andrew Grimm wrote:
>>
>> Who['s] also got a copy of the human genome.
>
> I have trillions of copies. None machine-readable unfortunately,
> except on one old machine I've had, like, forever - and I haven't
> figu
On 23/03/2010, at 12:16 AM, Brett Goulder wrote:
Due to a prior commitment, I have to bow out of the hack night
tomorrow. My slot opened for anybody who wants it.
I can't make it either gang! I meant to remove myself from the wiki
long ago, but it seems I must have simply imagined removing
17 matches
Mail list logo
