You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@commons.apache.org by "Bruno P. Kinoshita" <br...@yahoo.com.br> on 2012/06/01 06:31:48 UTC

Re: [jcs] Request for review

 Hi, 

[jcs] seems very interesting :-) and maybe I could use it in some applications too (or at least knowing about it may help).

I'm not an expert in generics, but some time ago similar task was done in [functor], maybe we could use that as base. I will read the changes in [functor] and will try to review the warnings in [jcs], and then comment or propose patches.

In case you would like to take a look on what was done in [functor], check r1188373 until r1188409 more or less. There may be more to be done in [functor] as it hasn't been released yet, but the new version with generics is working perfectly, no broken tests, no changes in the functionality, and I think there are no warnings. 


Hope that helps,

Bruno P. Kinoshita
http://kinoshita.eti.br
http://tupilabs.com

>________________________________
> From: Thomas Vandahl <tv...@apache.org>
>To: Commons Developers List <de...@commons.apache.org> 
>Sent: Thursday, 31 May 2012 1:31 PM
>Subject: Re: [jcs] Request for review
> 
>On 31.05.12 03:09, sebb wrote:
>> DoubleLinkedList looks incomplete.
>> It is parameterised by the node type (which must extend
>> DoubleLinkedListNode), but the node content type is not specified.
>> 
>> Unfortunately, I don't know how to fix this as I am no generics expert either.
>
>Yeah, I failed on this one, too. I decided not to silence the warnings
>so that something reminds me...
>
>> Eclipse generates a lot of other generics-related warnings, so it
>> could take a while to fix them.
>
>I know, it's work in progress. Thanks for your contributions.
>If I just were able to fix those spurious test failures.
>
>Bye, Thomas.
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>For additional commands, e-mail: dev-help@commons.apache.org
>
>
>
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Re: [jcs] Request for review

Posted by Gary Gregory <ga...@gmail.com>.
Hi all,

Note that replacing arrays with collections should not be done without
understanding the performance implications...

Gary

On Jun 5, 2012, at 1:14, "Bruno P. Kinoshita"
<br...@yahoo.com.br> wrote:

> Hi Thomas,
>
>> Thanks, I'm trying to pull things straight to make it usable it current
>> environments. I use the old version a lot. The code has more than 400
>> unit tests and a few disabled ones. Some tests tend to fail every now
>> and then (timing dependent). Those problems are hard to spot. I could
>> need some help...
>
> I don't have a project using [jcs], but I can help debugging, running tests in Linux with different JVM's, and other minor issues :-) As I have used ehcache in some JEE projects, I'm specially interested in learning if it is a good idea to replace ehcache by [jcs], as it looks like it has better performance [1] and more features.
>
>> I'm afraid this is not enough information to understand what you mean.
>> Could you be a bit more specific, please?
>
> Sure. I was only suggesting that maybe you could try adding generics to [jcs] and when in doubt, consult the [functor] SVN history and look for what was done. But I had a look on [jcs] code, and its codebase is much bigger than [functor]'s. And I've also noticed that some of the warnings are due generics in arrays. In some cases it is simply not possible to generify the types used in arrays (you can use collections, or suppress the warning), not sure if it is the case in [jcs] though.
>
> I will read more the [jcs] codebase to see if I can spot some places to generify the code and remove some warnings :-) There is something about generics in arrays, and 'generifying' legacy code in [2].
>
> All the best,
>
> [1] http://commons.apache.org/jcs/JCSvsEHCache.html
> [2] http://java.sun.com/j2se/1.5/pdf/generics-tutorial.pdf
>
> Bruno P. Kinoshita
> http://www.kinoshita.eti.br
> http://www.tupilabs.com
>
> On 06/02/2012 04:28 PM, Thomas Vandahl wrote:
>> On 01.06.12 06:31, Bruno P. Kinoshita wrote:
>>>  Hi,
>>>
>>> [jcs] seems very interesting :-) and maybe I could use it in some applications too (or at least knowing about it may help).
>>
>> Thanks, I'm trying to pull things straight to make it usable it current
>> environments. I use the old version a lot. The code has more than 400
>> unit tests and a few disabled ones. Some tests tend to fail every now
>> and then (timing dependent). Those problems are hard to spot. I could
>> need some help...
>>
>>> I'm not an expert in generics, but some time ago similar task was done in [functor], maybe we could use that as base. I will read the changes in [functor] and will try to review the warnings in [jcs], and then comment or propose patches.
>>
>> Well, my impression with generics is that you can overdo it easily.
>> Maybe it's not necessary in the deeper layers of the library and it only
>> causes trouble. That's why I'm asking for reviews.
>>
>>>
>>> In case you would like to take a look on what was done in [functor], check r1188373 until r1188409 more or less. There may be more to be done in [functor] as it hasn't been released yet, but the new version with generics is working perfectly, no broken tests, no changes in the functionality, and I think there are no warnings.
>>
>> I'm afraid this is not enough information to understand what you mean.
>> Could you be a bit more specific, please?
>>
>> Bye, Thomas.
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
>> For additional commands, e-mail: dev-help@commons.apache.org
>>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Re: [jcs] Request for review

Posted by "Bruno P. Kinoshita" <br...@yahoo.com.br>.
Hi Thomas,

> Thanks, I'm trying to pull things straight to make it usable it current
> environments. I use the old version a lot. The code has more than 400
> unit tests and a few disabled ones. Some tests tend to fail every now
> and then (timing dependent). Those problems are hard to spot. I could
> need some help...

I don't have a project using [jcs], but I can help debugging, running 
tests in Linux with different JVM's, and other minor issues :-) As I 
have used ehcache in some JEE projects, I'm specially interested in 
learning if it is a good idea to replace ehcache by [jcs], as it looks 
like it has better performance [1] and more features.

> I'm afraid this is not enough information to understand what you mean.
> Could you be a bit more specific, please?

Sure. I was only suggesting that maybe you could try adding generics to 
[jcs] and when in doubt, consult the [functor] SVN history and look for 
what was done. But I had a look on [jcs] code, and its codebase is much 
bigger than [functor]'s. And I've also noticed that some of the warnings 
are due generics in arrays. In some cases it is simply not possible to 
generify the types used in arrays (you can use collections, or suppress 
the warning), not sure if it is the case in [jcs] though.

I will read more the [jcs] codebase to see if I can spot some places to 
generify the code and remove some warnings :-) There is something about 
generics in arrays, and 'generifying' legacy code in [2].

All the best,

[1] http://commons.apache.org/jcs/JCSvsEHCache.html
[2] http://java.sun.com/j2se/1.5/pdf/generics-tutorial.pdf

Bruno P. Kinoshita
http://www.kinoshita.eti.br
http://www.tupilabs.com

On 06/02/2012 04:28 PM, Thomas Vandahl wrote:
> On 01.06.12 06:31, Bruno P. Kinoshita wrote:
>>   Hi,
>>
>> [jcs] seems very interesting :-) and maybe I could use it in some applications too (or at least knowing about it may help).
>
> Thanks, I'm trying to pull things straight to make it usable it current
> environments. I use the old version a lot. The code has more than 400
> unit tests and a few disabled ones. Some tests tend to fail every now
> and then (timing dependent). Those problems are hard to spot. I could
> need some help...
>
>> I'm not an expert in generics, but some time ago similar task was done in [functor], maybe we could use that as base. I will read the changes in [functor] and will try to review the warnings in [jcs], and then comment or propose patches.
>
> Well, my impression with generics is that you can overdo it easily.
> Maybe it's not necessary in the deeper layers of the library and it only
> causes trouble. That's why I'm asking for reviews.
>
>>
>> In case you would like to take a look on what was done in [functor], check r1188373 until r1188409 more or less. There may be more to be done in [functor] as it hasn't been released yet, but the new version with generics is working perfectly, no broken tests, no changes in the functionality, and I think there are no warnings.
>
> I'm afraid this is not enough information to understand what you mean.
> Could you be a bit more specific, please?
>
> Bye, Thomas.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org


Re: [jcs] Request for review

Posted by Thomas Vandahl <tv...@apache.org>.
On 01.06.12 06:31, Bruno P. Kinoshita wrote:
>  Hi, 
> 
> [jcs] seems very interesting :-) and maybe I could use it in some applications too (or at least knowing about it may help).

Thanks, I'm trying to pull things straight to make it usable it current
environments. I use the old version a lot. The code has more than 400
unit tests and a few disabled ones. Some tests tend to fail every now
and then (timing dependent). Those problems are hard to spot. I could
need some help...

> I'm not an expert in generics, but some time ago similar task was done in [functor], maybe we could use that as base. I will read the changes in [functor] and will try to review the warnings in [jcs], and then comment or propose patches.

Well, my impression with generics is that you can overdo it easily.
Maybe it's not necessary in the deeper layers of the library and it only
causes trouble. That's why I'm asking for reviews.

> 
> In case you would like to take a look on what was done in [functor], check r1188373 until r1188409 more or less. There may be more to be done in [functor] as it hasn't been released yet, but the new version with generics is working perfectly, no broken tests, no changes in the functionality, and I think there are no warnings. 

I'm afraid this is not enough information to understand what you mean.
Could you be a bit more specific, please?

Bye, Thomas.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org