about fast lineDo and...

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
11 messages Options
Reply | Threaded
Open this post in threaded view
|

about fast lineDo and...

Stéphane Ducasse
Hi nicolas

I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
since in moose we are computing a lot LOC :)
By any chance could we get the same in pharo?
Else I will harvest them.

Stef
_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Nicolas Cellier
Don't know if I'll have time, but changes are fairly easy to grab
(independent of each other).

Nicolas

2010/1/19 Stéphane Ducasse <[hidden email]>:

> Hi nicolas
>
> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
> since in moose we are computing a lot LOC :)
> By any chance could we get the same in pharo?
> Else I will harvest them.
>
> Stef
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Stéphane Ducasse
ok I understand. I will keep your energy for the important one.
but since I'm nearly the only that have a look at fixes in squeak
may be I will simply just stop to look at them because the community
does not seem to interested in getting them or producing them for pharo.

Stef


On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:

> Don't know if I'll have time, but changes are fairly easy to grab
> (independent of each other).

Ok

>
> Nicolas
>
> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>> Hi nicolas
>>
>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>> since in moose we are computing a lot LOC :)
>> By any chance could we get the same in pharo?
>> Else I will harvest them.
>>
>> Stef
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Nicolas Cellier
2010/1/20 Stéphane Ducasse <[hidden email]>:
> ok I understand. I will keep your energy for the important one.

http://code.google.com/p/pharo/issues/detail?id=1856

> but since I'm nearly the only that have a look at fixes in squeak
> may be I will simply just stop to look at them because the community
> does not seem to interested in getting them or producing them for pharo.
>

Sure some people are interested as much as you are.
But waiting for someone to do the job for you, or wasting your own
time is rarely a dilemna except for the happy few owners of "own time"
;)

It's important to keep up with what's going on in other branches :
kind of technology forecasting :)
Just lurking the daily commit logs, and registering interesting stuff
as pharo issue is already valuable.

Nicolas

> Stef
>
>
> On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:
>
>> Don't know if I'll have time, but changes are fairly easy to grab
>> (independent of each other).
>
> Ok
>>
>> Nicolas
>>
>> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>>> Hi nicolas
>>>
>>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>>> since in moose we are computing a lot LOC :)
>>> By any chance could we get the same in pharo?
>>> Else I will harvest them.
>>>
>>> Stef
>>> _______________________________________________
>>> Pharo-project mailing list
>>> [hidden email]
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Stéphane Ducasse
Yes but I do not work on
        - new first class instance variable
        - byecode loader
        - new package to speed up everything
        - scripting syntax

so at the end I'm not sure what we lose but they are
simple actions that other people can do too to distribute the load.
So may be I will stop looking at squeak log.

Stef

> 2010/1/20 Stéphane Ducasse <[hidden email]>:
>> ok I understand. I will keep your energy for the important one.
>
> http://code.google.com/p/pharo/issues/detail?id=1856
>
>> but since I'm nearly the only that have a look at fixes in squeak
>> may be I will simply just stop to look at them because the community
>> does not seem to interested in getting them or producing them for pharo.
>>
>
> Sure some people are interested as much as you are.
> But waiting for someone to do the job for you, or wasting your own
> time is rarely a dilemna except for the happy few owners of "own time"
> ;)
>
> It's important to keep up with what's going on in other branches :
> kind of technology forecasting :)
> Just lurking the daily commit logs, and registering interesting stuff
> as pharo issue is already valuable.
>
> Nicolas
>
>> Stef
>>
>>
>> On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:
>>
>>> Don't know if I'll have time, but changes are fairly easy to grab
>>> (independent of each other).
>>
>> Ok
>>>
>>> Nicolas
>>>
>>> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>>>> Hi nicolas
>>>>
>>>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>>>> since in moose we are computing a lot LOC :)
>>>> By any chance could we get the same in pharo?
>>>> Else I will harvest them.
>>>>
>>>> Stef
>>>> _______________________________________________
>>>> Pharo-project mailing list
>>>> [hidden email]
>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>> [hidden email]
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Nicolas Cellier
2010/1/20 Stéphane Ducasse <[hidden email]>:
> Yes but I do not work on
>        - new first class instance variable
>        - byecode loader
>        - new package to speed up everything
>        - scripting syntax
>

Never mind, when I said "you", it was not specifically you, but anyone
else prefering to wait because of no free time.

> so at the end I'm not sure what we lose but they are
> simple actions that other people can do too to distribute the load.
> So may be I will stop looking at squeak log.
>

Sure, the danger is to become a follower of random activity driven by others.
If time is sparse, you might prefer to keep control.
But some changes like file buffering really makes a difference.

> Stef
>
>> 2010/1/20 Stéphane Ducasse <[hidden email]>:
>>> ok I understand. I will keep your energy for the important one.
>>
>> http://code.google.com/p/pharo/issues/detail?id=1856
>>
>>> but since I'm nearly the only that have a look at fixes in squeak
>>> may be I will simply just stop to look at them because the community
>>> does not seem to interested in getting them or producing them for pharo.
>>>
>>
>> Sure some people are interested as much as you are.
>> But waiting for someone to do the job for you, or wasting your own
>> time is rarely a dilemna except for the happy few owners of "own time"
>> ;)
>>
>> It's important to keep up with what's going on in other branches :
>> kind of technology forecasting :)
>> Just lurking the daily commit logs, and registering interesting stuff
>> as pharo issue is already valuable.
>>
>> Nicolas
>>
>>> Stef
>>>
>>>
>>> On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:
>>>
>>>> Don't know if I'll have time, but changes are fairly easy to grab
>>>> (independent of each other).
>>>
>>> Ok
>>>>
>>>> Nicolas
>>>>
>>>> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>>>>> Hi nicolas
>>>>>
>>>>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>>>>> since in moose we are computing a lot LOC :)
>>>>> By any chance could we get the same in pharo?
>>>>> Else I will harvest them.
>>>>>
>>>>> Stef
>>>>> _______________________________________________
>>>>> Pharo-project mailing list
>>>>> [hidden email]
>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>>
>>>>
>>>> _______________________________________________
>>>> Pharo-project mailing list
>>>> [hidden email]
>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>> [hidden email]
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Stéphane Ducasse
>
>> Yes but I do not work on
>>        - new first class instance variable
>>        - byecode loader
>>        - new package to speed up everything
>>        - scripting syntax
>>
>
> Never mind, when I said "you", it was not specifically you, but anyone
> else prefering to wait because of no free time.
>
>> so at the end I'm not sure what we lose but they are
>> simple actions that other people can do too to distribute the load.
>> So may be I will stop looking at squeak log.
>>
>
> Sure, the danger is to become a follower of random activity driven by others.
> If time is sparse, you might prefer to keep control.
> But some changes like file buffering really makes a difference.

Yes but tracking changes you do not know is also taking time.
I started but so far failed two times because you have to go slowly not to break too much the system.
Tomorrow I will follow the advice of levente and check.
The problem also is that the level of the tools is not really good.
Some behavior of the history browser add to changeset other not.

We will have to improve that.


>> Stef
>>
>>> 2010/1/20 Stéphane Ducasse <[hidden email]>:
>>>> ok I understand. I will keep your energy for the important one.
>>>
>>> http://code.google.com/p/pharo/issues/detail?id=1856
>>>
>>>> but since I'm nearly the only that have a look at fixes in squeak
>>>> may be I will simply just stop to look at them because the community
>>>> does not seem to interested in getting them or producing them for pharo.
>>>>
>>>
>>> Sure some people are interested as much as you are.
>>> But waiting for someone to do the job for you, or wasting your own
>>> time is rarely a dilemna except for the happy few owners of "own time"
>>> ;)
>>>
>>> It's important to keep up with what's going on in other branches :
>>> kind of technology forecasting :)
>>> Just lurking the daily commit logs, and registering interesting stuff
>>> as pharo issue is already valuable.
>>>
>>> Nicolas
>>>
>>>> Stef
>>>>
>>>>
>>>> On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:
>>>>
>>>>> Don't know if I'll have time, but changes are fairly easy to grab
>>>>> (independent of each other).
>>>>
>>>> Ok
>>>>>
>>>>> Nicolas
>>>>>
>>>>> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>>>>>> Hi nicolas
>>>>>>
>>>>>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>>>>>> since in moose we are computing a lot LOC :)
>>>>>> By any chance could we get the same in pharo?
>>>>>> Else I will harvest them.
>>>>>>
>>>>>> Stef
>>>>>> _______________________________________________
>>>>>> Pharo-project mailing list
>>>>>> [hidden email]
>>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Pharo-project mailing list
>>>>> [hidden email]
>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>
>>>>
>>>> _______________________________________________
>>>> Pharo-project mailing list
>>>> [hidden email]
>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>> [hidden email]
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project


_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Levente Uzonyi-2
On Wed, 20 Jan 2010, Stéphane Ducasse wrote:

>>
>>> Yes but I do not work on
>>>        - new first class instance variable
>>>        - byecode loader
>>>        - new package to speed up everything
>>>        - scripting syntax
>>>
>>
>> Never mind, when I said "you", it was not specifically you, but anyone
>> else prefering to wait because of no free time.
>>
>>> so at the end I'm not sure what we lose but they are
>>> simple actions that other people can do too to distribute the load.
>>> So may be I will stop looking at squeak log.
>>>
>>
>> Sure, the danger is to become a follower of random activity driven by others.
>> If time is sparse, you might prefer to keep control.
>> But some changes like file buffering really makes a difference.
>
> Yes but tracking changes you do not know is also taking time.
> I started but so far failed two times because you have to go slowly not to break too much the system.
> Tomorrow I will follow the advice of levente and check.
> The problem also is that the level of the tools is not really good.
That shouldn't be hard: in a updated Trunk image, file out
MultiByteFileStream and StandardFileStream. Open a File Browser in your
Pharo image select one of the files, right click and select "code-file
browser". This will show you all the methods in the file and will diff it
with the current versions (if it exists). The annotation pane shows the
timestamp and information about the method (new, identical, modified, etc).


Levente

> Some behavior of the history browser add to changeset other not.
>
> We will have to improve that.
>
>
>>> Stef
>>>
>>>> 2010/1/20 Stéphane Ducasse <[hidden email]>:
>>>>> ok I understand. I will keep your energy for the important one.
>>>>
>>>> http://code.google.com/p/pharo/issues/detail?id=1856
>>>>
>>>>> but since I'm nearly the only that have a look at fixes in squeak
>>>>> may be I will simply just stop to look at them because the community
>>>>> does not seem to interested in getting them or producing them for pharo.
>>>>>
>>>>
>>>> Sure some people are interested as much as you are.
>>>> But waiting for someone to do the job for you, or wasting your own
>>>> time is rarely a dilemna except for the happy few owners of "own time"
>>>> ;)
>>>>
>>>> It's important to keep up with what's going on in other branches :
>>>> kind of technology forecasting :)
>>>> Just lurking the daily commit logs, and registering interesting stuff
>>>> as pharo issue is already valuable.
>>>>
>>>> Nicolas
>>>>
>>>>> Stef
>>>>>
>>>>>
>>>>> On Jan 20, 2010, at 8:27 AM, Nicolas Cellier wrote:
>>>>>
>>>>>> Don't know if I'll have time, but changes are fairly easy to grab
>>>>>> (independent of each other).
>>>>>
>>>>> Ok
>>>>>>
>>>>>> Nicolas
>>>>>>
>>>>>> 2010/1/19 Stéphane Ducasse <[hidden email]>:
>>>>>>> Hi nicolas
>>>>>>>
>>>>>>> I saw that you commit some speed up for lineending to the squeaktrunk and I would love to have them
>>>>>>> since in moose we are computing a lot LOC :)
>>>>>>> By any chance could we get the same in pharo?
>>>>>>> Else I will harvest them.
>>>>>>>
>>>>>>> Stef
>>>>>>> _______________________________________________
>>>>>>> Pharo-project mailing list
>>>>>>> [hidden email]
>>>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> Pharo-project mailing list
>>>>>> [hidden email]
>>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Pharo-project mailing list
>>>>> [hidden email]
>>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>>>
>>>>
>>>> _______________________________________________
>>>> Pharo-project mailing list
>>>> [hidden email]
>>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>>
>>> _______________________________________________
>>> Pharo-project mailing list
>>> [hidden email]
>>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>>>
>>
>> _______________________________________________
>> Pharo-project mailing list
>> [hidden email]
>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
>
> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>
_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Stéphane Ducasse
>
> Yes but I do not work on
>>>>       - new first class instance variable
>>>>       - byecode loader
>>>>       - new package to speed up everything
>>>>       - scripting syntax
>>>>
>>>
>>> Never mind, when I said "you", it was not specifically you, but anyone
>>> else prefering to wait because of no free time.
>>>
>>>> so at the end I'm not sure what we lose but they are
>>>> simple actions that other people can do too to distribute the load.
>>>> So may be I will stop looking at squeak log.
>>>>
>>>
>>> Sure, the danger is to become a follower of random activity driven by others.
>>> If time is sparse, you might prefer to keep control.
>>> But some changes like file buffering really makes a difference.
>>
>> Yes but tracking changes you do not know is also taking time.
>> I started but so far failed two times because you have to go slowly not to break too much the system.
>> Tomorrow I will follow the advice of levente and check.
>> The problem also is that the level of the tools is not really good.
>
> That shouldn't be hard: in a updated Trunk image, file out MultiByteFileStream and StandardFileStream. Open a File Browser in your Pharo image select one of the files, right click and select "code-file browser". This will show you all the methods in the file and will diff it with the current versions (if it exists). The annotation pane shows the timestamp and information about the method (new, identical, modified, etc).

ok I will try like that.
usually I check all the deltas in chronological order and skip what is not needed and wished.

Stef
_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Nicolas Cellier
2010/1/21 Stéphane Ducasse <[hidden email]>:

>>
>> Yes but I do not work on
>>>>>       - new first class instance variable
>>>>>       - byecode loader
>>>>>       - new package to speed up everything
>>>>>       - scripting syntax
>>>>>
>>>>
>>>> Never mind, when I said "you", it was not specifically you, but anyone
>>>> else prefering to wait because of no free time.
>>>>
>>>>> so at the end I'm not sure what we lose but they are
>>>>> simple actions that other people can do too to distribute the load.
>>>>> So may be I will stop looking at squeak log.
>>>>>
>>>>
>>>> Sure, the danger is to become a follower of random activity driven by others.
>>>> If time is sparse, you might prefer to keep control.
>>>> But some changes like file buffering really makes a difference.
>>>
>>> Yes but tracking changes you do not know is also taking time.
>>> I started but so far failed two times because you have to go slowly not to break too much the system.
>>> Tomorrow I will follow the advice of levente and check.
>>> The problem also is that the level of the tools is not really good.
>>
>> That shouldn't be hard: in a updated Trunk image, file out MultiByteFileStream and StandardFileStream. Open a File Browser in your Pharo image select one of the files, right click and select "code-file browser". This will show you all the methods in the file and will diff it with the current versions (if it exists). The annotation pane shows the timestamp and information about the method (new, identical, modified, etc).
>
> ok I will try like that.
> usually I check all the deltas in chronological order and skip what is not needed and wished.
>
> Stef

Good old change set ?

Nicolas

> _______________________________________________
> Pharo-project mailing list
> [hidden email]
> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
>

_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
Reply | Threaded
Open this post in threaded view
|

Re: [Pharo-project] about fast lineDo and...

Stéphane Ducasse
> ok I will try like that.
>> usually I check all the deltas in chronological order and skip what is not needed and wished.
>>
>> Stef
>
> Good old change set ?

yes I use that from time to time.

Stef
_______________________________________________
Pharo-project mailing list
[hidden email]
http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project