about package naming

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

about package naming

stepharo
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following
the seaside convention - and I do not understand it either) why the test
package for FileSystem-Core is named FileSystem-Tests-Core

and not just

     FileSystem-Core-Tests

     like that we could have

         FileSystem-Core-Help

         FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically
FileSystem-Core-Tests

Now we cannot.

Stef


Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Uko2
Hi, I like the FileSystem-Core-Tests more, because then we can also have FileSystem-Core-Rules and FileSystem-Core-Rules-Tests. But I think that the other way to think about this is that if you want to (un)load all the test from filesystem you just work with FileSystem-Tests and not do it individually for each submodule. However this is just my guess.

Uko

> On 04 Sep 2016, at 10:17, stepharo <[hidden email]> wrote:
>
> Hi guys
>
> Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core
>
> and not just
>
>    FileSystem-Core-Tests
>
>    like that we could have
>
>        FileSystem-Core-Help
>
>        FileSystem-Core-Examples
>
> And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests
>
> Now we cannot.
>
> Stef
>
>


Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Tudor Girba-2
In reply to this post by stepharo
Hi,

That naming convention is due to the fact that there was a time when Monticello committed the content of packages based on prefix matching.

So, if you had:
        FileSystem-Core
        FileSystem-Core-Tests
committing FileSyste-Core would also commit FileSystem-Core-Tests as a subcategory.

Actually, we still have the bug that if you would change something in FileSystem-Core, FileSystem-Core-Tests would be marked as dirty.

Still, it would be much better to have naming of a package reflect a categorization. So, FileSystem-Core-Tests would be much more preferable.

Cheers,
Doru


> On Sep 4, 2016, at 10:17 AM, stepharo <[hidden email]> wrote:
>
> Hi guys
>
> Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core
>
> and not just
>
>    FileSystem-Core-Tests
>
>    like that we could have
>
>        FileSystem-Core-Help
>
>        FileSystem-Core-Examples
>
> And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests
>
> Now we cannot.
>
> Stef
>
>

--
www.tudorgirba.com
www.feenk.com

"If you can't say why something is relevant,
it probably isn't."


Reply | Threaded
Open this post in threaded view
|

Re: about package naming

CyrilFerlicot
Le 04/09/2016 à 20:52, Tudor Girba a écrit :

> Hi,
>
> That naming convention is due to the fact that there was a time when Monticello committed the content of packages based on prefix matching.
>
> So, if you had:
> FileSystem-Core
> FileSystem-Core-Tests
> committing FileSyste-Core would also commit FileSystem-Core-Tests as a subcategory.
>
> Actually, we still have the bug that if you would change something in FileSystem-Core, FileSystem-Core-Tests would be marked as dirty.
>
Exactly!

At Synectique we had to rename a lot of packages because this bug was
really annoying to know what to commit.

We opened an issue:
https://pharo.fogbugz.com/f/cases/18712/Monticello-have-strange-behavior-if-a-package-begin-by-the-name-of-another-package

> Still, it would be much better to have naming of a package reflect a categorization. So, FileSystem-Core-Tests would be much more preferable.
>
> Cheers,
> Doru
>
>
> --
> www.tudorgirba.com
> www.feenk.com
>
> "If you can't say why something is relevant,
> it probably isn't."
>
>

--
Cyril Ferlicot

http://www.synectique.eu

2 rue Jacques Prévert 01,
59650 Villeneuve d'ascq France


signature.asc (836 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Guillermo Polito
In reply to this post by stepharo

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef



DSC_0214.JPG (146K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Uko2
In reply to this post by Tudor Girba-2

On 04 Sep 2016, at 20:52, Tudor Girba <[hidden email]> wrote:

Actually, we still have the bug that if you would change something in FileSystem-Core, FileSystem-Core-Tests would be marked as dirty.

I think that this happens because we pretend to have packages, but behind the scenes it is still based on the old smalltalk categories…

Uko
Reply | Threaded
Open this post in threaded view
|

Re: about package naming

stepharo
In reply to this post by CyrilFerlicot


Le 4/9/16 à 21:15, Cyril Ferlicot D. a écrit :

> Le 04/09/2016 à 20:52, Tudor Girba a écrit :
>> Hi,
>>
>> That naming convention is due to the fact that there was a time when Monticello committed the content of packages based on prefix matching.
>>
>> So, if you had:
>> FileSystem-Core
>> FileSystem-Core-Tests
>> committing FileSyste-Core would also commit FileSystem-Core-Tests as a subcategory.
>>
>> Actually, we still have the bug that if you would change something in FileSystem-Core, FileSystem-Core-Tests would be marked as dirty.
>>
> Exactly!
>
> At Synectique we had to rename a lot of packages because this bug was
> really annoying to know what to commit.
>
> We opened an issue:
> https://pharo.fogbugz.com/f/cases/18712/Monticello-have-strange-behavior-if-a-package-begin-by-the-name-of-another-package

You should mention it on the mailing list when you open a bug and this
one is important.

>
>> Still, it would be much better to have naming of a package reflect a categorization. So, FileSystem-Core-Tests would be much more preferable.
>>
>> Cheers,
>> Doru
>>
>>
>> --
>> www.tudorgirba.com
>> www.feenk.com
>>
>> "If you can't say why something is relevant,
>> it probably isn't."
>>
>>
>


Reply | Threaded
Open this post in threaded view
|

Re: about package naming

stepharo
In reply to this post by Guillermo Polito

No guillermo. This is really annoying.

With long package list like Collections* then you do not see well the tests.

So we should check the bug like that we are free.

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef



Reply | Threaded
Open this post in threaded view
|

Re: about package naming

stepharo
In reply to this post by Guillermo Polito
Guille

do you know if iceberg has the same problem?
I do not remember what nicolas mentionned about if MC entities were kept
around and used.
Stef

Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Guillermo Polito
In reply to this post by stepharo
<offtopic>
Wow, I do not know where that email came from O_o.
It looks a spammish email. Maybe it came from my phone?

If it happens again, please tell me so I can take action.
</offtopic>

Now in-topic: I agree with having conventions and with the tests one particularly. But as you say, we should also think what are the benefits in terms of tooling so people do not feel forced to use it: they feel attracted to use it :).


-------- Original Message --------

No guillermo. This is really annoying.

With long package list like Collections* then you do not see well the tests.

So we should check the bug like that we are free.

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef




Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Guillermo Polito
As far as I understand, Iceberg uses RPackages. And RPackages do not necessarily follow MC conventions.

-------- Original Message --------
<offtopic>
Wow, I do not know where that email came from O_o.
It looks a spammish email. Maybe it came from my phone?

If it happens again, please tell me so I can take action.
</offtopic>

Now in-topic: I agree with having conventions and with the tests one particularly. But as you say, we should also think what are the benefits in terms of tooling so people do not feel forced to use it: they feel attracted to use it :).


-------- Original Message --------

No guillermo. This is really annoying.

With long package list like Collections* then you do not see well the tests.

So we should check the bug like that we are free.

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef





Reply | Threaded
Open this post in threaded view
|

Re: about package naming

Nicolas Passerini
Yes Icebergs uses RPackages, I have RPackages and Tags and I do not remember having problems.

On Mon, Sep 5, 2016 at 9:50 AM, Guille Polito <[hidden email]> wrote:
As far as I understand, Iceberg uses RPackages. And RPackages do not necessarily follow MC conventions.


-------- Original Message --------
<offtopic>
Wow, I do not know where that email came from O_o.
It looks a spammish email. Maybe it came from my phone?

If it happens again, please tell me so I can take action.
</offtopic>

Now in-topic: I agree with having conventions and with the tests one particularly. But as you say, we should also think what are the benefits in terms of tooling so people do not feel forced to use it: they feel attracted to use it :).


-------- Original Message --------

No guillermo. This is really annoying.

With long package list like Collections* then you do not see well the tests.

So we should check the bug like that we are free.

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef






Reply | Threaded
Open this post in threaded view
|

Re: about package naming

stepharo

Nice. I wonder where the MC problem is actually coming (probably a notification handled the old way).


Le 5/9/16 à 09:58, Nicolas Passerini a écrit :
Yes Icebergs uses RPackages, I have RPackages and Tags and I do not remember having problems.

On Mon, Sep 5, 2016 at 9:50 AM, Guille Polito <[hidden email]> wrote:
As far as I understand, Iceberg uses RPackages. And RPackages do not necessarily follow MC conventions.


-------- Original Message --------
<offtopic>
Wow, I do not know where that email came from O_o.
It looks a spammish email. Maybe it came from my phone?

If it happens again, please tell me so I can take action.
</offtopic>

Now in-topic: I agree with having conventions and with the tests one particularly. But as you say, we should also think what are the benefits in terms of tooling so people do not feel forced to use it: they feel attracted to use it :).


-------- Original Message --------

No guillermo. This is really annoying.

With long package list like Collections* then you do not see well the tests.

So we should check the bug like that we are free.

Je suis pas de place  que no necesita activar gustamucho la 🎶 que ya el otro lado


Le 4 sept. 2016 10:18, "stepharo" <[hidden email]> a écrit :
Hi guys

Sorry to be that dull and stupid but I do not get why (beside following the seaside convention - and I do not understand it either) why the test package for FileSystem-Core is named FileSystem-Tests-Core

and not just

    FileSystem-Core-Tests

    like that we could have

        FileSystem-Core-Help

        FileSystem-Core-Examples

And when we are looking for FileSystem-Core we can see automatically FileSystem-Core-Tests

Now we cannot.

Stef