Re: [Pharo-project] How to define a ConfigurationOf with a #development version that depends on a baseline?
Alexandre, I had the same feeling, when I had to use it starting around 1997 in our company :-) We are now C#'pers, but it is interesting to see, that one item is remembered by all persons in our team from the Smalltalk-epoche in our company in a friendly way: ENVY. This was the item, which seem to impressed most of the people - especially *after* switching to VS. Marten On 07.04.2013 04:48, Alexandre Bergel wrote: Hi Marten, Frankly speaking, I find Envy way too complicated. I am using it these days and I am really not happy with. If I have to read a book to understand how it works, than there is no way I can teach students on how to use it in a reasonable amount of time. Configuration Map, Application, edition, scratch version and many more are way too many concepts, in my opinion. attachment: marten.vcf
Re: [Pharo-project] How to define a ConfigurationOf with a #development version that depends on a baseline?
In all the discussions about how to manage/organize Smalltalk code - I would like to add, that ENVY (though it's so old) with its config maps, applications, prerequisites is a pretty good example how it works. Take a look at the only available ENVY implementation today (VAST) and think about the way they did it in the eighties ... The ENVY-(concept - not the implementation) is still a master peace in this area - even when compared with VS or Eclipse - and it supports the power of Smalltalk very much. There are questionable parts of ENVY and I do not use these parts of it very often. But even these questionable parts of ENVY have not been better solved by other Smalltalk source code management systems. I've repositories, which are around 10 years old - and ALL code is still there, all versions for all different platforms - down to methods differences within a mouse click and the speed within a LAN is so good. Marten On Apr 5, 2013, at 11:28 PM, Alexandre Bergelalexandre.ber...@me.com wrote: Just a question: Is there any plan for a Metacello 2? attachment: marten.vcf
Re: [Pharo-project] Deploying without .changes and .sources files
The good thing about stackoverflow is, that its an open qa system for various languages and a system mentioned there means additional advertisement. I sometimes find myselfs there due to a question and suddenly was reading answer for different systems. Marten On 07.02.2013 10:08, Marcus Denker wrote: On Feb 7, 2013, at 10:00 AM, Damien Cassoudamien.cas...@gmail.com wrote: Hi, Norbert asked about the possibility to deploy a pharo image without .changes and .sources files http://stackoverflow.com/questions/14737695/is-it-possible-to-deploy-a-pharo-image-without-changes-and-sources-files Could people participate to the discussion there? I think it is important. Stackoverflow is just one channel to much for me… I need to controll how this or I will do nothing but just read and filter. Marcus attachment: marten.vcf
Re: [Pharo-project] Fwd: Lua scripting with nativeboost?
And this is the biggest reason to use Lua as a scripting language. Marten On 12.12.2012 10:47, Igor Stasenko wrote: i think the easiest solution is to run a separate image. if you want to run in same image, then you have to modify compiler/environment for sandboxed code, where nothing can sneak through and access private data. attachment: marten.vcf
Re: [Pharo-project] Fwd: Lua scripting with nativeboost?
On 12.12.2012 17:18, Esteban Lorenzano wrote: IMO, that's non-sense. Well, that is a matter of view. I've done this also with Lua and VASmalltalk and I did this not because of other Smalltalkers in mind using my software, but perhaps of potential end users willing to enhance the application in a more known language. Because of this I also tried to use Rexx as an extension language, which is in my view even more suitable to be used as an embedded language - but the market decided to use Lua and not Rexx. Marten attachment: marten.vcf
Re: [Pharo-project] I don't agree with DummyXXX
Nicht um 1600 Uhr ? Wolltest Du nicht noch Unterlagen rumschicken ? marten On 21.11.2012 16:47, Fernando Olivero wrote: I totally agree. Dummy does not tell me about the class. Are there just 3 uses of Dummy in the core? We should either use Null or use a specific prefix always, so its form a consisten pattern. DummyUIManager - NullUIManager Fernando On Wed, Nov 21, 2012 at 4:34 PM, Mariano Martinez Peck marianop...@gmail.com mailto:marianop...@gmail.com wrote: From what I understand, Dummy means something different. I have using that for class names. More accurate names can be found. I would rename: DummyUIManager - HeadlessUIManager DummySystemProgressItem - NullSystemProgressItem DummySoundSystem - NullSoundSystem what do you think? do you have better names? -- Mariano http://marianopeck.wordpress.com attachment: marten.vcf
Re: [Pharo-project] A trend and an unfair comparison about js everywhere
With HTML and the upcoming WebSockets standard we will see a shift in web programming. Web applications are getting true client server systems in a much more difficult environment than a traditional (high-speed, low latency) company IT world. Than on the other hand - and we've seen examples in the smalltalk world already - solution will be available to produce desktop apps (using the same technology as mentioned above - but in an all-in-one-server app) and Internet based web applications. The base key technologies in this area will be how to handle the client approach and what the part of Smalltalk can be here and a way to do data exchange between internet database and local database. The server part ? Well integrate the WebSocket technology by the vendors and free Smalltalk projects and you have Smalltalk there. Marten attachment: marten.vcf
Re: [Pharo-project] [NB] Speaking about callbacks
Me too. Marten Am 16.09.2012 20:42, schrieb Stéphane Ducasse: for what it is worth I like the 3rd form. Now I'm not sure on: is the right selector. attachment: marten.vcf
[Pharo-project] SqueakSource3: ByteString(Object)doesNotUnderstand: #goferReferences
When I copy the Gofer source (listed on the project pages) to a Pharo 1.4 and tries to execute stuff like: Gofer new repository: 'http://ss3.gemstone.com/ss/ICU'; package: 'ConfigurationOfICU'; package: 'ICU-Core'; package: 'ICU-NativeBoost-Core'; package: 'ICU-Tests-Core'; package: 'ICU-Tests-NativeBoost'; load I get an error message: ByteString(Object)doesNotUnderstand: #goferReferences Marten attachment: marten.vcf
Re: [Pharo-project] Fwd: Working with Pharo 1.4/2.0 …
Just some additional remarks about the setting framework and how VA has it since its latest release. Due to its source code management (ENVY) each application has its own configuration section in a simple ascii based *.ini file. Therefore each application has a text section like [name of application] key = value ... The developer just write 3 methods on the class side of its application class, which more or less do something: one method defines all possible entries expected in the text files and what type it has one method defines current AND default values for each setting one method defines how to transfer the values to your application. That's pretty simple - but also limited - and you copy the implementation of these three methods from another application, change them according to your needs and that's it. The documentation describing the Pharo settings framework is 20 pages long and offers many more features - I actually do not need. And I do not need a GUI to edit my setting values - but for others this might be useful. The documentation is not straight forward - three or four pages with introductions. And even now I do not know how the settings are stored outside the system. But as I said before - I have not done much work with Squeak/Pharo yet - and this only reflects my early impressions when working with this system. And I even do not know, if Settings is part of Pharo 1.4 or Pharo 2.0. Just to give you an overview of my ideas here. and another point I want to say about using the system: Working with VASmalltalk/ENVY the methods called after starting, resuming, loading, unloading (and various other events) to each class/application are well documented. I do not find documents/informations about this in Pharo/Squeak and this is a very important information when initializing an application. Marten Feldtmann Am 06.09.2012 17:30, schrieb Sean P. DeNigris: * Settings – I find this one *really* interesting. Will you highlight the key differences to VA? Maybe we can improve the framework in Pharo, or at minimum create a UI that makes things simpler and more straightforward for you attachment: marten.vcf
Re: [Pharo-project] Fwd: Working with Pharo 1.4/2.0 …
Another hint here: I like the refactoring browser's way: multi-select items in the list of attributes and default accessors are created - no need to edit the code here. I am in the browser, already looking at this class - and then I change the code in the browser. Marten Am 06.09.2012 17:30, schrieb Sean P. DeNigris: * creating accessor methods for instance attributes is not speedy and takes too much clicks attachment: marten.vcf
Re: [Pharo-project] [NativeBoost] Question: zero-based or one-based indexes?
Igor, for me it does not matter. I've created my own interface package for all base types, array access - and all zero-index based - either for Smalltalk-memory or external memory access. That's good enough for me and my projects. Now I'm waiting for your base package with all these new accessor methods. Marten attachment: marten.vcf
Re: [Pharo-project] [NativeBoost] Question: zero-based or one-based indexes?
Hello Igor,
0-based - because it's oriented towards C structures (and therefore
memory oriented).
Another way to see this is: it's not an index, its an byte-offset from
the beginning of the memory. Therefore I would not name the parameter as
something like index, but more like zeroBasedOffset. By the way:
VASmalltalk does it the same way.
The same problem might occur, when dealing with pointers to array of
primitive data types in ByteArray or instances of NBExternalAddress.
Even here I would suggest using zero based index structures - though it
is contrary to Smalltalk typical 0ne-based array index. In VASmalltalk
this situation is zero-based access and the parameter is not a byte
index, but a data element index. And an index of 0 returns the start
of the array (and the first element).
Marten
Am 03.09.2012 02:28, schrieb Igor Stasenko:
Hi, there
i added a bunch of accessors to ByteArray and NBExternalAddress which
having uniform look:
nbtypeAt:
nbtypeAt: put:
where types are:
{U}Int8/16/32/64
or
Float32/64
the problem is that indexes in ByteArray are 1-based,
but in NBExternalAddress is 0-based.
And there's already different places in code which following these rules.
But i feel like this should be uniform (so same piece of code can be
used for either bytearrays or external addresses),
to avoid confusion and mistakes.
i just not sure which one to leave: 1-based or 0-based ?
From one side, all collections in smalltalk is 1-based..
but from other side accessing memory at 1-based offset looks also unnatural.
What you think?
attachment: marten.vcf
Re: [Pharo-project] [NativeBoost] 1.4 Bleeeding edge release
Hello Igor, I think, that unifom indexing is a must. The usage/handling of NBExternalAddress must be equal to instances of ByteArray: at least as long as we have synchronous callouts, where a garbage collector is not able to move the byte array around in Smalltalk memory during the callout. Marten attachment: marten.vcf
