Squeak-dev October 2002

squeak-dev@lists.squeakfoundation.org

183 participants
406 discussions

[ENH] RBBrowseLocal-nk
by Ned Konz 27 Oct '02

27 Oct '02

from preamble: "Change Set: RBBrowseLocal-nk Date: 27 October 2002 Author: Ned Konz Adds the browse local senders/implementors to the shifted message list menu in the Refactoring Browser. "!

1 0

[ENH] TextColor-nk
by squeak-dev＠lists.squeakfoundation.org 27 Oct '02

27 Oct '02

--==CelesteAttachment52650== from preamble: "Change Set: TextColor-nk Date: 27 October 2002 Author: Ned Konz This little change set lets you set a default text color other than black. If you want to have dark backgrounds and light text, you probably have to turn off the #alternativeWindowLook preference. This doesn't touch menus yet. Change text color like this: Preferences textColor: Color white. "! --==CelesteAttachment52650== Content-type: application/octet-stream Content-transfer-encoding: base64 Content-disposition: attachment;filename="TextColor-nk.cs.gz" H4sIAAAAAAAAAK1UwW7bMAw9++B/YNJDLluWpO26eqciQ4BtXRcsBXooukGx6ViwInkS1az9 +lGyGyRFsB3akyVSfHx8pDmYWbOGxW+Poj4eTsCUMB7DF68eYDIaTeBWCUJH4JuCDxkcnZyf vr8Do2FyBt9zMku07UtBcJ6NPmTjExDrQS/tTyuhVwgLpCxJrvEPTY0y9q2u008BKkmSZxDp hafK2OC5wgK+Gv2YpteVdKAkkULIW0SHBArJwYPx8SKgwFJ4RUCcBvKQBwxVjEscAksl8noI n8sYsRGaHxqoxD1CIWwNS3avrPG6cCB0welWVYv1JkY01izFkiWJIRxK3mqWqmR4hCOhCK0W JO/xRurCbC6NqTkIS7Socxx2VRQGnR6E3D6vYI3acwlI7O6k2mGvZM13jsrSdL5FcvFJ1DGD +IFNJSlk6PfSNO3NlV+tmCpeSkffjG1CHha1cLMQMiis2Ei9GoAjsW7YoGsYj95Nzt7FHo64 gyfcO68rliDKsECFOUmj08TFIxYtriyvDF1JlcHtviNv6R1k/bHrYXHXg5bwzqtcCef2+TbC Cjagdf+iPGbK2xRp8pP5lLANvSAe262beV8sHWqae7bfdirGCYG7YWT1OqSOd0llrSpRwxLC 0M6fQPbIkeleDtNOoAVZbtiBVkotSQolH0Xozv86Gl7fSKqmRhMX7zIQLTKUbOHbjD+A66YS jodue5qaApm1b/hf2mbkeUvaMf11uM3sD7Dsjrh8fcJj0y40e/g8M7kPnlIoF0xRpPw50645 ryhIW0z6kvq0VPvVjQ6W1DUzbMEXMT9l5t2u64a9/wPjNgqb6GkLllIpBnnglVmGoWKXxRx5 P9k+/x6Ha4oU/wIMNREDDwYAAA== --==CelesteAttachment52650==--

1 0

[ENH] TextColor-nk
by Ned Konz 27 Oct '02

27 Oct '02

from preamble: "Change Set: TextColor-nk Date: 27 October 2002 Author: Ned Konz This little change set lets you set a default text color other than black. If you want to have dark backgrounds and light text, you probably have to turn off the #alternativeWindowLook preference. This doesn't touch menus yet. Change text color like this: Preferences textColor: Color white. "!

1 0

Fwd: [Squeak-fr] Appel du SqueakNic 4
by Stephane Ducasse 27 Oct '02

27 Oct '02

> ----------------------------------------------------------------------- > ---------------------------------------------------------------- > > The french Smalltalk community would like to organize the 4th > SqueakNic in Paris. > What is a SqueakNic ? It's something like a mix beetween a picnic and > a coding party around Squeak and Smalltalk. The idea is to show what > you want around Squeak (demonstrations, tools, research projects, ...) > in a very friendly environment around computers (and pizza ?). We > invite all Squeakers and other developpers from all Smalltalk flavors > to join this event to share their enthusiasm and knowledge. > > Contact : Serge dot Stinckwich at info dot unicaen dot fr > > When : Monday 11th November from 10PM. > Where : ? > Phone number : 0624164645 > > Please add your name in the FSUG (French Smalltalk User Group) Wiki > web site if you want to join the party : > http://www.iutc3.unicaen.fr:8000/fsug/ > > Previous editions of SqueakNic was held in Berne (Switzerland) and > Karlsruhe (Germany). > > ----------------------------------------------------------------------- > -------------------------- > > La communauté Smalltalk françophone souhaite organiser le 4ième > SqueakNic à Paris > Qu'est ce qu'un SqueakNic ? C'est quelque chose comme un mélange entre > un picnique et une coding party autour de Squeak et de Smalltalk. > L'idée est de montrer ce que vous souhaitez à propos de Squeak > (démonstrations, outils, projets de recherche, ...) dans une ambiance > amicale avec des ordinateurs (et des pizzas ?). Nous invitons tous les > fanas de Squeak ainsi que les autres développeurs d'autres Smalltalk > de se joindre à l'événement afin de partager leurs enthousiasmes et > leurs connaissances. > > Contact : Serge point Stinckwich chez info point unicaen point fr > > Quand : le lundi 11 novembre à partir de 10h. > Ou : ? > Numéro de téléphone : 0624164645 > > Merci d'ajouter votre nom sur le wiki du FSUG (French Smalltalk User > Group) si vous voulez nous joindre à nous : > http://www.iutc3.unicaen.fr:8000/fsug/ > > Les éditions précédentes des SqueakNic se sont déroulés à Berne > (Suisse) et Karlsruhe (Allemagne). > > > -- > Serge Stinckwich -< ) > multiagent.com > Université de Caen>CNRS UMR 6072>GREYC>MAD /~\ > squeak.org > http://www.iutc3.unicaen.fr/serge/ (/ | zope.org > Smalltalkers do: [:it | All with: Class, (And love: it)] _|_/ > debian.org > Dr. Stéphane DUCASSE (ducasse(a)iam.unibe.ch) http://www.iam.unibe.ch/~ducasse/ "if you knew today was your last day on earth, what would you do different? ... especially if, by doing something different, today might not be your last day on earth" Calvin&Hobbes

2 1

About BC and macrobenchmark
by Stephane Ducasse 27 Oct '02

27 Oct '02

Hi Anthony This is great to see all the work you did. Have you compared the 2 VMs with the macroBenchmarks because they usually give more reasonable and concrete numbers? Stef

1 0

BC + call/cc
by squeak-dev＠lists.squeakfoundation.org 27 Oct '02

27 Oct '02

--==CelesteAttachment62531== Avi Bryant <avi(a)beta4.com> wrote: > Still tring to implement #callCC in the block closure image. I get a DNU > #trimBottomTo: in Process2>>swapSender:. Anthony, what am I doing wrong? > How does #swapSender: ever work? > I have to say, I think the BC work is great, but I much prefer working > with the non-BC MethodContexts... Hi Avi and all, swapSender: only works on suspended processes. In fact, all context manipulations only work on suspended processes. This is to allow running process to be internalized by the VM. It is pretty easy to temporarily suspend a running process and manipulate it by using forkAndWait, ie: proc _ thisContext process. [ proc suspend. "manipulate proc here" proc resume. ] forkAndWait. forkAndWait will cause the active process to wait until the forked block finishes. It still needs to be suspended, however, to convert it to object form for manipulation. Upon resume it resumes waiting for the fork to complete. Then the fork completes and the proc resumes running. Attached is a Continuation changeset for VI4. It works, accept if you write to the Transcript it crashes the VM. I will try to fix this tomorrow. Cheers, Anthony --==CelesteAttachment62531== Content-type: application/octet-stream Content-transfer-encoding: base64 Content-disposition: attachment;filename="Continuations-ajh.cs" J0Zyb20gU3F1ZWFrMy4yW1ZJNF0gb2YgMTEgSnVseSAyMDAyIFtsYXRlc3QgdXBkYXRlOiAj NDk1Nl0gb24gMjcgT2N0b2JlciAyMDAyIGF0IDU6NTQ6MzcgYW0nIQ1PYmplY3Qgc3ViY2xh c3M6ICNDb250aW51YXRpb24NCWluc3RhbmNlVmFyaWFibGVOYW1lczogJ3N1c3BlbmRlZFBy b2Nlc3MgJw0JY2xhc3NWYXJpYWJsZU5hbWVzOiAnJw0JcG9vbERpY3Rpb25hcmllczogJycN CWNhdGVnb3J5OiAnS2VybmVsLUV4ZWN1dGlvbichDQ0hQ29udGludWF0aW9uIGNvbW1lbnRT dGFtcDogJ2FqaCAxMC8yNy8yMDAyIDA1OjUwJyBwcmlvcjogMCENSSBjb250YWlucyBhIHN1 c3BlbmRlZCBwcm9jZXNzIHRoYXQgd2lsbCBnZXQgY29waWVkIGFuZCByZXN1bWVkIHVwb24g ZXZhbHVhdGlvbiAoYWxhIFNjaGVtZSkuDQ1UaGUgZm9sbG93aW5nIGV4YW1wbGUgd2lsbCBl eGVjdXRlICdoZWxsbycgaW5zcGVjdCB0d2ljZSAodGhlIHdob2xlIHRoaW5nIGlzIGluIGEg Zm9yayBzbyB0aGUgY29udGludWF0aW9uIGRvZXMgbm90IHJlc3VtZSBhIHNlY29uZCBVSSBs b29wKToNDVsJfCByZXN1bWVWYWx1ZSB8DQlyZXN1bWVWYWx1ZSA6PSBbOmNvbnRpbnVhdGlv biB8IGNvbnRpbnVhdGlvbl0gY2FsbENDLg0JJ2hlbGxvJyBpbnNwZWN0Lg0JcmVzdW1lVmFs dWUgaWZOb3ROaWw6IFsNCQkicmVzdW1lVmFsdWUgZXF1YWxzIHRoZSBjb250aW51YXRpb24g dGhlIGZpcnN0IHRpbWUuDQkJIE5vdyByZXN1bWUgdGhlIGNvbnRpbnVhdGlvbiB3aXRoIGEg cmVzdW1lIHZhbHVlIG9mIG5pbCINCQlyZXN1bWVWYWx1ZSB2YWx1ZTogbmlsXS4NXSBmb3Jr Lg0NVGhhbmtzIHRvIEF2aSBCcnlhbnQgZm9yIGV4cGxhaW5pbmcgY29udGludWF0aW9ucyBh bmQgcHJvdmlkaW5nIHRoZSBhYm92ZSBleGFtcGxlLg0hDV1zdHlsZVsoMTMwIDE1IDQ0Milm MSxmMWksZjEhDQ0NIUJsb2NrQ2xvc3VyZSBtZXRob2RzRm9yOiAnc2NoZWR1bGluZycgc3Rh bXA6ICdhamggMTAvMjcvMjAwMiAwNTo1NCchDWNhbGxDQw0JIkNhbGwgd2l0aCBjdXJyZW50 IGNvbnRpbnVhdGlvbiwgYWxhIFNjaGVtZS4NCUV2YWx1YXRlIHNlbGYgYWdhaW5zdCBhIGNv cHkgb2YgdGhlIHNlbmRlcidzIGNhbGwgc3RhY2ssIHdoaWNoIGNhbiBiZSByZXN1bWVkIGxh dGVyIg0NCV4gc2VsZiB2YWx1ZTogdGhpc0NvbnRleHQgc2VuZGVyIGNvcHlTdGFjayBhc0Nv bnRpbnVhdGlvbiEgIQ0NDSFDb250aW51YXRpb24gbWV0aG9kc0ZvcjogJ2FzIHlldCB1bmNs YXNzaWZpZWQnIHN0YW1wOiAnYWpoIDEwLzI3LzIwMDIgMDQ6NDMnIQ1zdXNwZW5kZWRQcm9j ZXNzOiBhU3VzcGVuZGVkUHJvY2Vzcw0NCXN1c3BlbmRlZFByb2Nlc3MgXyBhU3VzcGVuZGVk UHJvY2VzcyEgIQ0NIUNvbnRpbnVhdGlvbiBtZXRob2RzRm9yOiAnYXMgeWV0IHVuY2xhc3Np ZmllZCcgc3RhbXA6ICdhamggMTAvMjcvMjAwMiAwNTozMSchDXZhbHVlOiByZXN1bWVWYWx1 ZQ0JIkNvcHkgYW5kIHJlc3VtZSBteSBzdXNwZW5kZWRQcm9jZXNzLiAgcmVzdW1lVmFsdWUg d2lsbCBiZSB0aGUgcmV0dXJuIHZhbHVlIG9mIGNhbGxDQyINDQlzdXNwZW5kZWRQcm9jZXNz IGNvcHkNCQlzdGFja1JlcGxhY2VUb3A6IHJlc3VtZVZhbHVlOw0JCXJlc3VtZSEgIQ0NDSFN ZXRob2RDb250ZXh0MiBtZXRob2RzRm9yOiAnc2VuZGVyJyBzdGFtcDogJ2FqaCAxMC8yNy8y MDAyIDA0OjM1JyENY29udGV4dFdpdGhUYWc6IGFDb250ZXh0VGFnDQkiU2VhcmNoIG15IHNl bmRlcnMgZm9yIHRoZSBjb250ZXh0IHdpdGggYUNvbnRleHRUYWciDQ0JXiBzZWxmIGNvbnRl eHRXaXRoVGFnOiBhQ29udGV4dFRhZyBpZk5vbmU6IFtzZWxmIGVycm9yTm9Db250ZXh0V2l0 aFRhZzogYUNvbnRleHRUYWddISAhDQ0hTWV0aG9kQ29udGV4dDIgbWV0aG9kc0ZvcjogJ2Nv cHlpbmcnIHN0YW1wOiAnYWpoIDEwLzI3LzIwMDIgMDQ6MzMnIQ1jb3B5DQ0Jc2VsZiBlcnJv cjogJ0luZGl2aWR1YWwgY29udGV4dHMgc2hvdWxkIG5vdCBiZSBjb3BpZWQsIGNvcHkgc3Vz cGVuZGVkIHByb2Nlc3NlcyBpbnN0ZWFkJyEgIQ0NIU1ldGhvZENvbnRleHQyIG1ldGhvZHNG b3I6ICdjb3B5aW5nJyBzdGFtcDogJ2FqaCAxMC8yNy8yMDAyIDA0OjM4JyENY29weVN0YWNr DQkiQ29weSBteSBwcm9jZXNzIHRydW5jYXRlZCBkb3duIHRvIHRoaXMgY29udGV4dCINDQl8 IHRhZyBwcm9jIGN0eHQgfA0JdGFnIF8gc2VsZiBjb250ZXh0VGFnLiAgIkNyZWF0ZSBpdCBi ZWZvcmUgY29weWluZyINCXByb2MgXyB0aHJlYWQgY29weS4NCWN0eHQgXyBwcm9jIGNvbnRl eHRXaXRoVGFnOiB0YWcuDQlbY3R4dCBpc1RvcF0gd2hpbGVGYWxzZTogW3Byb2MgcG9wVG9w Q29udGV4dF0uDQleIHByb2MhICENDQ0hUHJvY2VzczIgbWV0aG9kc0ZvcjogJ2NvbnRleHRz JyBzdGFtcDogJ2FqaCAxMC8yNy8yMDAyIDA0OjM1JyENY29udGV4dFdpdGhUYWc6IHRhZw0N CV4gbGFzdEFjdGl2ZUNvbnRleHQgY29udGV4dFdpdGhUYWc6IHRhZyEgIQ0NIVByb2Nlc3My IG1ldGhvZHNGb3I6ICdwcml2YXRlJyBzdGFtcDogJ2FqaCAxMC8yNy8yMDAyIDA0OjE4JyEN Y29weQ0NCXwgY29weSB8DQlbCXNlbGYgc3VzcGVuZC4NCQljb3B5IF8gc2VsZiBjb3B5Lg0J CXNlbGYgcmVzdW1lLg0JXSBmb3JrQW5kV2FpdC4NCV4gY29weSEgIQ0NDSFTdXNwZW5kZWRQ cm9jZXNzIG1ldGhvZHNGb3I6ICdjb3B5aW5nJyBzdGFtcDogJ2FqaCAxMC8yNy8yMDAyIDA0 OjQyJyENYXNDb250aW51YXRpb24NDQleIENvbnRpbnVhdGlvbiBuZXcgc3VzcGVuZGVkUHJv Y2Vzczogc2VsZiEgIQ0NIVN1c3BlbmRlZFByb2Nlc3MgbWV0aG9kc0ZvcjogJ2NvcHlpbmcn IHN0YW1wOiAnYWpoIDEwLzI3LzIwMDIgMDQ6MzEnIQ1jb3B5DQ0JXiBzZWxmIHNoYWxsb3dD b3B5IHBvc3RDb3B5ISAhDQ1Qcm9jZXNzMiByZW1vdmVTZWxlY3RvcjogI3Bvc3RDb3B5IQ1N ZXRob2RDb250ZXh0MiByZW1vdmVTZWxlY3RvcjogI3Bvc3RDb3B5IQ0= --==CelesteAttachment62531==--

1 0

Re: BC + call/cc
by Anthony Hannan 27 Oct '02

27 Oct '02

Avi Bryant <avi(a)beta4.com> wrote: > Still tring to implement #callCC in the block closure image. I get a DNU > #trimBottomTo: in Process2>>swapSender:. Anthony, what am I doing wrong? > How does #swapSender: ever work? > I have to say, I think the BC work is great, but I much prefer working > with the non-BC MethodContexts... Hi Avi and all, swapSender: only works on suspended processes. In fact, all context manipulations only work on suspended processes. This is to allow running process to be internalized by the VM. It is pretty easy to temporarily suspend a running process and manipulate it by using forkAndWait, ie: proc _ thisContext process. [ proc suspend. "manipulate proc here" proc resume. ] forkAndWait. forkAndWait will cause the active process to wait until the forked block finishes. It still needs to be suspended, however, to convert it to object form for manipulation. Upon resume it resumes waiting for the fork to complete. Then the fork completes and the proc resumes running. Attached is a Continuation changeset for VI4. It works, accept if you write to the Transcript it crashes the VM. I will try to fix this tomorrow. Cheers, Anthony

1 0

Re: [ANN] DVS rev. 1.34
by danielv＠netvision.net.il 27 Oct '02

27 Oct '02

[A pluggable architecture for Installers] This could help us get the most out of the information SqueakMap makes available. Well, first a summary of where we are and what functionality we want to get. Right now we have several responsibilities bunched up together, we might want to separate, that are all handled by the installers - - Getting something from the net and saving it locally - Decompressing it - Unpacking (for sar files) - Installing And we want to add more logical operations on the code delivered, such as - - Upgrade/uninstall/reinstall - 'browse code' (regular or maybe via DVS) And I think we will also want to enable more operations on the package itself (not just the downloaded code), like: - Upload my changes to this package as a new version and update the SM entry, asking me for the required details. (I'm the maintainer for the code and the SM card) - Post my changes to this package to the mailing list, as a suggested patch (I'm a user and contributor). I'm assuming this covers more or less what we'll need soon. Any other ideas? Now, a design proposal (basically, the dynamicOpenMenu design reapplied)- If we want to have all these services available in the UI, we need to have proper registration of services that want to handle packages, including how they should be displayed in a menu. Then, when generating the menu for a select package, we do a (services select: [:service | service canHandle: aSMCard]) do: [:service | aMenu add: (service menuSpec)]. Services register when their class is installed, so people can freely add SM packages that implement new package handling services, including handling new card types, such as .pr files. How about it? Daniel

3 2

Re: [ANN] DVS rev. 1.34
by danielv＠netvision.net.il 27 Oct '02

27 Oct '02

Avi Bryant <avi(a)beta4.com> wrote: > > On Sat, 26 Oct 2002 danielv(a)netvision.net.il wrote: > > > Filing in via DVS will also remove methods? > > Yes. The whole point is that it brings the package in the image in sync > with whatever's in the file out - by adding, changing, *or removing* > methods to do so. Cool. That makes it quite a bit better than what we have now, in terms of filing things in, whether for comparing version or for updating packages, because it should prevent messes like we had when SqueakMap had some an implementation of #unpack removed, but people updating got both versions. > > if so, I need to start using > > it for my packages. Are you going to create an SMDVSInstaller? > > Yes, I probably should. I wrote a message about refactoring SM, saying I though installers are just a subset of code handlers that should be usable on the information provided by SqueakMap, and that the collection of handler should be pluggable, so that new handlers should show up in the tools easily. This would allow, for example, to plug in variants that update using DVS, and variants that warn you when updating a package will revert modifications made by yourself to a package. What do you think? > > Could you elaborate on this? > > Right. The basic idea is: [snip] > That help? Yes, alot, thanks. > Avi

3 3

[ANN] Seaside 2.0
by Avi Bryant 27 Oct '02

27 Oct '02

Hi all, I'd like to announce an initial release of the Seaside 2.0 web application framework. This is a complete reimplementation of Seaside, with a highly layered design and simpler, more "Smalltalk-like" semantics than the original. A changeset is available from http://beta4.com/seaside2 , as is a tutorial (for those who want to get a general sense of how it works) and some low-level design documents (for those who have an interest in how it's implemented). This is a preview release - Seaside 2.0 is still a work in progress. What this release includes: - Linear control flow. Rather than providing handlers at the individual request level, Seaside models an entire session as a continuous piece of code. To put it another way, Seaside web applications block on user input, allowing a much more natural and powerful development style. - Callback-based event system. Seaside provides an HTML generator that associates blocks, not urls, with links, buttons, and form inputs. - Solid component model. Implement stateful widgets and pages as reusable and embeddable component classes. - Backtracking safety. All of the above are designed to work consistently and reliably in the face of the browser back button. A transaction system allows the developer to selectively limit the scope of backtracking. - In-browser development tools. Seaside provides an integrated web-based inspector, system browser, profiler, and notifier. What this release does not yet include: - A fully fledged View library. In time, I expect to be able to provide a large number of widgets and layout managers to facilitate programmatic UI design. - A templating system. Earlier versions of Seaside had a very sophisticated template system, which would be nice to adapt and improve. - Good ways of dealing with static content, including images, .css files, and non-generated HTML. All of these features are missing largely because I believe they will be best explored through community feedback. After people get a chance to play with the base framework, I'd love to hear some discussion on how the above could work. Cheers, Avi

6 11

← Newer
1
...
7
8
9
10
11
12
13
...
41
Older →

Jump to page: