Squeak-dev December 2002

squeak-dev@lists.squeakfoundation.org

173 participants
456 discussions

Apparent bug in Squeak's exception handling
by Chris Muller 30 Dec '02

30 Dec '02

In a stock 3.4 image, print this in a workspace: [ InMidstOfFileinNotification signal ] on: Notification do: [ :noti | noti resume ] This expression should return false, but it returns nil. InMidstOfFileinNotification is an Exception included with Squeak that has a defaultAction to return false. But when wrapped in a Notification handler, as above, it seems to not return the correct value. Is this a bug? Any assistance would be appreciated. - Chris __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com

1 0

LGPL and SqueakMap
by Nevin Pratt 30 Dec '02

30 Dec '02

I very strongly disagree with the statement on SqueakMap that reads: > Licenses/LGPL - GNU Lesser General Public License. Only suitable for > plugins. > I believe this statement was derived from a post by Andrew C. Greenberg, wherein he stated: > "I am an IP and Patent lawyer. After a careful analysis, and > extensive discussions with FSF and RMS on the subject matter, it is my > present view that GPL is unsuitable for use with a monolithic object > image system unless all code of that system is to be GPL'd. While > LGPL can be made to work for things like plugins, GPL simply is "too > viral" to work in an open community such as ours." Notice that Andrew's statement was primarily about GPL code, with only a somewhat passing comment about LGPL. As to the use of LGPL for external libraries (like those found on SqueakMap), Andrew later said: > "LGPL, as opposed to GPL, is non-viral to calling programs. It is > usable for plugins and attached libraries, particularly if > dual-licensed with Squeak-L. I wouldn't use it to distribute > Smalltalk code...I have not given any thought to LGPL Smalltalk code > distributed outside the image. Let me consider this." The key phrase here is Andrew's admission that "I have not given any thought to LGPL Smalltalk code distributed outside the image". GLORP is licensed as LGPL. GLORP began life as an Object People project, and for various political reasons will probably remain LGPL. However, I can categorically state that many within Cincom have plans to eventually "officially" support GLORP as the preferred O/R mapping framework for VisualWorks. In other words, there are plans for GLORP to see an active life as a supported part of a major commercial Smalltalk implementation. And this, in and of itself, goes directly against the SqueakMap statement that LGPL is "only suitable for plugins". GLORP also has an attorney retained to investigate such issues. At some point an "official" opinion might be solicited from that attorney (with, of course, his usual fees to investigate the issue sufficiently to render an "official" opinion). However, for now, it isn't considered to be a high enough priority. Thus, for now, a simpler approach to quelch all uncertainties might be to get an "official" opinion rendered directly by the FSF concerning LGPL (just as it appears Andrew did for GPL), since they are the creators of the LGPL license. At some point (hopefully soon) I plan on attempting to do just that. But in any case, it certainly *is* the current opinion of many that LGPL is certainly good for more than "just VM plugins". I am one of those people. I strongly disagree with the statement on SqueakMap that states it that way (though I personally completely agree with Andrew's statement about GPL code). Nevin (This message has been sent to the Squeak list, and CC'd to Andrew C. Greenberg)

8 24

Unable to save a BookMorph
by Joseph Frippiat 30 Dec '02

30 Dec '02

Hi, I created a BookMorph and I want to save it to disk with "send all pages to server". It doesn't work. I use Squeak3.4a-5125 on Windows 98. Problem 1: I got a debugger with the message "Error: Server name not found". I change the method "ServerDirectory class>>defaultStemUrl", I replace 'DaniOnJumbo' with 'Bobs SuperSwiki'. Then I got the Problem 2 Problem 2: I got to the window asking the "url of the place to store a typical page in this book...". I introduce "file://E:/squeak/3.4alpha/myProjects/myPID/book1.sp" . I got a debugger with the message "Error: Fully qualified path expected".

1 0

monticello remote repositories
by Derek Brans 29 Dec '02

29 Dec '02

Is the current version of Monticello able to do remote repositories? If so, how do you set it up (I have everything installed xml-rpc, comanche, etc but I don't know how to run Monticello as a server). thanks. Nerd on a Wire: Web and Information Solutions Website Design - Database Systems - Site Hosting 604.874.6463 mailto:info@nerdonawire.com For more information, visit http://nerdonawire.com

2 1

Re: Workspace menu cancel
by Torge.Husfeldt＠gmx.de 29 Dec '02

29 Dec '02

Hi, To make it even worse: you can also do 'accept' on a workspace -- with no other effect than that the 'modified' marker will vanish and that any further 'cancel's will just be interpreted as no-ops. So the quick fix would be to do an accept on every workspace open. The slighlty cleaner fix would be to remove _both_ entries from the menu so people don't get irritated any more. The right thing to do would certainly be to add a 'lastAcceptedText' instVar to Workspace so that accept updates this text, cancel reverts. More to the point: Do not expect the menu choice 'cancel' to do nothing. It is meant to cancel some operation that is in progress. In the case of a workspace that would be the editing operation and so it _is_ reasonable behavior to revert to the empty text when no 'accept' has taken place since the workspace was opened. (note i said reasonable, not desirable). If you're in the process of a method accept and are given the menu choice to 'cancel' this will mean you back out of the accept but the edits are not reverted. I think users should be able to deal with these 'modes' (and have been since the dawn of squeak). I've attached my accepting workspace fwiw. HTH, Torge Markus Fritsche <Fritsche.Markus(a)gmx.net> wrote: > Hi! > > Is it right, that (windows-) right-click (should I buy nail polish for > my mouse buttons?) on a workspace and select "cancel" (when I want to > cancel the menu - I got used to cancel unwanted menus with "cancel", > call me windows infected if you like), the workspace is cleared. No undo > (cmd-z) is possible. > > Is this the expected behaviour? > > Regards, Markus > > -- > http://reauktion.de/archer/

1 0

Re: Pocket PC Performance
by Torge.Husfeldt＠gmx.de 29 Dec '02

29 Dec '02

Hi All, Can someone who encounters the performance problems mentioned in this thread please try out the following code snippets and report on the outcome?! First try in a workspace: | lookupLengths | lookupLengths _ SortedCollection new. Behavior allSubInstancesDo:[:class | | md | md _ class methodDict. lookupLengths addAll:(md keys asSortedCollection collect:[:sel | (((md scanFor: sel) - sel identityHash) \\ md basicSize) -> (class -> sel)]) ]. lookupLengths asBag sortedCounts inspect. lookupLenghts last:100 inspect. This will give you two inspectors. The first will show the sorted counts of a bag which entries should be interpreted the following: #occurences -> #lookupLength -> sampleClass -> sampleSelector Please report on the differences between a slow image and an acceptable image (preferrably on the same system) The second will give you the details of the 100 Methods with the highest lookupLenghts. Please look swiftly over this list if you can detect any Morphic specific selectors with long lookup lenghts. The second thing i want you to try is to grow all your MethodDictionaries that have exessive lookupLenghts. The following code snippet will do this for you. | lookupLengths | Behavior allSubInstancesDo:[:class | | md | md _ class methodDict. md isEmpty ifFalse:[ lookupLengths _ SortedCollection new. lookupLengths addAll:(md keys asSortedCollection collect:[:sel | (((md scanFor: sel) - sel identityHash) \\ md basicSize) -> (class -> sel)]). (lookupLengths last key > 9) ifTrue:[md grow]]] Please report if your image "feels" any swifter after this operation. Note#0: Be sure not to have any PackagePaneBrowser (aka 5-pane browser) open when you do your tests because these beasts will stop all morphic updating (and maybe event dispatch) for up to one second every second on a slow machine. This is due to a design bug which can very easily be avoided using a changeset i once posted to the list but don't have the patience to dig up right now. Note#1: These operations might take a _very_ long time (especially on a slow system) so be pationent) (on my 1700+ it was in the second range but since you're especially encountering problems on slow systems you will probably do the tests there, too -- so don't say i didn't warn you ). Note#2: LookupLenghts stand for the amount of probes the vm has to do in _a single method dictionary_ to find a method corresponding to a selector. This is just a minimum measure because it doesn't count the number of probes spent while following the superclass chain. These numbers are typically small for almost empty method dictionaries but may become huge when all superclasses have long probe chains and the selector is only implemented in ProtoObject Note#3: It is nowhere near guranteed that this will change anything because lookup lenghts aren't _supposed_ to make a difference. It is widely believed that the vm's lookup cache mechanism should deal with the performance hit that would result from long probe chains. I have, whatsoever, two strong hints that lookupLenghts _might_ be part of the problem you encounter. These are: Hint#1: The problem has arisen rather gradually and noone has yet been able to find any particular change that made the difference hint#2: The Lookup cache (as i understand it) seems to be rather small for as big a system as morphic (i only saw space for 512 entries last time i looked) and gets flushed on several accasions (such as gc's). Looking forward to your feedback, Torge

1 0

[ANN] [CFP] Conference on Creating, Connecting and Collaborating through Computing (C5)
by Shin'ichi Konomi 29 Dec '02

29 Dec '02

(My apology if you received this posting multiple times) C5 -- Conference on Creating, Connecting and Collaborating through Computing This event will be held in Kyoto, Japan next month. The conference's topics include Squeak and Learning. Latest information can be obtained from the following URL: http://www.db.soc.i.kyoto-u.ac.jp/alankay/e-home.htm -- Shin'ichi Konomi Kyoto Software Applications, Inc. and Kyoto University ----------------------------------------------------- <http://www.db.soc.i.kyoto-u.ac.jp/alankay/e-home.html> <http://www.db. soc.i.kyoto-u.ac.jp/alankay/new_page1.html> Conference on Creating, Connecting and Collaborating through Computing (C5) Date: Friday, January 31 Place: Heian Jingu <http://www.heianjingu.or.jp/index_e.html> Shrine Heian Jingu Shrine is well-known as a sightseeing spot -- the conference will be held at Heian Jingu Kaikan which is a part of the shrine. Participation Fee (including a ticket for the banquet): Free (Speakers) 5000 JPY (General Attendees) Call for Papers Abstract: January 5, 2003 (2 pages; IEEE proceedings <http://www.computer.org/cspress/instruct.htm> format) Please send a PDF file with your paper to the contact address below. Notification: January 16, 2003 Camera Ready: January 23, 2003 (IEEE proceedings <http://www.computer.org/cspress/instruct.htm> format) Contact: Shin'ichi Konomi (konomi(a)db.soc.i.kyoto-u.ac.jp) The main topics of the conference include Squeak and Learning/Education. The proceedings will be published from IEEE Computer Society. Invited speakers/overseas invitees include Alan Kay, Kim Rose, Yoshiki Ohshima, Roxanne Maloney, Leo Burd, Andreas Raab, David Smith, Michael Rueger and Jeff Pierce. Japanese Squeakers: including Kazuhiro Abe, Masashi Umezawa, Satoshi Nishihara, Masato Sumi. Learning/Education experts: including Kumiyo Nakakoji. <Tentative Schedule> - 9:00AM--11:00AM Dr. Alan Kay's Lecture (Capacity of the lecture room: 200 people) -11:00AM--11:20AM Coffee Break -11:20AM--12:50PM 3 technical presentations (Single session) -12:50PM--1:50PM Lunch Break -1:50PM--5:00PM Parallel Sessions (Capacity of each room: 100 people) (Room A) "computer science session" 7 technical presentations (Room B) "application session" 4 technical presentations, Interactive event ("Squeak Festival"), Poster presentations Registration: Please fill in this <http://www.db.soc.i.kyoto-u.ac.jp/alankay/new_page_3.htm> form and send it to Hitomi Takada (takada(a)db.soc.i.kyoto-u.ac.jp).

1 0

Workspace menu cancel
by Markus Fritsche 29 Dec '02

29 Dec '02

Hi! Is it right, that (windows-) right-click (should I buy nail polish for my mouse buttons?) on a workspace and select "cancel" (when I want to cancel the menu - I got used to cancel unwanted menus with "cancel", call me windows infected if you like), the workspace is cleared. No undo (cmd-z) is possible. Is this the expected behaviour? Regards, Markus -- http://reauktion.de/archer/

1 0

[ENH] linesInHierarchy #2
by sps2000＠mail.com 28 Dec '02

28 Dec '02

from preamble: "Change Set: linesInHierarchy Date: 28 December 2002 Author: Steven Swerling Add a preference called showLinesInHierarchyViews. When true, feint lines are drawn in hierarchy views to make it more clear how the elements in the view are related. Added this version: Added drawing optimizations to the last version of this changeset, and a color fix. Also, incorporates Ned Konz' fix for list items that have icons (for the StarBrowser)"!

1 0

[ENH] FileListEmptyDirectoryFix
by sps2000＠mail.com 28 Dec '02

28 Dec '02

from preamble: "Change Set: FileListEmptyDirectoryFix Date: 28 December 2002 Author: Steven Swerling Currently in the hierarchy list FileList2, all folders have toggles that make them appear to have subfolders. This changeset removes the toggle from folders that do not contain subfolders. The old behavior of assuming that a subfolder exists is maintained in cases that would incur a performance hit (such as nonlocal directories). Changes this version: Catch Error in #hasContents causes default behavior (assume subfolder) instead of walkback."!

1 0

← Newer
1
2
3
4
5
6
7
8
...
46
Older →

Jump to page: