From: Neil W. <row...@ho...> - 2007-09-01 08:19:30
|
Morning Greg, Kenneth, Am I write in thinking that we Jython users are mainly a sub-group of Python users? If so, I wonder what can be done to attract Java hackers over. Perhaps the wiki could occasionally diverger into two sections? Jython docs for Java guys & Jython docs for Python Guys? Neil. -- please ignore the following crap, which has been appended by Micro$oft without my permission. _________________________________________________________________ Can you see your house from the sky? Try Live Search Maps http://maps.live.com |
From: Kenneth M. <ken...@sb...> - 2007-09-01 09:24:42
|
At the moment, I'd guess that Jython users are a (small) sub-group of Python hackers, i.e. not users but rather the subgroup who both use and really enjoy playing with Python. This is an unfortunate consequence of the fact that Jython went dormant for such a long time; there's going to be a certain hesitancy about Jython until people believe it's here to stay. I myself confess to using JRuby at the moment, though I'd (probably) prefer Jython, which is why I'm looking at it :-). IMHO, for Jython (or any Java scripting language) to work in the long term, it needs to satisfy the fact that Java is in use in a more heavy-duty factory-line sort of environment than Python; I think a much higher percentage of Java users just want to get the job done, as opposed to a larger percentage of Pythonistas who also enjoy playing with tech. This makes sense--who'd think that playing with tech was fun if you had to program Java all day? To work with that sort of mind set, Jython needs several things. 1) It should be easy to set up and to start to do simple things with. So far, so good. 2) It needs to be really, really solid, both with respect to correctness and (to a useful level) completeness in its interoperability with Java. I don't know what Jython's status is at this point. Both Groovy (as of a few months ago) and Ruby (as of right now) have at least a couple of major problems (though I do have a lot of admiration for both teams--they seem like bright people.) 3) Here is the big one that will be the biggest stumbling block for most of these efforts; it needs good documentation so that a person can start doing useful things immediately, and also be able to go back to find the highly technical information they'll need as they start tacking more complex problems. For me, the biggest problem with Jython right now is the lack of a site I can go to that is kept up to date and says, "here is the status of Jython, here is what it can do, here is what it can't do, and here is everything you need to know about interoperating with Java". I don't know about others, but my main interest in getting Python or Ruby working with Java is to (finally) have a (half-)decent, true, crossplatform GUI platform that I can write a decent high-level API around. To get more people interested in Jython (assuming it's solid and complete enough right now to focus on this), perhaps illustrate Jython by example by starting to develop a UI library on top of Swing, and documenting aspects of that? This approach has these advantages: 1) It can solve a real problem; writing Swing apps in Java is absurdly nasty, that's why all these GUI dev tools are needed. 2) It's visually appealing, so you'll get eyeballs. 3) It provides lots of scope for illustrating how to do things like subclass Java classes in Jython (if that's possible), define Java classes in Jython (if that's possible), etc. Just a thought. Hope this helps, Ken Neil Wallace wrote: > Morning Greg, Kenneth, > > Am I write in thinking that we Jython users are mainly a sub-group of > Python users? > If so, I wonder what can be done to attract Java hackers over. > > Perhaps the wiki could occasionally diverger into two sections? Jython > docs for Java guys & Jython docs for Python Guys? > > Neil. > > |
From: Moore, G. <Gre...@ad...> - 2007-09-04 19:55:16
|
Greetings Neil, Kenneth, other readers. :) I cant speak for others only my own uses and things I've been involved with in a Jythonic (is that a word?) way. First a little background, I am, and have been a software QA engineer for many years. Mostly testing internet related technologies. Much of what I've worked on has been Java however in the last few years that has been increasingly python. For me Jython is wonderful because it allows me to quickly develop tests for java "stuff" that would be much more labor intensive with a pure java solution. Additionally, Junior qa testers can pick up python much quicker than a more formal language like Java which makes them better more versatile testers.=20 I do a lot of test tool development and most of that is done using Jython and swing because every one has a JVM on their PC and jar files are well understood. So the I have to agree with Kenneth when he says "Java users just want to get the job done" Jython lets develop complex tools much more rapidly then pure Java but allows me to use the existing java code base. I'm not sure if I'd consider myself a python hacker but I've come to love python and the flexibility that python/jython provide and the ease of use. I have no experience with Groovy or Ruby. I've looked at both didn't care for either but then again I don't like Eclipse either. I guess that make me "Special". LOL.=20 I believe that that documentation that accurate and current is critical. Maintaining 2 sets of documentation one from the python perspective and one from the Java perspective would be a PITA. One document that has notes where needed for the Java programmer or for the Python programmer that tells them ok this is how you do it in Java/Python and this is how you do 'it' in jython. The documentation need to have EXAMPLES!!! In an ideal world there would be a Dive into Jython book. I think that should be our goal. Reference manuals are great and wonderful but I refer to dive in to python more then I do the Python reference. Even something like the Python Quick ref. (http://rgruet.free.fr/PQR25/PQR2.5.html ) would be a great start. Personally I'd like to see more examples of converting java to Jython like appendix E of Jython Essentials or http://blogs.sun.com/sundararajan/entry/java_integration_beanshell_and_j ython =20 We can talk (or type) about this until the cows come home so whats the next step toward making better Jython docs a reality? Greg. Neil Wallace wrote: > Morning Greg, Kenneth, > > Am I write in thinking that we Jython users are mainly a sub-group of > Python users? > If so, I wonder what can be done to attract Java hackers over. > > Perhaps the wiki could occasionally diverger into two sections? Jython > docs for Java guys & Jython docs for Python Guys? > > Neil. > > =20 ------------------------------------------------------------------------ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Jython-users mailing list Jyt...@li... https://lists.sourceforge.net/lists/listinfo/jython-users This message and any attachments are intended only for the use of the add= ressee and may contain information that is privileged and confidential. I= f the reader of the message is not the intended recipient or an authorize= d representative of the intended recipient, you are hereby notified that = any dissemination of this communication is strictly prohibited. If you ha= ve received this communication in error, please notify us immediately by = e-mail and delete the message and any attachments from your system. |
From: Frank W. <fwi...@gm...> - 2007-09-04 20:23:12
|
On 9/4/07, Moore, Greg <Gre...@ad...> wrote: > We can talk (or type) about this until the cows come home so whats the > next step toward making better Jython docs a reality? better docs in the wiki would be a good start, and anybody can do it (you just need to create a user for yourself) -- I'd especially like to see wiki documents in ReST format, though I have only realized this is possible lately (and don't know if others will want to do this) -- the website is in ReST so moving from sort of unofficial wiki docs to official website docs would become pretty easy. To do a ReST doc in the wiki you start it with: #format rst Here is a link to ReST documentation: http://docutils.sourceforge.net/rst.html -Frank |
From: <ast...@gm...> - 2007-09-03 03:40:56
|
On 9/1/07, Kenneth McDonald <ken...@sb...> wrote: > At the moment, I'd guess that Jython users are a (small) sub-group of > Python hackers, > i.e. not users but rather the subgroup who both use and really enjoy > playing with Python. > This is an unfortunate consequence of the fact that Jython went dormant > for such a long time; > there's going to be a certain hesitancy about Jython until people > believe it's here to stay. > I myself confess to using JRuby at the moment, though I'd (probably) > prefer Jython, which > is why I'm looking at it :-). > I'm 100% Jythonista. No dabbling with Python whatsoever. As for JRuby, it sucks. Just compare the two: JRuby: require 'java' include_class "javax.swing.JFrame" include_class "javax.swing.JLabel" frame = JFrame.new() frame.getContentPane().add(JLabel.new("This is an example.")) frame.pack() frame.setVisible(true) Jython: from javax.swing import JFrame, JLabel frame = JFrame('no more new keyword', defaultCloseOperation=JFrame.EXIT_ON_CLOSE) frame.contentPane.add(JLabel('Too_many_underscores_are_ugly')) frame.pack() frame.show() With Jython, it's easy to convert Java code to Jython code. Plus, you can do this in Jython: iframe = JInternalFrame('Title', internalFrameClosing=myfunc, internalFrameActivated=etc.) That means you can make a subclass of a JFrame/JInternalFrame, et. al. without having to implement an interface. Example: class MDIChildFrame(JInternalFrame): def __init__(self, strTitle): JInternalFrame.__init__(self, strTitle, 1, 1, 1, 1, internalFrameActivated=self.onActive, internalFrameIconified=self.onIconify) self.contentPane.add(JButton('press me', actionPerformed=self.onButton)) def onActive(self, event): print 'This iframe was activated!' def onIconify(self, event): print 'This iframe was iconified!' def onButton(self, event): print 'Dont try this at home kids' Basically, I've covered the event-handling for both the JInternalFrame AND its JButton in one class. It's simple, straightforward, clean, manageable, not confusing, etc. And I seriously hate the def/end in Ruby. Reminds me of VB. > 2) It needs to be really, really solid, both with respect to correctness > and (to a useful level) > completeness in its interoperability with Java. I don't know what > Jython's status is at this point. > Both Groovy Groovy is a joke. Sorry if there are any groovy developers out there :-/ > > 3) Here is the big one that will be the biggest stumbling block for most > of these efforts; it needs > good documentation so that a person can start doing useful things > immediately, and also be able to go > back to find the highly technical information they'll need as they start > tacking more complex > problems. > I guess that's true. The docs will come when Jython gains more popularity. > I don't know about others, but my main interest in getting Python > or Ruby working with Java is to (finally) have a (half-)decent, true, > crossplatform GUI platform > that I can write a decent high-level API around. To get more people > interested in Jython > (assuming it's solid and complete enough right now to focus on this), > perhaps illustrate Jython > by example by starting to develop a UI library on top of Swing, and > documenting aspects of that? > This approach has these advantages: > > 1) It can solve a real problem; writing Swing apps in Java is > absurdly nasty, that's why all > these GUI dev tools are needed. > > 2) It's visually appealing, so you'll get eyeballs. > > 3) It provides lots of scope for illustrating how to do things like > subclass Java classes in Jython > (if that's possible), define Java classes in Jython (if that's > possible), etc. > It's true that writing a Swing app sucks. This is not the case in Jython, however. And I don't think a UI lib should be written on top of Jython. No need to reinvent the wheel, IMHO. I'm developing a desktop application in Jython right now and I think this has been the easiest language to use when writing GUI apps. Here's a sample line from my project: s = (105, 25) self.button = JToggleButton(self._getTitle(), self._getIcon(), focusable=0, preferredSize=s, minimumSize=s, maximumSize=s, font=self._getFont(), horizontalAlignment=SwingConstants.LEFT, margin=(1, 2, 1, 2), actionPerformed=self.onButton) And, you can still do self.button.setHorizontalAlignment(something) later on. Even if we compare Jython and WxPython, the former is more Pythonic than the latter. Here's a screenshot of the first generation of my app, written purely in Jython: http://astigmatik.googlepages.com/jeprox.jpg Sorry for the quality of the screenshot. I'm rewriting it right now, and the newer version looks better and has more features. And that concludes my $0.02 on this discussion :-) Best regards, astigmatik |
From: Kenneth M. <ken...@sb...> - 2007-09-03 09:07:16
|
ast...@gm... wrote: > I'm 100% Jythonista. No dabbling with Python whatsoever. As for JRuby, > it sucks. Just compare the two: > > JRuby: > require 'java' > include_class "javax.swing.JFrame" > include_class "javax.swing.JLabel" > frame = JFrame.new() > frame.getContentPane().add(JLabel.new("This is an example.")) > frame.pack() > frame.setVisible(true) > > Jython: > from javax.swing import JFrame, JLabel > frame = JFrame('no more new keyword', > defaultCloseOperation=JFrame.EXIT_ON_CLOSE) > frame.contentPane.add(JLabel('Too_many_underscores_are_ugly')) > frame.pack() > frame.show() > > Hopefully there is an implicit smiley face in the above; if not, I suggest you reexamine your epistemological methodologies. This is off the top of my head (no testing) so it will have errors, but the JRuby example is actually something closer to: include Java frame = javax.swing.JFrame.new frame.contentPane.add(JLabel.new("example")) frame.pack frame.visible = true Ruby does have several features that are much nicer than those available in Python, including a cleaner way of implementing properties, a somewhat cleaner approach to OO programming (few or no special functions), a defined documentation markup syntax, etc. Similarly, Python is much nicer than Ruby in some areas, enough so that I rather strongly prefer Python. But "JRuby sucks" is neither accurate nor helpful. Ken |
From: Frank W. <fwi...@gm...> - 2007-09-03 17:47:13
|
On 9/2/07, ast...@gm... <ast...@gm...> wrote: > As for JRuby, it sucks. > Groovy is a joke. I really wish we could all refrain from these sorts of characterizations. Comparing the pros and cons of competing languages is fine, language advocacy is fine, even Jython zeal is fine, but I'd rather not see this sort of name-calling. -Frank |
From: <ast...@gm...> - 2007-09-04 04:23:14
|
My apologies to all... a bit overzealous... astigmatik On 9/4/07, Frank Wierzbicki <fwi...@gm...> wrote: > On 9/2/07, ast...@gm... <ast...@gm...> wrote: > > > As for JRuby, it sucks. > > > Groovy is a joke. > > I really wish we could all refrain from these sorts of > characterizations. Comparing the pros and cons of competing languages > is fine, language advocacy is fine, even Jython zeal is fine, but I'd > rather not see this sort of name-calling. > > -Frank > |
From: Moore, G. <Gre...@ad...> - 2007-09-04 22:03:38
|
I think the point that David made is a good one but I think a stand alone non-wiki version of the docs would be good I still like looking at paper. A wiki version would be nice because of could be commented on and such... more of a 'living' document. I think we need to be careful when (if?) writing about future changes/features. Maybe in a separate document or in a "use at your own risk" appendix=20 A few questions come to mind: Could a PDF be generated from a ReST document? Would there be some sort of review of the changes to ensure accuracy and maybe quality? Where would one find an editable version of the user guide? Where would one obtain info on whats changed and, as David put it, "a bunch of 2.3-2.5 functionality waiting in the wings". I'm sure there are other questions but I cant think of any more at the moment? Thanks for the ReST link, Frank Regards, Greg. -----Original Message----- From: jyt...@li... [mailto:jyt...@li...] On Behalf Of Frank Wierzbicki Sent: Tuesday, September 04, 2007 1:23 PM To: Moore, Greg Cc: jyt...@li... Subject: Re: [Jython-users] Where best to find documentation? On 9/4/07, Moore, Greg <Gre...@ad...> wrote: > We can talk (or type) about this until the cows come home so whats the > next step toward making better Jython docs a reality? better docs in the wiki would be a good start, and anybody can do it (you just need to create a user for yourself) -- I'd especially like to see wiki documents in ReST format, though I have only realized this is possible lately (and don't know if others will want to do this) -- the website is in ReST so moving from sort of unofficial wiki docs to official website docs would become pretty easy. To do a ReST doc in the wiki you start it with: #format rst Here is a link to ReST documentation: http://docutils.sourceforge.net/rst.html -Frank ------------------------------------------------------------------------ - This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Jython-users mailing list Jyt...@li... https://lists.sourceforge.net/lists/listinfo/jython-users This message and any attachments are intended only for the use of the add= ressee and may contain information that is privileged and confidential. I= f the reader of the message is not the intended recipient or an authorize= d representative of the intended recipient, you are hereby notified that = any dissemination of this communication is strictly prohibited. If you ha= ve received this communication in error, please notify us immediately by = e-mail and delete the message and any attachments from your system. |
From: Frank W. <fwi...@gm...> - 2007-09-05 14:01:12
|
On 9/4/07, Moore, Greg <Gre...@ad...> wrote: > A few questions come to mind: > > Could a PDF be generated from a ReST document? This is doable -- though a little tricky because the path is to first turn the ReST doc into LaTex and then to PDF. Both steps are programmatic and I actually kind of like LaTex -- or at least I like the print that results from its use. No, on second thought I think I actually like putting my docs through a compile step :). > Would there be some sort of review of the changes to ensure accuracy and > maybe quality? I think this would occur when docs are moved from the wiki to the website. I think the website, or at least certain pages on the website would be the target of pdf generation if we end up doing that. The possibility of creating pdfs was actually one of the reasons I moved to ReST on the website. > Where would one find an editable version of the user guide? At the bottom of every page on the website is a "view source" link -- this gives you the ReST source. > Where would one obtain info on whats changed and, as David put it, "a > bunch of 2.3-2.5 functionality waiting in the wings". The ReST sources can also be found in the subversion repo here: https://jython.svn.sourceforge.net/svnroot/jython/trunk/website The ReST sources are in the Project directory. -Frank |
From: Frank W. <fwi...@gm...> - 2007-09-05 14:04:19
|
On 9/5/07, Frank Wierzbicki <fwi...@gm...> wrote: > LaTex oops that should have been LaTeX -- I knew that didn't look right. -Frank |
From: Moore, G. <Gre...@ad...> - 2007-09-05 17:45:42
|
Frank,=20 Thanks for the answers. Which of course generated a few more questions. :) >> >> Could a PDF be generated from a ReST document? > This is doable -- though a little tricky because the path is to first > turn the ReST doc into LaTex and then to PDF. Ok that's interesting. I thought the LaTeX work I did a while back would be a one time thing. Good to see that wont go to waste. :) I've generated some really nice PDF's using this method. LaTeX is pretty amazing one you get into it.=20 >> Would there be some sort of review of the changes to ensure accuracy and >> maybe quality? > > I think this would occur when docs are moved from the wiki to the > website. =20 Forgive my ignorance, but why move them from the wiki to the web site. Wouldn't the link to the docs or doc area on the wiki sufficient? Maybe a link to an online viewable version and a link to downloadable PDF version. Wouldn't that be less work for you or Charlie?=20 >> Where would one obtain info on whats changed and, as David put it, "a >> bunch of 2.3-2.5 functionality waiting in the wings". > The ReST sources can also be found in the subversion repo here: > https://jython.svn.sourceforge.net/svnroot/jython/trunk/website >=20 > The ReST sources are in the Project directory. This is helpful but I was thinking of terms of Jython functionality that is not yet documented. Like new functions or operators, etc should javadocs be generated and available? This message and any attachments are intended only for the use of the add= ressee and may contain information that is privileged and confidential. I= f the reader of the message is not the intended recipient or an authorize= d representative of the intended recipient, you are hereby notified that = any dissemination of this communication is strictly prohibited. If you ha= ve received this communication in error, please notify us immediately by = e-mail and delete the message and any attachments from your system. |
From: Frank W. <fwi...@gm...> - 2007-09-05 22:09:51
|
On 9/5/07, Moore, Greg <Gre...@ad...> wrote: > Frank, > > Thanks for the answers. Which of course generated a few more questions. > :) > > >> > >> Could a PDF be generated from a ReST document? > > This is doable -- though a little tricky because the path is to first > > turn the ReST doc into LaTex and then to PDF. > > Ok that's interesting. I thought the LaTeX work I did a while back would > be a one time thing. Good to see that wont go to waste. :) I've > generated some really nice PDF's using this method. LaTeX is pretty > amazing one you get into it. > > >> Would there be some sort of review of the changes to ensure accuracy > and > >> maybe quality? > > > > I think this would occur when docs are moved from the wiki to the > > website. > > Forgive my ignorance, but why move them from the wiki to the web site. > Wouldn't the link to the docs or doc area on the wiki sufficient? Maybe > a link to an online viewable version and a link to downloadable PDF > version. Well, if we generated pdf off of the wiki content, it would probably be a manual process and would likely get out of date quickly. The website is a more stable source and so would be a safer bet for pdfs (also the website is generated by a script -- pdf support could be added there) The main thing I was saying is if we are going to check the wiki content I was thinking we might be able to create vetted "official" content from it -- Reviewing for quality would be easier this way, at least for me. This is just a thought though. > Wouldn't that be less work for you or Charlie? Not if we want to do ongoing reviews for quality. Since anyone can change the wiki at any time such reviews possibly be hard. > > >> Where would one obtain info on whats changed and, as David put it, "a > >> bunch of 2.3-2.5 functionality waiting in the wings". > > The ReST sources can also be found in the subversion repo here: > > https://jython.svn.sourceforge.net/svnroot/jython/trunk/website > > > > The ReST sources are in the Project directory. > > This is helpful but I was thinking of terms of Jython functionality that > is not yet documented. The best place to find this info is the NEWS file which (for the 2.2 release) is here: https://jython.svn.sourceforge.net/svnroot/jython/tags/Release_2_2/jython/NEWS > Like new functions or operators, etc should > javadocs be generated and available? Generating the javadocs is not a bad idea, although the comments are sparse in much of the codebase. |
From: Moore, G. <Gre...@ad...> - 2007-09-06 16:27:48
|
> The main thing I was saying is if we are going to check > the wiki content I was thinking we might be able to create vetted > "official" content from it -- Humm I hadn't thought of that. That's a good point. Sound good to me. Ok so one final question, for now at least :) I know generally what I'd like to see but I maybe be missing areas from my 'list' so, especially for you and the other developers; What are the things that need to be added / changed / deleted / rewritten with what currently exists?=20 Regards, Greg. -----Original Message----- From: Frank Wierzbicki [mailto:fwi...@gm...]=20 Sent: Wednesday, September 05, 2007 3:10 PM To: Moore, Greg Cc: jyt...@li... Subject: Re: [Jython-users] Where best to find documentation? On 9/5/07, Moore, Greg <Gre...@ad...> wrote: > Frank, > > Thanks for the answers. Which of course generated a few more questions. > :) > > >> > >> Could a PDF be generated from a ReST document? > > This is doable -- though a little tricky because the path is to first > > turn the ReST doc into LaTex and then to PDF. > > Ok that's interesting. I thought the LaTeX work I did a while back would > be a one time thing. Good to see that wont go to waste. :) I've > generated some really nice PDF's using this method. LaTeX is pretty > amazing one you get into it. > > >> Would there be some sort of review of the changes to ensure accuracy > and > >> maybe quality? > > > > I think this would occur when docs are moved from the wiki to the > > website. > > Forgive my ignorance, but why move them from the wiki to the web site. > Wouldn't the link to the docs or doc area on the wiki sufficient? Maybe > a link to an online viewable version and a link to downloadable PDF > version. Well, if we generated pdf off of the wiki content, it would probably be a manual process and would likely get out of date quickly. The website is a more stable source and so would be a safer bet for pdfs (also the website is generated by a script -- pdf support could be added there) The main thing I was saying is if we are going to check the wiki content I was thinking we might be able to create vetted "official" content from it -- Reviewing for quality would be easier this way, at least for me. This is just a thought though. > Wouldn't that be less work for you or Charlie? Not if we want to do ongoing reviews for quality. Since anyone can change the wiki at any time such reviews possibly be hard. > > >> Where would one obtain info on whats changed and, as David put it, "a > >> bunch of 2.3-2.5 functionality waiting in the wings". > > The ReST sources can also be found in the subversion repo here: > > https://jython.svn.sourceforge.net/svnroot/jython/trunk/website > > > > The ReST sources are in the Project directory. > > This is helpful but I was thinking of terms of Jython functionality that > is not yet documented. The best place to find this info is the NEWS file which (for the 2.2 release) is here: https://jython.svn.sourceforge.net/svnroot/jython/tags/Release_2_2/jytho n/NEWS > Like new functions or operators, etc should > javadocs be generated and available? Generating the javadocs is not a bad idea, although the comments are sparse in much of the codebase. This message and any attachments are intended only for the use of the add= ressee and may contain information that is privileged and confidential. I= f the reader of the message is not the intended recipient or an authorize= d representative of the intended recipient, you are hereby notified that = any dissemination of this communication is strictly prohibited. If you ha= ve received this communication in error, please notify us immediately by = e-mail and delete the message and any attachments from your system. |