IRC log of #zope3-dev for Wednesday, 2005-08-10

*** _drzoltron has quit IRC00:02
*** gintas has quit IRC00:06
*** niemeyer has quit IRC00:07
*** vinsci has quit IRC00:22
*** hazmat has joined #zope3-dev00:27
*** timte has quit IRC00:29
*** vinsci has joined #zope3-dev00:40
*** RaFromBR1 has joined #zope3-dev00:40
*** vinsci has joined #zope3-dev00:40
*** hazmat_ has joined #zope3-dev00:55
*** hazmat has quit IRC00:55
*** hazmat_ is now known as hazmat00:55
*** ChanServ sets mode: +o hazmat00:55
*** RaFromBRC has quit IRC00:56
*** hazmat has left #zope3-dev00:59
*** hazmat has joined #zope3-dev00:59
*** ChanServ sets mode: +o hazmat00:59
*** RaFromBRC has joined #zope3-dev01:00
*** RaFromBR1 has quit IRC01:11
*** benji_york has quit IRC01:15
*** hazmat has quit IRC01:41
*** hazmat has joined #zope3-dev01:44
*** bradb is now known as bradb-away02:00
*** hazmat_ has joined #zope3-dev02:11
*** hazmat_ has joined #zope3-dev02:40
*** hazmat has quit IRC02:57
*** stub has joined #zope3-dev02:58
*** d2m has quit IRC03:07
*** RaFromBRC is now known as RaFromBRC|away03:47
*** MiUlEr has joined #zope3-dev04:04
*** hazmat_ has quit IRC04:06
*** RaFromBRC|away has quit IRC04:06
*** RaFromBRC has joined #zope3-dev04:08
*** hazmat_ has joined #zope3-dev04:11
*** RaFromBRC|away has joined #zope3-dev04:11
*** RaFromBRC|away has quit IRC04:12
*** hazmat_ has quit IRC04:22
*** hazmat_ has joined #zope3-dev04:22
*** hazmat_ has quit IRC04:41
*** hazmat_ has joined #zope3-dev04:41
*** hazmat_ has quit IRC04:45
*** hazmat_ has joined #zope3-dev04:45
*** tvon has quit IRC06:52
*** tvon has joined #zope3-dev07:36
*** bob2 has quit IRC07:43
*** bob2 has joined #zope3-dev07:44
*** RaFromBRC has quit IRC08:04
*** tvon has quit IRC08:13
*** tvon has joined #zope3-dev08:25
*** vinsci has quit IRC08:32
*** vinsci|2 has joined #zope3-dev08:32
*** vinsci|2 is now known as vinsci08:33
*** hdima has joined #zope3-dev09:06
*** yota has joined #zope3-dev09:23
*** hazmat has joined #zope3-dev09:35
*** MJ has quit IRC09:52
*** d2m has joined #zope3-dev10:07
*** MiUlEr has left #zope3-dev10:15
*** sashav has joined #zope3-dev10:32
*** philiKON has quit IRC10:46
*** mgedmin has joined #zope3-dev10:48
*** MJ has joined #zope3-dev10:48
*** ignas has joined #zope3-dev10:59
*** Aiste has joined #zope3-dev11:01
*** ignas has quit IRC11:01
*** ignas has joined #zope3-dev11:01
*** MJ has quit IRC11:11
*** MJ has joined #zope3-dev11:15
*** bob2 has quit IRC11:23
*** bob2 has joined #zope3-dev11:23
*** bob2 has quit IRC11:26
*** bob2 has joined #zope3-dev11:26
*** hazmat has quit IRC12:01
*** jinty has joined #zope3-dev12:12
*** timte has joined #zope3-dev12:13
*** jinty has quit IRC12:23
*** jinty has joined #zope3-dev12:23
*** projekt01 has joined #zope3-dev12:51
*** mkerrin has joined #zope3-dev13:33
*** stub has quit IRC13:44
*** SteveA has quit IRC13:48
*** SteveA has joined #zope3-dev13:51
*** anguenot has joined #zope3-dev14:01
*** bskahan has joined #zope3-dev14:11
*** ignas has quit IRC14:36
*** juka has joined #zope3-dev15:05
*** juka has left #zope3-dev15:11
*** efge has joined #zope3-dev15:15
*** bskahan has quit IRC15:16
*** bskahan has joined #zope3-dev15:48
*** bradb-away is now known as bradb16:06
*** zagy has joined #zope3-dev16:25
*** benji_york has joined #zope3-dev16:27
*** ignas has joined #zope3-dev16:29
*** srichter has quit IRC16:30
*** srichter has joined #zope3-dev16:34
*** strichter has joined #zope3-dev16:35
*** BjornT_ has joined #zope3-dev16:41
*** bskahan has quit IRC16:43
*** BjornT has quit IRC16:46
*** bskahan has joined #zope3-dev17:02
*** ChanServ sets mode: +o srichter17:12
*** strichter has quit IRC17:12
*** niemeyer has joined #zope3-dev17:12
*** sashav has quit IRC17:25
srichterhey anyone here?17:29
srichter...and willing to discuss some thoughts of mine?17:29
srichterok, so one of my goals for 3.2 in to make apidoc a standalone app17:30
srichterthe reason being that you want to be able to run APIDOC even when Zope 3 itself does not startup17:31
benji_yorkah, ok17:31
srichterbasically, right now apidoc is an addon for Zope 3 that requires Zope 3 to start up17:31
srichteralso, it does a lot of work when you first look at the code browser, for example17:32
srichterso Jim suggested a long time back to store all of the data in the Data.fs17:32
srichternow the question is: should only the views be stored or the data structure?17:33
mgedminsrichter, YAY!17:34
srichteranother option would be to write some fancy version of wget17:34
benji_yorkhmm, interesting17:35
srichterso I simply do not know the best route to take17:35
srichtercomments, ideas, anything?17:35
benji_yorkI'm a big fan of 80/20, so my initial reaction would be to store the views17:36
benji_yorkhmm, it might even be interesting to allow generation of html on disc, so no server required, just point your browser at a file:// url17:37
hdimasrichter: like pydoc?17:37
srichterhdima: I have not looked at pydoc for a long time, so I dunno17:37
srichterbenji_york: right, so you just do a recursive wget and that should be it, right?17:37
benji_yorkyep (as long as the recursion terminates) :)17:38
srichterI think wget does this automatically17:38
srichterso what would be disadvantages of the wget approach?17:39
hdimapydoc can store pages on disk or you can run it as server with -p PORT option17:39
benji_york       -l depth17:39
benji_york       --level=depth17:39
benji_york           Specify recursion maximum depth level depth.  The default maximum depth is 5.17:39
srichterit would be easy to find out what that would be for apidoc17:40
benji_yorkor if you are confident there are no loops, use --level=inf (for infinite recursion)17:42
srichterI am positive there are loops17:42
mgedminwget does a graph traversal17:43
mgedminloops are ok, as long as the graph itself is finite17:43
srichteryes, the graph is finite17:44
benji_yorksorry, I mis-spoke, instead of loops, let's say "spirals"  :)17:44
srichterI think the tricky part will be to find cases like: objectX/@@index.html versus objectX/index.html versus objectX/17:45
*** hdima has quit IRC17:46
srichterso I guess a little validator in Python would be sensible17:46
benji_yorkalso note that wget isn't installed by default on some systems (*cough* Windows *cough*)17:49
srichterI know, tough luck for them17:49
*** philiKON has joined #zope3-dev17:49
srichterthey have the choice to: have the live version or use the distributed package17:49
srichterphiliKON: we are discssing a way of making apidoc available offline; wget seems to be the way to go17:50
benji_york"the distributed package" being the result of running it on a stock install?17:50
benji_yorkso, what would it take for a user to update their copy of the static docs?  Start the server, run a command.17:52
projekt01srichter, +1 from me, but would be nice also to use pyDoc style comments, but perhaps this is out of scope for zope.17:53
srichterpreferably on Unix: make docs17:53
srichterand: make docs-package17:53
srichterprojekt01: what are "pyDoc style comments"?17:54
projekt01something like:17:55
projekt01 * Construct a new XMLHttp.17:55
projekt01 * @class This is the basic XMLHttp class.17:55
projekt01 * @constructor17:55
projekt01 * @param {string} url URL pointing to the server17:55
projekt01 * @return A new XMLHttp17:55
projekt01 */17:55
srichterOh, apidoc would need to support this; this is a different issue17:55
srichternote that I do not think that this syntax is that useful for Zope, since we have interfaces17:56
srichterWe really should support the two draft proposals from Jim that call for (optional) heavier interface declarations17:56
projekt01Hm, right, but it would be nice to have this handy java doc format for documentation17:57
benji_yorkI'm not sure I've ever seen those words use together before: "handy java doc format"  :)17:59
projekt01What's really missing is the documentation of return values, parameters, etc17:59
srichterprojekt01: yes, Jim's proposal covers this.18:00
projekt01Hm, I can't find them18:01
projekt01srichter, that looks a little bit heavy at the first time, but could be a good way for additional information18:04
srichterbut I think it is much better than the @comment thingies, because they cannot be used by the code, whereby MethodSpecifications can be used by the code to verify correct functioning18:06
*** gintas has joined #zope3-dev18:09
philiKONsrichter, so,you want to save static HTML?18:19
srichterwhatever pre-eval I am doing it will be static anyways18:20
philiKONfair enough18:20
philiKONwould be cool if there was a way to produce reST18:20
srichterso whether I only store static data or final files does not matter18:20
srichterno, it would not :-)18:20
srichterbecause the useful part about APIdoc is the linkage stuff18:20
srichterI would rather work on a PDF export18:21
philiKONgood point18:21
srichterwhich can be done with some effort using pdftemplate I developed for projekt0118:21
philiKONor by producing docbook18:22
philiKONactually, docbook format is pretty versatile18:22
philiKONand easy to generate (XML)18:22
srichteryeah, probably that will be better18:22
*** dman13_ has joined #zope3-dev18:23
*** dman13 has quit IRC18:23
*** vlado has joined #zope3-dev18:28
*** MJ has quit IRC18:34
*** vlado has quit IRC18:48
*** Aiste has quit IRC18:56
*** alga has joined #zope3-dev19:15
*** zagy_ has joined #zope3-dev19:31
*** bradb is now known as bradb-lunch19:34
*** zagy has quit IRC19:43
*** gintas has quit IRC19:46
*** SureshZ has joined #zope3-dev19:48
*** mgedmin has quit IRC20:05
*** kryptoer has joined #zope3-dev20:06
*** efge has quit IRC20:09
kryptoerhello all. I plan to build an app with zope and I am wondering if anyone is aware if zope 3 is currently supporting ZEOS and sessions. I don't see any reference to it anywhere in current docs20:10
*** sashav has joined #zope3-dev20:12
* genconc is away: gone home20:15
yotaphiliKON: ping20:20
philiKONyota, pong20:20
philiKONkryptoer, zope3 supports zeo; i haven't tested on any zeo setup yet, but it's supposedly modelled to work with zeo20:20
*** sashav has quit IRC20:21
philiKONyota, what's up?20:21
*** sashav has joined #zope3-dev20:22
yotaphiliKON: what is the right tool for write a book ?20:24
philiKONyota, what kind of book?20:24
yotatechnical book20:24
philiKONalready got a publisher?20:24
philiKONyota, what formats does the publisher accept?20:25
VladDrac(well, that would have been  the answer of my former professor, Tanenbaum ;)20:25
philiKONVladDrac, right...20:25
yotano, just an idea20:25
philiKONyota, well, i used docbook20:25
philiKONand liked it a lot20:25
kryptoerThanks for the feedback philiKON. That helps in my decision making as to where to devout my development effort. Basically want to make sure that when its time to handle scaling that I can workout replication and session failover20:25
VladDrac(he seriously still writes his books using troff)20:25
philiKONVladDrac, amazing...20:25
philiKONVladDrac, so, you were a student of Tanenbaum?20:26
VladDracon minix probably20:26
VladDracanyway, food20:26
VladDracphil: yeah, and I worked as a scientific programmer later on for him (the Globe project)20:26
philiKONkryptoer, in any case, there's enough knowledge in the zope community regarding zeo and sessions so that minor glitches can definitely be sorted out20:27
yotait's my first choice but I don't know how transform docbook with images to pdf and what editor to use20:28
philiKONi use emacs and nxml-mode20:28
philiKONit can validate xml documents according to relaxNG schemas while editing20:29
philiKONvery nice20:29
yotaah, emacs addict :)20:29
philiKONyes :)20:29
philiKONif you only need PDF output, i suggest you transform docbook to XSL-FO using the docbook-xsl stylesheets20:29
philiKONand then use a FO processor (e.g. fop) to make it into PDF20:29
philiKONfop is part of apache jakarta, though there haven't been any new releasees for over a year now ;(20:30
philiKONthere might be other xsl-fo processors20:30
philiKONthe route i went with the book was to produce latex because the publsiher wanted it20:31
philiKONi used the db2latex stylesheet package20:31
philiKONthis too is unfortunately not developed anymore20:31
philiKONbut a fork, dblatex, is20:31
philiKON(i think i'll switch to that one soon)20:31
yotaoki, thx20:32
philiKONlemme know if you have any questions20:32
philiKONthe nice thing about docbook is that you have a semantic definition of your text20:33
philiKONyou can transform it to anything basically20:33
yotayes, I like very much the concept : separate content and presentation20:34
yotaI hate wysiwyg app20:35
*** bradb-lunch is now known as bradb20:36
yotaphiliKON:  do you think to upgrade your book for next releases ?20:37
*** ignas has quit IRC20:41
kryptoerthanks philiKON! I am sure when I meet those crossroads that I will be more vocal. Thanks Again!20:41
*** kryptoer has quit IRC20:42
philiKONyota, at least i have plans to keep it up to date; we'll see about publishing. maybe a second edition for zope 3.2 or 3.3, but i'm not the one to decide that20:50
*** alga has quit IRC21:00
*** sashav has joined #zope3-dev21:03
*** bskahan has quit IRC21:04
*** hazmat_ has quit IRC21:15
*** hazmat_ has joined #zope3-dev21:16
*** hazmat_ has quit IRC21:22
*** hazmat_ has joined #zope3-dev21:23
*** fdrake has joined #zope3-dev21:24
fdrakesrichter, you there?21:24
srichterfdrake: yes21:25
srichtergood to see you here21:25
fdrakei just made changes to zpkg  :-)21:25
fdrakeare you planning to update your checkout of zpkg before making another Z3 release?21:26
srichteryes :-)21:26
srichterI was just about to ask you about it :-)21:26
fdrakethen I'll probably need to change some things in the Z3 tree  ;-)21:26
fdrakethe goal is to make the distribution-building do less, so we can eventually use zpkg for checkout-based builds as well21:27
fdrakethis change makes public include files get handled in the distutils phase21:28
srichterI am trying to bring the SchoolTool community to zpkgtools as well21:28
fdrakeand causes the component-based approach to produce fewer surprises21:28
fdrakebe sure to file any issues that arise in the Z3d collector21:28
fdrakethe twisted folks still have some issue w/ zpkg, I think21:29
srichterwe have some other refactoring to do before that21:29
fdrakewe == SchoolTool?21:29
fdrakeok, i'll take a look at using zpkg for Z3.1 now21:30
fdrakesomeone needs to run the zpkg tests on Windows, BTW  :-)21:30
fdrakemy Windows-fu isn't sufficient right now21:30
srichterhe he21:30
fdrakewell, it doesn't work out of the box now21:48
*** bradb is now known as bradb-bbl21:48
*** gintas has joined #zope3-dev21:48
fdrakehmm; the required changes go back to the ZODB packages as well  :-(21:52
fdrakei'll ask Tim how he wants to deal with them; he's generally approved of the changes for the ZODB source tree, but I'm not done with everything that'll be needed there21:54
srichterohm well, better do it now than later21:54
fdrakemight be easier to wait, though; i don't know how much time I'll have for this this week  :-(21:55
fdrakeif you want to do that, just don't pick up the last zpkg update from today21:55
fdrakethat way this doesn't interfere with the release schedule, and the changes can be backported for 3.1.121:56
srichterok, I have not updated yet21:56
*** bskahan has joined #zope3-dev21:57
fdrakewe'll see how far I can get21:58
srichterphiliKON: are you there?22:23
srichterphiliKON: benji_york: The problem with the wget approach is that we cannot use search functions22:30
benji_yorkgrep?  :)22:33
benji_yorkinstead of wget, could we have a minimal zope just with the "captured" apidoc info?22:34
benji_yorkperhaps with its own ZODB?22:36
srichterright this is what Jim suggested22:40
srichtermaybe we can have a dumbed down version of the search22:40
srichterthat is based on JS22:40
benji_yorkor, we could replace the search facility in the static copy with a simple-minded javascript that searches an html index22:40
benji_yorkhey, that was my idea :)22:41
srichterbasically you have a list of say all interfaces and a JS simply hides them22:41
srichterI guess that would be easier22:41
benji_yorkI like the idea.22:41
srichterI have to write some custom views anyways, since I do not wnat to show the preferences on the static version either22:42
benji_yorkah, yep22:42
srichterok, someone agrees, this is good22:42
srichterI also need to make the menu JS based22:42
srichterotherwise it is downloading all the menu states22:42
benji_yorkhmm, it would be nice if we could compile a CHM of the docs... I'd really like that.22:43
benji_york(then no search would be needed)22:43
benji_yorkoh... that's windows HTML based help format, Python docs come as CHM now22:43
benji_yorkI like them so much I use xCHM on my linux boxes to view them22:44
srichterinteresting, though this would require rewriting a lot of the templates22:45
benji_yorkwhy do you say that? (not that I know; one way or the other)22:45
srichterit was more a question than a statement, sorry22:46
benji_yorkoh, I don't know, but it appears to be fairly easy.  The Python CHM is just a straight rip of the python HTML docs22:47
d2mCHM is HTML pages compiled into 1 big binary file, with indexes attached22:48
srichterI see22:48
d2myou could as well python pickle a site and use it through a frontend22:49
benji_yorkI just asked Tim and Fred (they did the original CHM for python)..22:49
projekt01fdrake, what is to do, to get zpkg work on windows?22:49
benji_yorkthey said that building the CHM-specific index took a bit of work, but other than that the HTML was a stright copy22:50
benji_yorkand since they already wrote the indexer, we could probably reuse it22:50
srichterthere is a pyCHM project out there22:50
benji_yorkI think that's  more about reading CHMs, not writing them22:51
srichterah, ok22:52
srichtercan we create the CHM file on Linux as well?22:52
benji_yorkyou have to use Microsoft's CHM compiler (on Windows)22:54
benji_yorkI can help if need be22:54
fdrakeprojekt01, you mean run the tests?22:54
srichterI wonder whether it will run under Wine?22:54
projekt01fdrake, no build the .exe22:55
srichterbenji_york: whether the CHM compiler will run using wine22:56
projekt01fdrake, I would like to use the zpkg for our Tiks framework, but I like to do it on a windows box22:57
*** mkerrin has quit IRC22:58
benji_yorksrichter, right.  I don't know if it'll run.  I would think so, because it's not a complex piece of software.22:59
benji_yorkwell, I have to go.  I'll be back later.22:59
projekt01fdrake, right now I use BitRock for building the installer.exe and I have to add the packages by hand.22:59
projekt01fdrake, I think this could be done by zpkg, right?23:00
*** SureshZ has joined #zope3-dev23:02
fdrakeprojekt01, by .exe, do you mean the windows installed23:05
*** RaFromBRC has joined #zope3-dev23:06
projekt01But the important part is to "package" the source for the installer23:06
fdrakezpkg will generate a tarball; that's a "normal" distutils package23:06
fdrakeyou can then run "python2.4 bdist_wininst" in an unpacked installer23:06
fdrakei don't know what bitrock is23:07
fdrake(no, I'm not a Windows user!)23:07
projekt01It's just a cross plattform installer where can build the installer for all plattforms.23:07
fdrakezpkg collects all the metadata that's needed to drive distutils23:07
fdrakethen what you need is a distutils command like bdist_bitrock23:08
philiKONhi fdrake23:08
fdrakehey, philiKON !23:08
fdraketestRunIgnoresParentSignals is evil  :-(23:08
philiKONsrichter, re: wget on docs23:09
philiKONsrichter, well, i thought the point of off-line docs was that they'd be searchable on the filesystem?!?23:09
fdrakephiliKON, we'd get that with CHM files, at least if xCHM is any good23:10
fdrakeit's been over a year since I looked at it though23:10
philiKONyup; CHM is just one format, though. point is, if we have it in some filesystem format (docbook, HTML, CHM, PDF. etc), people can search it with their tools23:11
fdrakeand HTML is the easiest to scrape out  :-)23:11
philiKONi hate grepping html23:11
fdrakebut it can be converted to other things, like CHM23:11
fdrakeit's an easy way to get the docs out of Zope23:12
srichterKDE has a good CHM viewer23:12
philiKONi would volunteer to write docbook views for apidoc23:12
fdrakesomething more structured would be ideal, I think23:12
philiKONat least to help with them23:12
fdrakeooh, that would be nice23:12
fdrakei'll contribute a few pointy brackets  :-)23:12
srichterphiliKON: I take you up on that23:13
projekt01fdrake, do you know what's exactly missing for running zpgk on windows?23:13
srichterI will definitely do an HTML export too23:13
philiKONi thought the point about the off-line tools were a) people can use their favourite tools to search and view (no browser) and b) people don't have to run a Zope3 instance23:13
philiKONof course, point b) will also take the interactivity of apidoc away23:13
philiKONmeaning, you've registered your own adapter, it will show up in apidoc23:14
fdrakeprojekt01, no; did you have troubles?23:14
philiKONbut not in the static docs unless you regenerate the docs23:14
projekt01Yes, popen doesn't work23:14
philiKONof course, this way apidoc can become a pretty good API dcoumentation generation facility... :)23:14
srichterI am only going for point (b) at this stage23:14
srichterthis was the use case brought to me most23:14
philiKONi have no idea how to view CHM on mac23:15
philiKONis CHM pure XML or is it zipped/processed in some other way?23:15
philiKONcool, thanks :)23:16
* philiKON should've googled, of course23:16
projekt01fdrake, srichter, I think Thomas from SchoolTools reported this also some time ago, I'm not sure if he fixed it in his own workspace. Perhaps srichter knows more about this issue?23:16
srichterwhich issue?23:17
projekt01srichter, zpkg on windows23:17
fdrakeprojekt01, ok, there's a calll to os.spawnlp() that doesn't appear to be supported on Windows; the popen() call is likely in the implementation of that23:17
fdrakei'll have to look into it a bit more23:18
projekt01py 2.4.1 has a new popen implementation ;-)23:18
srichterI dunno anything about Tom having troubles :-)23:18
srichterwow, wget runs since more than 1 hour now23:19
fdrakeprojekt01, what Python are you using?23:19
srichtermainly getting menu entries23:19
fdrakei could use the tarfile module, but that doesn't deal with compression23:19
projekt01fdrake, now 2.4.1, But the time I tried I used 2.3.523:19
srichterphiliKON: is it possible to get a fully expanded tree with at once?23:19
fdrakethe compression would be available on Windows, but not necessarily on Unix23:19
gintasprojekt01, which Tom from SchoolTool did you mean?23:20
fdrakeprojekt01, try again w/ 2.4.1 then  :-)23:20
*** stub has joined #zope3-dev23:20
projekt01gintas, I'm not sure but I think his name is Thomas or not?23:21
projekt01fdrake, Ok will try23:21
gintaswell, we have Tom von Schwerdtner and Tom Hoffman23:22
fdrakeif it still doesn't work, post an issue on the Z3d collector23:22
gintasfor some reason I think you have Marius in mind23:22
* srichter just discovered his Aertze MP3 CD... I am in a good mood now :-)23:22
projekt01gintas, Tom Hoffman23:22
projekt01See, and search for "The Zope3 Windows installer is indeed built by disutils23:23
philiKONsrichter, ew, i don't remember23:24
philiKONsrichter, there might be something like expandAllNodes...?23:24
philiKONbut i don't think i ever had a use case; i actually had the use case than i didn't want to load all nodes to construct the tree (tree might come from SQL, for example)23:25
srichterI want it for the static documentation23:26
srichterI want to show the entire tree23:26
projekt01srichter, here's a better link about the problem from Tom,
srichterand use JS to hide elements23:26
philiKONsrichter, you could probably easy enhance it to have that functionality23:27
philiKONanyway, i'm out23:27
philiKONhave a good night23:27
srichtersee ya23:27
*** MJ has joined #zope3-dev23:29
*** fdrake has quit IRC23:52
projekt01fdrake, ayt?23:54
*** zagy_ has quit IRC23:58
*** zagy_ has joined #zope3-dev23:58

Generated by 2.15.1 by Marius Gedminas - find it at!