seaside

seaside@lists.squeakfoundation.org

2 participants
8164 discussions

[Feature Suggestion] FileLibrary that serves from disk (A mix of WAFileMetadataFileLibrary and WAExternalFileLibrary)
by Jupiter Jones 05 May '24

05 May '24

My project repo has a directory structure that includes... <MyProjectRepo>/web_root/files/MyFileLibrary … that stores all the files (images / scripts / etc) that are used in my Seaside app. In production, the web server serves files from the web_root directory so all the WAFileMetadataLibrary urls (ie. /files/MyFileLibrary/*) work as expected. In development, my subclasses of WAFileMetadataLibrary serve files from the seaside file library handler... /files/MyFileLibrary This all works perfectly, since the developers don’t have a web server so need to get all files via the image, and both development and production urls are the same. It’s an ideal solution for web apps without a lot of assets. The biggest issue is that as our applications grow, all the files are on disk, are also in the image, and also end up in the source directory in a single text document MyFileLibrary.class.st <http://myfilelibrary.class.st/> which ends up huge, slow, and should be unnecessary. Ideally, in development, it would be nice to have a FileLibrary that works identically to a WAFileMetadataLibrary (ie. stores meta data and generates url's), except it doesn’t store the file contents in the image. Rather it streams them from disk via zinc when requested. I know there’s a WAExternalFileLibrary, however, AFAIK it can’t be served via /files/MyExternalFileLibrary, and doesn’t take advantage of the close integration of an application and its file libraries. Ie. myApplication addLibrary: MyFileLibrary. So the question… does anyone have anything like this? Or is there anyone else with growing applications that would rather not have multiple copies of all your web application assets, including a copy inside the image? Any thoughts? Cheers, Jupiter

2 2

Logging messages that play nicely with server tools
by Tim Mackinnon 03 May '24

03 May '24

Hi everyone - not strictly a seaside question - but as many of you deploy seaside on web servers, you may know what I'm talking about. I have a Pharo image in one of the wonderful BA Docker containers, and it runs well - and the host I'm using will show the logs for you, so you can figure out what is going on... well that is if your logs come out properly (and of course, if it gets really hairy then you can get a VNC session onto the image and figure stuff out) So logs are handy, and pharo these days has a nice headless mode that redirects the Transcript to stdout - and there are also a few decent logging frameworks as well. But as most things go to the Transcript, and that goes to stdout - it should be good. HOWEVER - flushing is the killer, as if things happen and the last thing goes wrong, but the output isn't flushed, then you aren't going to see it. So my question is how to properly flush? And I'm sure I've read something about this before, but I can't find it. From memory, you often need to have a Transcript cr. to flush your last line. BUT, most things in the pharo image seem to use "self crTrace:" these days, which is a cr to ensure the previous msg is separated from what you want to write, and then you write your line out. However, as there is no cr - you might not see it. So I tried changing my stuff to use "self traceCr:" (which is in the image), and that still didn't seem to work - the last failing line wasn't being output. Worse still, its confusing, as many things in the image are using crTrace: and so you get intermingled messages, which are hard to decipher. So I tried: Transcript cr; show: msg; flush But that didn't seem to work (which I don't understand) Eventually I did: Transcript show: msg; cr; flush And this seems to ensure things do reliably get outputted - but I'm wondering if anyone can shed light on this areas? Ideally I want to use: Transcript cr'; show: msg; flush. But having said this - I do wonder if tools like Papertrail or Betterstack won't work with this as they want a full line of text to parse into syslog format (and so they need the cr to terminate a line). Stef did mention on the Pharo list to check the Stdio class, which did mention an anomaly about doing a #sync after a #flush (which I've never heard of). So I'm just wondering what all of you do for your seaside logging? Do you use some of the modern web based server tools like other web kids seem to do? Tim p.s. While I'm not using Seaside, I've been having quite a bit of fun playing with CodeParadise, which borrows ideas from Seaside but with its WebSocket approach and direct Dom manipulation seems a bit more modern (although its a bit bleeding edge)

4 3

Seaside 3.5.6 installation error
by Rene Paul Mages (ramix) 26 Feb '24

26 Feb '24

Hello everyone, We are : - using Pharo 12.0.0 (64 bit) - trying to install Seaside 3.5.6 on a Linux-laptop with the following code : Metacello new baseline:'Seaside3'; repository: 'github://SeasideSt/Seaside:master/repository'; load The installation is broken : https://ramix.org/pharo/seaside/Capture2024-02-26.png Thanks a lot for your help. Nota Bene : If we use Pharo 11.0.0 (64 bit) the installation is completed correctly -- All the best Rene Paul Mages (ramix) GnuPG key : 0x9840A6F7 http://sites.google.com/view/les-logiciels-libres/pharo http://nosoftwarepatents.wikidot.com/smalltalk http://twitter.com/RenePaulMages

2 1

WAResponseGenerator>>#forbidden
by Norbert Schlemmer 15 Dec '23

15 Dec '23

Hello using Seaside 3.4.5 as part of VAST Platform 2023 running on Windows Server. Have added a REST API to our Smalltalk application and have an issue e.g. in case of basic auth leads to #forbidden. In this case the response is generated by WAResponseGenerator>>#forbidden as plain text. Would expect that a json response will be returned. Does anybody have a sub class of WAResponseGenerator like WAHtmlResponseGenerator for mime type JSON / "WAMimeType applicationJson" ? thanks Norbert Example curl.exe -X POST https://10.22.2.10/rest/activateMission --insecure ^ -H "accept: application/json" ^ -H "content-type: application/json" ^ -H "authorization: Basic xxx...xxx" ^ --data "{\"MissionNumber\":1,\"MissionName\":\"Mission one\"}" Error: you are forbidden to access "/rest/activateMission"

2 1

Strange behavior on IOS
by Annick Fron 26 Sep '23

26 Sep '23

Hi, I have deployed a very simple app on the web, but it runs on all browsers except on IOS. But it runs when I launch it inside another app. https://fencingfox.com/cartons/main Any idea why ? Best regards Annick

2 2

Seaside design issue
by Annick Fron 31 Aug '23

31 Aug '23

Hi, I use tables with items and use anchors for item lines to perform some particular action on an item. This action is for instance to call a specific component on that item. If I have numerous items, I am worried about the number of callbacks piling in and not being garbaged. Is it preferable to create a component per item holding only one callback ? Annick

2 3

Tab Order / Focus after replacement per Ajax
by Joachim Tuchel 23 Jun '23

23 Jun '23

Dear Seasiders, it seems this list is more or less abandoned, but I'll try anyways... My question is probably more JS related that Seaside.. I have a question regarding tab order and replaced components via Ajax. Let's assume we have a form with 3 inputs. Teh middle one has an onblur event handler that replaces the form. So far, so good. The problem is that the focus moves to the first text input element instead of the "next" one (which is not the same element as before, because we replaced all three with new ones). I tried a few things like using the tablndex attribute on elements, but found nothing that works. Does anybody have a clue what to do in this situation? I am pretty sure it is a JS only or even plan HTML solution, but cannot find it on my own... Joachim

2 3

None
by None 04 Apr '23

04 Apr '23

------- Instructions for using the "IVAP" Parser by Stephen Pair <spair(a)acm.org> "IVAP" is an acronym for "Instance Variable Acess Protecting"; thus what we have is an instance variable acess protecting parser. The IVAP Parser enables you to selectively compile message sends where instance variable accessing bytecodes would normally appear. This effectively enables you to trap all reads and writes to instance variables without needing to write special code in your class to do so. Additionally, the "inst var refs..." and "inst var defs..." functionality of the system is modified to be aware of such specially compiled methods. You can choose to enable this special compiling behavior on a class by class basis. Additionally, you can control which methods within a class use the special compiling behavior. To enable a class to trap inst var access, you first need to override #useInstVarAccessorMethods in your class (on the class side) to answer true. Then, you can override one or both of #selectorForInstVarAccess or #selectorForInstVarMutation to answer the specific messages that you want to have sent when an attempt is made to access an instance variable. See those methods for details on the signature of those methods. You do not need to override both methods; thus you can choose to trap only reads or only writes. That is it! Any methods that you compile in your class will now trap inst var reads and/or writes (it is up to you to actually implement the messages that are sent to access instance variables). You may also be more selective in which methods will actually trap reads and/or writes by overriding #normallyCompiledMethods and/or #useInstVarAccessorMethodsFor:. ------- I've seen this functionality needed in several applications including Kats and Seaside. I originally wrote this for Kats, but thought it was useful enough to stand on its own. I'm hoping that Avi might consider adopting this in Seaside so that we can move forward with a common implementation of this functionality. - Stephen

1 0

None
by None 04 Apr '23

04 Apr '23

1 0

None
by None 04 Apr '23

04 Apr '23

elements that get generated, and the S-Expr automagically closes ( </term> ) them for me. Then I just read the "@term" as an action, which would have a corresponding method, in the case of everything except the language elements for repeat and conditionals. In the case of the additional language elements, you'd just have to know (as on page 1 of the definitive Seaside guide) what those language elements are. In other words, if you're coming from HTML or probably Squeak land, and less from Lisp land, you can "see" a little better what it's doing, though that might not be what's happening under the sheets. Plus, in the first case, it's not clear that there is a repeat: statement by a simple glance. There are so many things surrounded by single quotes, you may miss the / as the loop indicator. That's basically the same complaint as: border = 0 being rewritten as: border: '0' coming from Squeak land, it's not as attractive as saying, "border: 0" With that said, it not quite sure how more powerful macros fit in the scheme. My guess is that someone hands me a HTML page that I have to translate into a Seaside #html method. Server side logic says that you need more "Umph" so you need a macro of some sort. I guess at the of the day, the less I have to type, or hopefully machine translate, the better. Notice that one thing in the above example that obscures what is going on in the S-Expr world is the (edit) and (remove) statements. For other readers, these spit out text that includes the parens. They don't really have anything to do with the S-Expr itself, just a bad selection of example. > To make something clearer: the transformation that's happening at this > level is from href->sea:id, not the other way around. That is, <a > href="@foo"> gets translated to <a sea:id="foo">. The sea:id > attribute is the canonical form, which is why I don't bother using > anything else in the s-epxr templates. That's also why it's irrelevant > what attribute you use (href/repeat/foo) because it all becomes a sea:id > once its parsed. The only purpose of the @ notation is that it lets us > sneakily use whatever little fields are provided in dialog boxes in > WYSIWYG editors to enter in the info which is actually important, ie, the > sea:id. > > Now, there is a separate "transformation" of sorts that happens much > later, when the element actually generates html, and then, indeed, an href > will be printed out... but that's a different kettle of fish. > > Making any sense? Yes, it's starting to sink in. Jim

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

seaside