Webbot Downloader

[Crunchy Theory]

Project Designer and Board Designer download fine, but WebbotLib doesn't start downloading after I set the download path. Is this a known issue or a bug?

I notice that the remote URL points to an index.html page instead of a file to download. I'm not sure if this was intentional, but I can't modify the path.

 I am using Mac OSX, by the way.

[Webbot]

Crunchy,

Here's the reason for the 'index.html'.
The URL points to SourceForge - ie the home page for WebbotLib on Source Forge. It then parses the content to find the URL for the 'download latest version' button.
This URL goes via several redirects - to end up with a URL for the .ZIP file on one of the Mirror download sites.
The downloader then uses that as the address to download the image from.
So it is isn't trying to download the 'index.html' its just trying to use it as the start point of the process.

Works fine on a PC but will take another look to see why its an issue on a Mac.

[Webbot]

SourceForge, where the WebbotLib releases are hosted, has been spending the last few weeks recovering from a massive attack and have only recently got most of the bits going again. They have also introduced a new Look'n'Feel. Unfortunately, for me, this is a backward step as the pages are now in 'unparsable' HTML rather than well-formed XHTML - and this is why Crunchy is having a problem.

I have logged a support ticket with SourceForge to see if this is going to change - and, if not, then I will have to resort to hosting the releases on my own server. I don't want to do that as I loose out on stats like the number of downloads (including via Webbot Downloader).

So at the moment - its broke - and I'll wait to hear what the SourceForge folk say.

[Webbot]

OK done a fix for it.
http://www.societyofrobots.com/robotforum/index.php?topic=7787.msg97991#msg97991

[Admin]

My coworker yesterday had the same problem, but I thought he was not doing something right . . .

Thanks for fixing it so fast!

[Crunchy Theory]

Still having a problem with Downloader downloading webbotlib on my machine. From the interface, it looks as though it's checking for the latest version (v1.28) and sees it, then displays a checkmark. But my filesystem doesn't contain v1.28. Are there any requirements for where Downloader needs to be or what install directories I need to use?

I noticed the Webbot Apps automatically create a default webbotlib directory in my Macintosh HD/Users/ directory, though I prefer to use my own location. Would this have something to do with it? The only thing in it is a directory called "Download" with a single file within called "launcher.xml"

Two other things: 1) It seems Downloader displays download progress per file rather than total download (most noticeable with the HTML docs) - at least I think this is why the percentage jumps around. No problem if this was intentional, but I'd argue that a total download progress percentage might be more useful. 2) When choosing a download directory for an application, Downloader does not automatically select the current directory you're in if you don't specify a directory from the list.

Both minor nitpicks, but I thought they'd be worth bringing up. Otherwise, everything works great!

(Edited: for issues after a clean install)

[Webbot]

Still having a problem with Downloader downloading webbotlib on my machine. From the interface, it looks as though it's checking for the latest version (v1.28) and sees it, then displays a checkmark. But my filesystem doesn't contain v1.28.

Try launching the downloader from a Command prompt window. Change to the directory where the WebbotDownload.jar file lives and then run:-
java -jar WebbotDownload.jar
The Command Prompt window will then display all sorts of trace messages. Maybe something in there that helps explain - especially any exceptions.

Are there any requirements for where Downloader needs to be or what install directories I need to use?

Nope. For the WebbotLib part of the download it will create a subdir ie "1.28" and will also create a "temp" subdir which should get deleted out after a successfull download. So make sure that it has got permissions to create directory folders.

I noticed the Webbot Apps automatically create a default webbotlib directory in my Macintosh HD/Users/ directory, though I prefer to use my own location. Would this have something to do with it? The only thing in it is a directory called "Download" with a single file within called "launcher.xml"

Yep the "webbotlib" folder is created in the current user home directory and is used as the default for storing projects (although I may add the ability at some stage to say what your preferred project folder is) but it is also used for storing other config settings. For example: when you code gen a project then some of the values (like where WebbotLib is loaded) are sticky - in Windows I use the registry but on Unix it creates a file in the "webbotlib" folder. As you have seen the WebbotDownload app also stores its settings in a subdir called "launcher.xml" - this contains the folders you have specified for each of the downloads. By all means open it and have a look - but don't edit and save just in case you corrupt it !!

Two other things: 1) It seems Downloader displays download progress per file rather than total download (most noticeable with the HTML docs) - at least I think this is why the percentage jumps around. No problem if this was intentional, but I'd argue that a total download progress percentage might be more useful.

It actually shows the percentage based on the number of bytes already downloaded vs the total number of bytes that it know it will have to download. Most of the apps are a single file but with the HTML docs then there are lots of individual files. It will only download the HTML files that have been changed and it doesn't know this until it has checked the timestamps of each file which it does as it goes. ie if it decides that it has to download one single HTML page it has no idea how many more it MAY need to download. The other alternative would be to have two passes: one that works out the files that have changed and a second to do the download now knowing what the total is. But this would be a lot slower. As with most progress bars - its really there to show you that its doing something!

2) When choosing a download directory for an application, Downloader does not automatically select the current directory you're in if you don't specify a directory from the list.

Since its an executable then there is no concept of 'current directory'. ie if you went to C:\ and typed in java -jar c:\foo\bar\WebbotDownload.jar then what is the current directory? Is it C:\  or C:\foo\bar?

Both minor nitpicks, but I thought they'd be worth bringing up. Otherwise, everything works great!

That's fine - keep 'em coming. But lets get your download working first!

[Crunchy Theory]

I ran the Downloader from Terminal/Command Prompt using the command you gave me and it looks like it's skipping over webbotlib. I can see it checking queued files for every app except webbotlib. Should I expect to see any printouts if Downloader is having write access issues copying webbotlib to the specified directory? There is still a check icon next to webbotlib in the GUI, so it seems to think webbotlib v1.28 is there, though it looks like it never even checks.

[Admin]

Its no longer downloading WebbotLib 1.28 for me as well . . .


In separate news, the type that will make you hit your head on the wall . . . I now need to click through about 300 'Applet Alert' windows before Webbot Download finishes it's job. Supposedly it has something to do with the Java sandbox . . .

So . . . did you change something, or did the network admins make another draconian change behind my back? And better yet, any ideas on how I can disable the Applet Alert for Webbot Download?

[Webbot]

Ok - have uploaded another version at the same place - this will work !!

In separate news, the type that will make you hit your head on the wall . . . I now need to click through about 300 'Applet Alert' windows before Webbot Download finishes it's job. Supposedly it has something to do with the Java sandbox . . .
So . . . did you change something, or did the network admins make another draconian change behind my back? And better yet, any ideas on how I can disable the Applet Alert for Webbot Download?

Nope I haven't changed anything in terms of how it works. Something screwy at your end - eg you have downloaded and are running a Java Application not an Applet (an applet runs in a web page - which Downloader doesn't).
My hunch is that Admin is only getting the problem (300 alerts) at work and not at home - ie its the same problem you had with the original JavaWebStart versions.
So Downloader gets around the cache problem - but something in your setup doesn't like the use of 'http HEAD' messages and seems to think (incorrectly) that its caused by an applet. I can't work around that - without downloading ALL files of ALL applications every time you run the downloader - which makes it pointless !

[Admin]

Webbot . . .

I just had my coworker run the previous version already on his laptop, and no Applet Alerts occurred.

Then I had him download the latest Webbot Downloader, and he got the 300 something alerts.

There's definitely something different between the two versions . . .

[Webbot]

Admin - check your personal emails. I've taken this offline for now. You seem to be the only one suffering (to my knowledge) - and if there are others then its probably the same cause. So lets sort it out 1-2-1 then I can post the result. Saves me posting all sorts of intermediate versions and confusing everyone.
Webbot

[Webbot]

Ok seemed to be a problem with jar signing. New version up there - but its only purpose is to fix issues like Admins in a corporate world.