Over the last week, someone posted a link to a complete OBEX ZIP file. I should have bookmarked it, but I got distracted. Anyone know of which I speak?
Thanks David. I thought it was you, just couldn't remember the name of the download file. My only complaint with the current OBEX is that the author's name no longer appears. I used to use the name to search files.
I downloaded complete_obex-2019-01-20.zip. My WinZip cannot open it.
I attached a copy of the error message.
That's very interesting. Were you able to extract the contents and just not that one file? Or were you not able to access anything?
Right now, the process just zips up the whole folder, but if I really need to I can walk through each file in the folder and rename anything with a colon (I'm guessing that's the problem) prior to adding it to the archive.
I pushed an update to the script that uses less parallelism. This resulted in a completely error-free run and a zip file that is 109 MB instead of 107 MB. Now it is the "complete" obex
Are you thinking in the folder name? Such as changing "1-wire Routines" to "1-wire Routines - Joe Schmoe"? Or maybe a text file inside the directory, named "author" or "author.txt" with the author's name as shown in the obex? Seems like if you have to open a file, you might as well look at the source code. But adding anything else onto the folder names may start getting a might excessive. Find for "1-wire Routines" but imagine adding a name to what is already crazy long: "Driver for AMD Am9511A Arithmetic processor (vintage 8-bit floating point coprocessor)" or "I2C PASM driver, I2C SPIN driver, I2C slave driver (PASM), lots of demos."
I'm open to ideas... not sure I like either of the above ideas, but they're the ones that popped into my head first.
I could also create a new directory level:
Root
- Joe Schmoe
- 1-wire Routines
- 2-wire Routines
- Joanne Doe
- 3-wire Routines
- 4-wire Routines
But that limits your ability to find objects if you don't know the author already. But maybe you do already know the author? It is only intended to be a backup, which means you probably know what you're looking for if you're using this?
No, please not use the author first, renaming the directory and adding the author is way better
Because if I need to look for 1-wire I certainly would prefer something made by @JonnyMac , but he can't do all of it and sometimes there are gems where I still do not know the name of the author, like JDserial for example.
Since the object title shows the title the author did chose it makes the most sense to use that first.
Are you thinking in the folder name? Such as changing "1-wire Routines" to "1-wire Routines - Joe Schmoe"? Or maybe a text file inside the directory, named "author" or "author.txt" with the author's name as shown in the obex? Seems like if you have to open a file, you might as well look at the source code. But adding anything else onto the folder names may start getting a might excessive. Find for "1-wire Routines" but imagine adding a name to what is already crazy long: "Driver for AMD Am9511A Arithmetic processor (vintage 8-bit floating point coprocessor)" or "I2C PASM driver, I2C SPIN driver, I2C slave driver (PASM), lots of demos."
I'm open to ideas... not sure I like either of the above ideas, but they're the ones that popped into my head first.
I could also create a new directory level:
Root
- Joe Schmoe
- 1-wire Routines
- 2-wire Routines
- Joanne Doe
- 3-wire Routines
- 4-wire Routines
But that limits your ability to find objects if you don't know the author already. But maybe you do already know the author? It is only intended to be a backup, which means you probably know what you're looking for if you're using this?
David,
What you have now is great. Don't change anything. A person can go to OBEX and sort by author name now that the author column has been put back in place.
@jonabel1971, glad you found this thread. I saw your other two posts about trying to find somewhere to post your copy of the obex. I can host it on my server, if you'd like, but I'd first ask: given that I've already got a "complete" copy of the obex in only 110mb... Do you still want to bother with yours?
I included my name and email in the comments. My object may require calibration. If any part of it is unclear I will know where to edit the documentation.
@jonabel1971, glad you found this thread. I saw your other two posts about trying to find somewhere to post your copy of the obex. I can host it on my server, if you'd like, but I'd first ask: given that I've already got a "complete" copy of the obex in only 110mb... Do you still want to bother with yours?
I looked at one of the zipped up OBEX files that was about 100 Megs, but all the filenames were about reading in the 10-100 bytes range. Something wasn't right. I would still like to post mine, since there continue to be complaints. Mine is tested, and acts just like the old OBEX website.
@jonabel1971, glad you found this thread. I saw your other two posts about trying to find somewhere to post your copy of the obex. I can host it on my server, if you'd like, but I'd first ask: given that I've already got a "complete" copy of the obex in only 110mb... Do you still want to bother with yours?
I looked at one of the zipped up OBEX files that was about 100 Megs, but all the filenames were about reading in the 10-100 bytes range. Something wasn't right. I would still like to post mine, since there continue to be complaints. Mine is tested, and acts just like the old OBEX website.
That seems about right - I have a local copy of the Git repository here, and the file names are generally around 10-100 bytes long. I also downloaded the zip, and its names agreed. The names seem reasonable, even though they can be rather uncomfortably long. What didn't seem right to you?
Classic Obex files generated by the PropTool archive function are automatically named "{projectName} - Archive [Date yyyy.mm.dd Time hh.mm].zip" so minimum of 45 characters plus whatever the creator named the project.
@jonabel1971, glad you found this thread. I saw your other two posts about trying to find somewhere to post your copy of the obex. I can host it on my server, if you'd like, but I'd first ask: given that I've already got a "complete" copy of the obex in only 110mb... Do you still want to bother with yours?
I looked at one of the zipped up OBEX files that was about 100 Megs, but all the filenames were about reading in the 10-100 bytes range. Something wasn't right. I would still like to post mine, since there continue to be complaints. Mine is tested, and acts just like the old OBEX website.
That seems about right - I have a local copy of the Git repository here, and the file names are generally around 10-100 bytes long. I also downloaded the zip, and its names agreed. The names seem reasonable, even though they can be rather uncomfortably long. What didn't seem right to you?
I don't see any .spin or .zip files. Even if I rename them - they still aren't the size of the files on the OBEX site - and they aren't organized well. If organized using an html index file - it looks and feels like the old OBEX. Can I send it somewhere - or not?
Why don't you upload it to Google Drive, Dropbox, OneDrive, or some other such cloud service. Ideally, you would just leave it there for others to access indefinitely, but if you don't want to do that, I will download it and put it under https://david.zemon.name/downloads/ for you
@jonabel1971, glad you found this thread. I saw your other two posts about trying to find somewhere to post your copy of the obex. I can host it on my server, if you'd like, but I'd first ask: given that I've already got a "complete" copy of the obex in only 110mb... Do you still want to bother with yours?
I looked at one of the zipped up OBEX files that was about 100 Megs, but all the filenames were about reading in the 10-100 bytes range. Something wasn't right. I would still like to post mine, since there continue to be complaints. Mine is tested, and acts just like the old OBEX website.
That seems about right - I have a local copy of the Git repository here, and the file names are generally around 10-100 bytes long. I also downloaded the zip, and its names agreed. The names seem reasonable, even though they can be rather uncomfortably long. What didn't seem right to you?
I don't see any .spin or .zip files. Even if I rename them - they still aren't the size of the files on the OBEX site - and they aren't organized well. If organized using an html index file - it looks and feels like the old OBEX. Can I send it somewhere - or not?
Oh, I think I see your problem. The actual spin files are under the "All" directory, not the individual categories - did you try looking there? This is kind of a pain. The individual categories contain symlinks to the real objects in the Git repo, but Github probably doesn't know what to do with them when it makes the zip file, since Windows doesn't support symlinks.
But if you want something organized as HTML and not an offline archive, why don't you just use the new frontend site David Zemon is developing?
Why not use an unfinished frontend - when I have a finished solution? Because I keep seeing the same problems get brought up (i.e. "It's kind of a pain"). And, I have a fully zipped OBEX directory that doesn't require any adjustments to make it look right.
So, make a decision. Do you want to help people get over their complaints - or keep acting like customer service? Where do I put my .zip file so people can access it?
Why don't you upload it to Google Drive, Dropbox, OneDrive, or some other such cloud service. Ideally, you would just leave it there for others to access indefinitely, but if you don't want to do that, I will download it and put it under https://david.zemon.name/downloads/ for you
Comments
Very Cool!
Why on earth did Parallax remove authors' names? Did somebody seriously think authors no longer wanted to be credited?
I have an email going to Parallax to ask that they include the author's name.
Yep... Someone screwed up. I cannot imagine Parallax wanting that.
I downloaded complete_obex-2019-01-20.zip. My WinZip cannot open it.
I attached a copy of the error message.
That's very interesting. Were you able to extract the contents and just not that one file? Or were you not able to access anything?
Right now, the process just zips up the whole folder, but if I really need to I can walk through each file in the folder and rename anything with a colon (I'm guessing that's the problem) prior to adding it to the archive.
@ DavidZemon. I would like to see the next ZIP file as Parallax has reinstituted the authors in the OBEX.
UnZips perfectly. Is it possible in the future to have author names listed next to the programs?
I'm open to ideas... not sure I like either of the above ideas, but they're the ones that popped into my head first.
I could also create a new directory level:
But that limits your ability to find objects if you don't know the author already. But maybe you do already know the author? It is only intended to be a backup, which means you probably know what you're looking for if you're using this?
Because if I need to look for 1-wire I certainly would prefer something made by @JonnyMac
Since the object title shows the title the author did chose it makes the most sense to use that first.
Enjoy!
Mike
David,
What you have now is great. Don't change anything. A person can go to OBEX and sort by author name now that the author column has been put back in place.
Thanks for all your work!
OBEX is not open-sourced anymore - if it has been made functionally inaccessible.
I looked at one of the zipped up OBEX files that was about 100 Megs, but all the filenames were about reading in the 10-100 bytes range. Something wasn't right. I would still like to post mine, since there continue to be complaints. Mine is tested, and acts just like the old OBEX website.
That seems about right - I have a local copy of the Git repository here, and the file names are generally around 10-100 bytes long. I also downloaded the zip, and its names agreed. The names seem reasonable, even though they can be rather uncomfortably long. What didn't seem right to you?
I don't see any .spin or .zip files. Even if I rename them - they still aren't the size of the files on the OBEX site - and they aren't organized well. If organized using an html index file - it looks and feels like the old OBEX. Can I send it somewhere - or not?
Oh, I think I see your problem. The actual spin files are under the "All" directory, not the individual categories - did you try looking there? This is kind of a pain. The individual categories contain symlinks to the real objects in the Git repo, but Github probably doesn't know what to do with them when it makes the zip file, since Windows doesn't support symlinks.
But if you want something organized as HTML and not an offline archive, why don't you just use the new frontend site David Zemon is developing?
Why not use an unfinished frontend - when I have a finished solution? Because I keep seeing the same problems get brought up (i.e. "It's kind of a pain"). And, I have a fully zipped OBEX directory that doesn't require any adjustments to make it look right.
So, make a decision. Do you want to help people get over their complaints - or keep acting like customer service? Where do I put my .zip file so people can access it?
@jonabel1971
I don't know how to send you a 256 MB file.