Well, you just write some code to generate the HTML files you need and then hook that into some sort of CI job that pushes them and all the static files you need onto the web server.
But I get you, when you know how to use a hammer, everything is a nail. (I'm still using Allegro 4 for simple graphics stuff on PC because that's what I always used, despite it being a buggy mess even on Win7...)
Hm... interesting. But yea... this is my hammer and everything is, most definitely, a nail
If you know how to get the data (for me the sticking point would be interfacing to GitHub) and you can make it presentable dynamically, as part of the build process you could "make it presentable" in a private space instead of the public internet, then use an automated website copier to scrape the "dynamic" content and render it as static files in the public place you do want to serve it from.
Oops! I just reread this thread and realized that electrodude already said something nearly identical!
Thank you. Please continue saying it.
Wouldn't it be so much less work for Parallax if they just let people manage their own projects as individual repositories within a Github organization created to manage OBEX objects, instead of having to deal with tons of pull requests?
Git has a `git subtree` command that should make it practical to write a script to automatically split the master repository into one repository for each object.
EDIT: And you could still have a master repository that includes all the child repositories as submodules.
i would donate my time to write a new obex....but i already asked what was actually wrong with it and heard that it was too expensive....they closed the post then....so i guess keep guessing
Come to think of it, I used HTTrack for something once, and it managed to scrape many copies of the same page with only the query parts differing. This is probably why jonabel's scraped copy is so much bigger than yours.
I have found a couple posts about the old OBEX site being more user-friendly than GITHUB. I can’t say I disagree however the move seems permanent. A while ago, some genius (name is somewhere on the forums I’m sure) saved the whole OBEX in a single .zip and I was fortunately enough to nab it up before it disappeared.
It’s a little over 100 MB and it’s a simple folder-based organization as seen below. There are no frills, but here it is. 😊
The 244 MB version I made with HTTrack has the entire website - and has all the files in the right places. People keep pushing 100 MB versions - one version didn't even have any code in it.
One thing to mention also is that the old obex had lots of EXTRA files,...
like images, and zips, that users attached to the obex entry, did that information carry over also?
That is most likely how the old obex was larger, also when I crawled it with my web copier it had multiple repeats of each entry due to the parsing.
I know why they got rid of the obex, it was a SPAM magnet.
When I crawled it with my web copier, I searched the files for spam, and I found 50+ spam links.
ALL the spam was on the comments section of various objects.
They could have just disabled comments.
Disorganizing the files - makes it harder to learn things.
I am becoming disillusioned, since I am unable to get the information I need to finish my PWM Phasing project on a P2.
There isn't enough documentation to understand how to replace/convert certain P1 calls.
I am currently stuck on NOP. There is no replacement - so what do I put in it's place?
What's NOP for?
Yesterday I tried to download Jon’s jm_serial. I just wanted the related 4 files.
After half an hour I gave up. Cannot download just the files in the bottom directory!
I’m not downloading the lot as I am on a metered network, and if I go over the monthly limit it’s $10 per GB. One month a W10 update cost me $40!!!
I had tried to write a simple program to test jm_serial but pnut 34x complained “no pub routine in “jm_serial” and effectively crashed so I tried an old pnut but had the same result.
So I looked for PropTool 2.2. Another chore to find it, downloaded it. It failed. You don’t set the top object in the same way as P1. Anyway, trying to compile as top object failed so I gave up.
If beginners get these results, they will be gone never to return. I know the P2 hasn’t officially been released yet, but these are fundamental things. If you cannot get these right now, then you’re building on Smile an there’s only one possible outcome! And these fundamentals are not even being considered, let alone a plan to remedy.
So I quit work on the P2 yesterday. The fun isn’t in the P2 anymore! Everything is a chore.
The only bright side is Eric’s fastspin, but that’s not going to fix the basics.
Disorganizing the files - makes it harder to learn things.
I am becoming disillusioned, since I am unable to get the information I need to finish my PWM Phasing project on a P2.
There isn't enough documentation to understand how to replace/convert certain P1 calls.
I am currently stuck on NOP. There is no replacement - so what do I put in it's place?
What's NOP for?
I would create a new thread for help on specific coding, you will most likely get better help.
Disorganizing the files - makes it harder to learn things.
I am becoming disillusioned, since I am unable to get the information I need to finish my PWM Phasing project on a P2.
There isn't enough documentation to understand how to replace/convert certain P1 calls.
I am currently stuck on NOP. There is no replacement - so what do I put in it's place?
What's NOP for?
I am currently stuck on NOP. There is no replacement - so what do I put in it's place?
What's NOP for?
I assume you are converting P1 assembly to P2. The NOP instruction on the P1 causes the execution of one instruction without changing the flags, cog RAM, special registers or hub RAM. It does nothing but expend the number of cycles normally used by an instruction. There are lots of ways to do a NOP, such as adding zero to a cog location, moving a cog location to itself, etc. Or you could just use the NOP instruction in the P2.
To get a single file from GitHub you click on the file name to view the file, and then click on "Raw". This will show only the text of the file, and nothing else on the screen, such as line numbers. On my browser I right-click somewhere on the text of the file, and then select "save as". There are two issues with saving it this way. It tacks on a .txt extension, so instead of jm_serial.spin2 I get jm_serial.spin2.txt. I just rename the file to remove the .txt.
The other issue is that the original Windows CRLF line endings are all converted to just LF. PNut doesn't like this type of line ending, and treats the entire file as one line. You can use WordPad to restore the CRLF line endings. Just open the file with WordPad, and then save it back. That should fix it.
I looked at one of my P1 objects, and it has this same problem with the missing CR's. I believe there's a setting in GitHub that is causing all the CR's to be stripped away when the files are being pushed onto GitHub. I'm not sure if this can be fixed.
Windows Notepad couldn't handle missing CR's in the past. However, the latest version of Notepad can now handle these files. I suggest the PNut should be modified to handle them as well.
To get a single file from GitHub you click on the file name to view the file, and then click on "Raw". This will show only the text of the file, and nothing else on the screen, such as line numbers. On my browser I right-click somewhere on the text of the file, and then select "save as". There are two issues with saving it this way. It tacks on a .txt extension, so instead of jm_serial.spin2 I get jm_serial.spin2.txt. I just rename the file to remove the .txt.
The other issue is that the original Windows CRLF line endings are all converted to just LF. PNut doesn't like this type of line ending, and treats the entire file as one line. You can use WordPad to restore the CRLF line endings. Just open the file with WordPad, and then save it back. That should fix it.
I looked at one of my P1 objects, and it has this same problem with the missing CR's. I believe there's a setting in GitHub that is causing all the CR's to be stripped away when the files are being pushed onto GitHub. I'm not sure if this can be fixed.
Windows Notepad couldn't handle missing CR's in the past. However, the latest version of Notepad can now handle these files. I suggest the PNut should be modified to handle them as well.
Yes. A total simple, easy, intuitive solution for Noobs. NOT!!!
And do this for each of the 4 files in the folder!!!
Try David Zemon's frontend to the OBEX on GitHub. It is located at https://obex.zemon.name/tree . You can progress through the tree by clicking on the ">" character to the left of an item. That will expand that item with everything at the next level. When you finally get to the file or project your looking for, you can then click on the downward pointing arrow to download it.
I think for the 4 files you want you will still have to do each one individually, but David Zemon's interface makes it a little easier.
The files ending with just LF instead of CRLF is a Git feature ( ), not a GitHub problem. It's also configurable. But if one were to clone the git repository onto a Windows computer, one would be able to choose to check out the files with CRLF line endings.
Unfortunately, no such option is provided when downloading files from GitHub. You're limited to whatever was committed, which in this case appears to be LF line endings.
And on my website, I clone the repository on a Linux box, and I've no interest in changing the default setting to checkout with CRLF.
I whole heartedly agree though, that pnut should be updated to support LF line endings. That seems like an innocent oversight, and one easily fixed.
Shouldn't parallax's entire official spin library (the same one that gets installed with prop tool) be here?
Im on linux, so I don't have prop tool with its default libraries.
Does someone need to upload them still? (I guess FullDuplexSerialPlus.spin is NOT in the prop tools default libraries, i thought it was once, or at least on the obex...)
But FullDuplexSerialPlus.spin is not HOT swappable with Parallax Serial Terminal they are different.
However FullDuplexSerialPlus.spin can be hot swapped with FullDuplexSerial.spin, which is in the github library.
So many objects use FullDuplexSerialPlus.spin, would it be wise to make a folder in the github library for FullDuplexSerialPlus, and put a note there that it was replaced with FullDuplexSerial.spin.
FullDuplexSerialPlus.spin has been replaced by FullDuplexSerial.spin, i guess?
Comments
Hm... interesting. But yea... this is my hammer and everything is, most definitely, a nail
Thank you. Please continue saying it.
Wouldn't it be so much less work for Parallax if they just let people manage their own projects as individual repositories within a Github organization created to manage OBEX objects, instead of having to deal with tons of pull requests?
Git has a `git subtree` command that should make it practical to write a script to automatically split the master repository into one repository for each object.
EDIT: And you could still have a master repository that includes all the child repositories as submodules.
I did this a few months back. It's about 1.3 GB.
It's only 339 MB if you just download the original git repository with git, instead of getting an HTMLified version with HTTrack.
I also ran weekly scrapes of the old obex and provided the results here: https://ci.zemon.name/buildConfiguration/ObexDownloader_Download?guest=1
Unfortunately, the last run was April 12, 2020 - so anything added after that is lost (to me). Here's a direct link to that last run: https://ci.zemon.name/repository/download/ObexDownloader_Download/5714:id/complete_obex-2020-04-12.zip?guest=1
This ZIP is only 111 MB
Come to think of it, I used HTTrack for something once, and it managed to scrape many copies of the same page with only the query parts differing. This is probably why jonabel's scraped copy is so much bigger than yours.
It’s a little over 100 MB and it’s a simple folder-based organization as seen below. There are no frills, but here it is. 😊
Complete OBEX:
https://jdpresents.com/temp/complete_obex.zip
like images, and zips, that users attached to the obex entry, did that information carry over also?
That is most likely how the old obex was larger, also when I crawled it with my web copier it had multiple repeats of each entry due to the parsing.
I know why they got rid of the obex, it was a SPAM magnet.
When I crawled it with my web copier, I searched the files for spam, and I found 50+ spam links.
ALL the spam was on the comments section of various objects.
They could have just disabled comments.
Oh well, its now on github, on ward HO!
I am becoming disillusioned, since I am unable to get the information I need to finish my PWM Phasing project on a P2.
There isn't enough documentation to understand how to replace/convert certain P1 calls.
I am currently stuck on NOP. There is no replacement - so what do I put in it's place?
What's NOP for?
After half an hour I gave up. Cannot download just the files in the bottom directory!
I’m not downloading the lot as I am on a metered network, and if I go over the monthly limit it’s $10 per GB. One month a W10 update cost me $40!!!
I had tried to write a simple program to test jm_serial but pnut 34x complained “no pub routine in “jm_serial” and effectively crashed so I tried an old pnut but had the same result.
So I looked for PropTool 2.2. Another chore to find it, downloaded it. It failed. You don’t set the top object in the same way as P1. Anyway, trying to compile as top object failed so I gave up.
If beginners get these results, they will be gone never to return. I know the P2 hasn’t officially been released yet, but these are fundamental things. If you cannot get these right now, then you’re building on Smile an there’s only one possible outcome! And these fundamentals are not even being considered, let alone a plan to remedy.
So I quit work on the P2 yesterday. The fun isn’t in the P2 anymore! Everything is a chore.
The only bright side is Eric’s fastspin, but that’s not going to fix the basics.
I would create a new thread for help on specific coding, you will most likely get better help.
Not the correct thread to ask this question.
NOP-No Operation. Adds a wait.
Cluso99 - After 15,000 posts, I would say you are invested.
I am not convinced that you are giving up
The other issue is that the original Windows CRLF line endings are all converted to just LF. PNut doesn't like this type of line ending, and treats the entire file as one line. You can use WordPad to restore the CRLF line endings. Just open the file with WordPad, and then save it back. That should fix it.
I looked at one of my P1 objects, and it has this same problem with the missing CR's. I believe there's a setting in GitHub that is causing all the CR's to be stripped away when the files are being pushed onto GitHub. I'm not sure if this can be fixed.
Windows Notepad couldn't handle missing CR's in the past. However, the latest version of Notepad can now handle these files. I suggest the PNut should be modified to handle them as well.
And do this for each of the 4 files in the folder!!!
I think for the 4 files you want you will still have to do each one individually, but David Zemon's interface makes it a little easier.
So it's likely just another git problem
Unfortunately, no such option is provided when downloading files from GitHub. You're limited to whatever was committed, which in this case appears to be LF line endings.
And on my website, I clone the repository on a Linux box, and I've no interest in changing the default setting to checkout with CRLF.
I whole heartedly agree though, that pnut should be updated to support LF line endings. That seems like an innocent oversight, and one easily fixed.
So where is FullDuplexSerialPlus.spin?
The most commonly used spin object ever. is.... where in the github pile?
Yes, you may have a local copy. But where is the obex copy.. not in github, i guess?
I cloned the entire repository, and there is not a single folder with that name, i see variations of it, what about the original?
I figured it would be here: https://github.com/parallaxinc/propeller/tree/master/libraries/official/p1
And the only thing there is a readme.md.
Shouldn't parallax's entire official spin library (the same one that gets installed with prop tool) be here?
Im on linux, so I don't have prop tool with its default libraries.
Does someone need to upload them still? (I guess FullDuplexSerialPlus.spin is NOT in the prop tools default libraries, i thought it was once, or at least on the obex...)
https://forums.parallax.com/discussion/163963/full-duplex-serial-plus-not-available-as-a-download-any-more
I guess it was replaced with PST? Parallax Serial Terminal?
This thread says it was replaced with parallax serial terminal?
https://forums.parallax.com/discussion/164916/parallax-serial-terminal-spin-vs-fullduplexserialplus-spin-vs-fullduplexserial-spin
But FullDuplexSerialPlus.spin is not HOT swappable with Parallax Serial Terminal they are different.
However FullDuplexSerialPlus.spin can be hot swapped with FullDuplexSerial.spin, which is in the github library.
So many objects use FullDuplexSerialPlus.spin, would it be wise to make a folder in the github library for FullDuplexSerialPlus, and put a note there that it was replaced with FullDuplexSerial.spin.
FullDuplexSerialPlus.spin has been replaced by FullDuplexSerial.spin, i guess?
Ready to get confused?
This file calls for them both FDSP and PST, so I just replaced FullDuplexSerialPlus.spin with FullDuplexSerial.spin
Seems to compile ok.
https://github.com/parallaxinc/propeller/blob/master/libraries/community/p1/All/WX ESP8266 SPIN Get Started/Text_Page_To_Micro_1.spin
And neither are located inside the WX folder. Roundy Roundy Roundy goes the merry go round, BLARG.