Jim Bailie

I would think just a correct nginx config is all you need to worry about. Here's where I started from: https://gist.github.com/LostKobrakai/b895e2e0e8a2c14b4da88cc7e16cf954

Thank you! That's ok, no rush! This is a slow moving rebuild of an ancient application. We have to be careful because once images are programmatically migrated by request to the old app, the links can never change from that point. And we have to make sure it's somewhat bullet-proof because of the non-technical skill levels of the admins. And MM would be great for their workflow. However, as far as pdf files go, which I'd like to have in place soon, I did find the module "Files Rename Replace" which gets me to a place where we can start testing/demo'ing

Two months or less? 😜

@kongondo Pre-sale question: Does MM have a setting for overwriting existing files/images when updating? It seems like it would. As an example let's say I have tiger.jpg and want to update the image with lion.jpg...but keep the name tiger.jpg while displaying the new lion.jpg...does that make sense 🤨

Ugh, I knew it was this straight forward. Thank you both! @Gideon So @gebeer

I'm sure this has been explained numerous times, but I'm struggling trying to find the best way to go about this scenario: So the user creates a new page and then is able to select from a drop-down, a number of pieces of existing content items that will render in the template in the order selected. I'm assuming these pieces of content will reside in their own individual, grouped pages without an associated template. In the main template I would then loop through these selected pages and output as desired.

@Guy Incognito Your above explanation on url segments is very interesting. Can you expand any further on how you set this up? I've read through https://processwire.com/docs/front-end/how-to-use-url-segments/ , but I'm still not fully absorbing what I think is your approach. I'm sure the parent could benefit from this approach as well.

I'm sorry I don't have any silver bullets, but I've been through this on a development machine and from what I remember, it came down to permissions/config files along with netplan and how I accessed the apps via nginx.conf (i.e server_name). For development, I've always used IP aliases in /etc/network/interfaces a'la ifupdown and set my nginx server blocks to listen to such IP addresses. Netplan doesn't allow this. It says it does, but I don't feel like writing a PhD thesis to get it to work, so I removed netplan and replaced it with ifupdown. I'm not telling you to do this! I did it because I understood the problem and just wanted to use the network stack like I always have.

Do all of your user/groups/permissions line up throughout your config files (nginx/php/php-fpm)? Are these all stock apt-get installs? And the new Debian based distos are using that goofy netplan s%#t. Is that working as advertised?

I had a similar situation a while back, but this thread helped...can't remember any details except that there was a permissions issue in the app's directory structure

Yeah, coming from several applications written from scratch, with codeigniter, and to a lesser extent, wordpress, workflow with Processwire has been tough for me personally, but I am adapting. I've sort of settled into a pattern where after having deployed the initial production application, I try to do as much non-destructive work as possible on this instance. If something may be destructive, then I'll clone the codebase and hit the same database and work on this. If and when everything looks good, I'll just point the DNS to this instance. You can read up on multi-instancing. The reason for the above approach is because with the app I'm working on, the DB is too big to be slinging back and forth to a development machine. Having said all this, while I'm working on production and staging instances, I've also got a half dozen R&D instances on my dev machine than are pretty darn close in structure to production. This all probably sounds nebulous and I hope it helps. And I would certainly keep search-engine'ing things like "processwire development staging production"

Yeah, I am eager to use ProCache. After having built some fairly complex caching functionality with Nginx/Openresty/Lua for a high traffic, public facing site, it would be nice just to let the framework take care of most of the heavy lifting. Most of my work now is for internal CRUD and content management, but as soon as we get an opportunity to build something public facing with PW, we'll whip out the credit card ?

Hello - I'm going to buy profields later today or tomorrow, but does anyone know if it ships with a site sample are other examples? I've searched and it seems like usage examples are pretty few and far between. Thanks!

@horst Yeah, I just went and reread the ProFields page which I should have done earlier. I'll probably need to resurrect this thread later, but I think I have enough of plan now to start experimenting. Thanks again for the reinforcement!

@horst Url segments look like they're pretty easy to manage and a date and date-created field will serve as an index so to speak. @monchu Yes! I think a single, uniform, parse-able url segment is the best approach. I know the client wants the urls to look like what's in the initial post, but that seems hokey to me. Now here's a follow-up question: 50+ document templates with 5-40 fields each is going to add up to A LOT of fields...even if a good handful of them will be shared. Does ProFields offer a way to "consolidate" these fields without adding too many tables to the DB? It seems like it would...

Hello All, I would like to accomplish this the PW way, but again, I'm struggling conceptually and don't want to burn days experimenting. So now the people signing the checks want users to be able generate dynamic documents. There are 50-70 documents to chose from and they all will have between 5 and 40 fields...and they need to create multiple versions of each document...repeatedly. /john/2019/6/28/tpr-report/v1/ /john/2019/6/28/tpr-report/v2/ /john/2019/6/28/tpr-report/final/ /john/2019/6/28/stfu-report/v1/ /john/2019/6/30/stfu-report/v2/ /john/2019/6/30/stfu-report/final/ /mary/2019/6/29/wth-summation/final/ /mary/2019/6/29/tpr-report/v1/ What I do know is that I will need ProFields and will be purchasing that. I also know that the reports will need to be their own templates...but the dates and the versioning are the problem for me. Thank you in advance for any guidance! P.S. Would one of the blogging modules be a good fit?

@Gideon So Yeah, it looks like that will work. It also looks like Text Blocks in ProFields could accomplish this as well. But it would be nice just to have a simple a dropdown for editors to select the block. Ok, now I have a starting point. Thanks!

Hello All, Can we pick this thread up again? What if I wanted to have a dropdown box of choices linking to 30-40 static content blocks for page editors to populate in a certain region of a given template. What would be the best approach? That's my chief complaint about PW. Everything works so efficiently, that I'm always afraid I'm missing the most efficient way to do things.

@BitPoet What a terrific write-up. Thanks! You've helped validate a lot of places my mind is going for this project.

Hello, So we're wrapping up an internal application using PW. We have about 6500+ pages in so far and 14 fields. It's a pretty vanilla setup. We've got a lot of positive feedback, etc. Now this app is getting a little more notice than we previously thought and there's interest in merging another application with it. This legacy application has a well established database that will serve the new PW pages very well, but I couldn't, in good conscience, try to convert it to the "Processwire Way". So I'm curious to hear any stories or anecdotes about how others may have legacy DB's working with new PW apps. Are you using two or more db's? Have you folded existing tables into the PW db? I hope this question isn't too nebulous.

@dragan Thanks for your help! I'll look for a hook and do this as a module then.

This is sort of the essence of what I'm dealing with, but there are some legacy routing and session details that need to be dealt with. And I think I have the solution now! A vast majority of the requests will be satisfied by the PW app. Why not just put all this redirect logic into a PW 404 template?? Would you agree?

Yeah, maybe my preamble made the whole thing too nebulous. Sorry. I need to include a php script on PW's index.php prior to PW starting up. That script will run and generate an array of data. If I do not need to redirect based on that data, I would like to include that data for later use in the _init.php file. What is the best way to get this array of data from index.php to _init.php I hope that's a better description of my dilemma. Thanks!

Hello All We're a month into development of our new PW application and what a breath of fresh air it is. We're planning to roll out 3400+ pages by 6/1. But we have come to the realization that we have to carry some old baggage forward. For the near future I need to pull some data in before I even start loading Processwire in the index.php file. This has to be done at this point due to possibility of 301/302'ing the request. Question: How can I EFFICIENTLY place an array of data, derived before bootstrapping, into the bowels of PW and use it later when needed?

Finally back in the saddle: @horst Thanks for these links! Even though it's in the new skyscraper code, I hadn't even noticed the region method initially, so these are valuable links for anyone diving into pw/ajax in the future. @neosin Thanks! Yeah, you're exactly right. Horst also pointed this out earlier in the thread...but it was in German