Ok, I know this is an old thread, but thought it might help someone if I come back on this issue!

I have found since then, one of the cause of the issue.  For us, it was the "html tips"!   Since I think version 13, the default parameter for 'TC' was changed, so that it automatically shows tips as html.  I=On our main menu, we had a bunch of small buttons that had a decription containing the symbol "<" which was detected as html markup.  SO, everytime we would mouse over, it would open a chromium object and at some point, would generate a "memory access violation".  I'm not sure if this was resolved in the latest version or not though.  We are still only running version 19...

I must say, that error came back recently and not sure the cause yet.  I'm suspecting the use of an "animated gif" for a loader.  That is the only recent change we did that would explain this.  I have removed that gif in most programs and the error is less frequent since.

If you want to keep your data in a string, I think you should PAD your string to the left before saving to your database if that field is part of a key. 

So in your database, the data would be something like:

"       300"
"       315"
"      3100"

I'm not sure what these numbers are, but if they are a unique no, like a customer number for example, we would pad to the left with zeros instead of blanks.

Perhaps there's a better way, but this is how we do it.

Good morning everyone,

I`m curious to know... Is there a way to just check if a subprogram and entry line exist, but not execute or load that program.

Just to explain my context, I want to check if the subprogram and entry line exist.  If it does, execute something and then call that subprogram and entry line.

I'm working on improving our grid loading routine to load faster though proplist and propvalue.  But in some case, I want to make sure the grid is loaded before exexuting a specific subroutine, but only if that subroutine exists since that subroutine may try to alter the grid.  That subroutine is in another program.

I could've use the LNO function, but it only works with the current loaded program.

Thanks again for all the comments.  It looks I have some homework to do!  Just to be clear, we've been using Eclipse IDE since the plugin came out like 15 years ago and we've never used Nomads except for database/file management.  So yes, I feel that most of PVX effort in the last few years were to develop Nomads and Webster+ and I may be missing on these.  I've played a bit with them, but not enough to see the real potential I guess! 
I think that PVX lacks of tutorials and that's what makes it hard to learn.  All popular languages have tons of info on the web and everyone I know learns languages by following tutorials and/or examples.  I mean, I could watch a 15 minutes youtube video and could learn the basics of any language probably.

Some have mentioned hiring somebody coming from BBX... How old was that person?  I mean, I would be surprise someone less than 60 years old knows BBX?!  My old boss, the original founder of our company, came from a BBX background and discovered Providex in early 2000s. He's well over 60 years old though!

We have used Java for Android mobile apps in the past, but we have now switched to Flutter which isn't perfect, but pretty good for what we do (and most importantly real cross plateform compatible).  But, my concern with a technology like Flutter is that it evolves too much.  There's frequent updates and a lot of obsolescence.  I'm not sure how many times we had issues after an update because of versions compatibility.  We tried React Native prior and it was way worst!   That's one thing I really like about PVX.  Our code is rarely broken because of an update.  That makes it easy to maintain and keep a stable software with low maintenance.

Again, thanks everyone!

Thanks everyone for the responses! 

Thanks for your response Rob!

Well, there's some flaws in our current structure where the main software only works through Windx.  I like Windx, but in Today's world, many prefer having the option to run the software from any OS and any type of devices.  I also find it more difficult to build GUI interface that adapts perfectly for all the different types of screen resolutions and screen sizes.  There's also a performance "issue" where I find that a more asynchonous approach would help with performance and change the way the interface is built and navigation.

Obviously, the "way it looks" is now a big sale point that gives a good "first impression" to future customers.  Even if the software performs better than competition, customers that do not know us are tempted to choose the most "modern looking" option.

Also, it's hard to seduce young developers to join the team since PVXPLUS has an old school approach that's really different to all the other popular "web based" technologies.  There's not a lot of documentation or info about the language that makes it hard for them to learn.

Anyway, nothing is perfect and it's just a matter of bein rational about it.  The "stability", low ressources requirements, easy maintenance and security are all super good qualities about the technology.  I really like how the "old school" mindset is still relevant Today. I mean, I remember a few years ago, everbody was pushing SQL for databases and nowadays, some are bragging about NoSql like it's a new thing, but it's pretty much what PVX is providing since forever.  There's also the security question that is more than relevant Today!

Hi guys,

I am currently debating with my team on the advantage of using PVX and why we should keep it.  I do have some arguments against them who try to push me out of PVX to use a more "standard" framework, but still, I am wondering what were your thoughts on it?

We've been using PVX for almost 25 years, but with all the new frameworks available that has a lot to offer, the question is relevant.  So, how to convince them that PVX is still relevant for us?  Our main product is an ERP system built 100% in pvxplus with WindX.  Our complimentary tools (web sites and mobile app) are built in Flutter and php mainly.  Note to say, we use our "in house" GUI engine... So no Nomads... 

I like the fact that PVX stays "retro" compatible, which makes it easy to maintain.  I also like the use of WindX, for control and security mainly.  The database is grest too.  All around, the performance is great for what we do, but there is some drawbacks that makes us rethink our strategy.

Thanks Mike, the info is helpful, but unfortunately the UCP solution throws an error 46.
I've tried with prefix, without, with other like $01$, $02$, etc... Always the same...

I'm thinking it may be related to the error 29 I'm getting while trying to simply print the data to screen.


Anyway, for now, I was able to reach out to the server admin so that I can receive the data uncompressed.


thanks

Hi, I would like to report 1 issue that I have juste encountered:

Unable to read compressed data like gzip for example.  Normally, I would specify in the header "accept-encoding: none" and it would result in a non compressed response, but on this particular server, it seems they force the response in gzip no matter what I do.  So, now I have to figure out how to handle this in PVX!

Hi,

I'm wondering if there was a way I could "force" a json value to be a numeric when the Array is set to string.   I have a case where I need to send the data in a specific format for the end user (not pvx) who uses variable types.

So my data is all sent as if they were string, but I'd like one specific value to be without quotation mark, so as a numeric value.

I thought that maybe there was a prefix I could put before the value so that PVX would recognize it as a numeric for the end result of :

"value": 1,

instead of

"value": "1",

Hi,

I'm wondering if there was a way I could "force" a json value to be a numeric when the Array is set to string.   I have a case where I need to send the data in a specific format for the end user (not pvx) who uses variable types.

So my data is all sent as if they were string, but I'd like one specific value to be without quotation mark, so as a numeric value.

I thought that maybe there was a prefix I could put before the value so that PVX would recognize it as a numeric for the end result of :

"value": 1,

instead of

"value": "1",

The Sage plugin doesn't work like it should... It does, but as a bunch of bugs... My colleague tried it and had to check all the programs to make sure it was compiling ok.  I believe one big issue was that it wasn't UTF8 compatible and was detecting the program as non PVX program since it wasn't recognizing some characters like "éàûêà, etc... " that may have been used in comments line.   We write comments mainly in french, so that happens a lot!  

After it "detects" that the program ain't a PVX program, it markes it as an "archive ressource" and won't try to compile until you manually change it.

I think there was another issue where it wasn't recognizing the /* marker for comments.  I'm not 100% about that one though...  Anyway, maybe it was just a configuration issue at some point but we had lost too much time on it so just decided to revert to an older Eclipse version and use the original plugin which works perfectly fine on first try. 

We would also like to use VS since that's what we use for non pvx coding as well.

Ok, thanks... will give it a try! 

Hi, I'm wondering if someone ever attempted to merge multiple PDF files into one via pvxplus?  I have some PDFs saved on the server where my goal is to open them up on the client side at once. (not file by file).  In some cases, there could be like 25 files to open at once.

I know there's some linux tools I could use, but wondering if it's doable in PVX?

Super!  Thanks guys!