celfred

It returns 'NULL'. How does that sound ?

Well, I've checked my localhost version and it gives 5.7.19. Anyway, I've changed my line taking the 'sort' out of my 'find' string and it works (actually, I believe my 'sort' command is not really useful on my page any longer so it's not really an issue for me any longer). I've taken care of most of my errors (I think) and it seems to work fine. I went ahead and made the changes on my live site. It looks ok. I'll test tomorrow when actually saving data 'for real', but on my localhost it works, so that's cool ! I'm running on PW 3.0.36 now Thanks a lot for your help ! If anything still matters to you and you suspect a weird behavior on PW side and you want to dig this a little, don't hesitate asking me for further testing or information.

Hello, I'm back to it and it's not easy 203 PDO queries in my Tracydebugger ! That sounds like a lot to me... I have tried to isolate the 'responsible' one and here it is (I think) : SELECT pages.id,pages.parent_id,pages.templates_id FROM `pages` JOIN field_team AS field_team ON field_team.pages_id=pages.id AND ((((field_team.data='85366') ) )) LEFT JOIN field_group AS _sort_group ON _sort_group.pages_id=pages.id LEFT JOIN pages AS _sort_page_group_name ON _sort_group.data=_sort_page_group_name.id WHERE (pages.templates_id=43) AND (pages.status<1024) GROUP BY pages.id ORDER BY _sort_page_group_name.name This corresponds to the following statement : $allPlayers = $pages->find("template=player, team=$team, sort=group"); Answering your question : the 'group' field is indeed a 'Page' type. It used to be a 'Text' type so I tried and changed my find query replacing 'sort=group' by 'sort=group.name' and the error is still there. Here's the resulting PDO query in Tracydebugger : SELECT pages.id,pages.parent_id,pages.templates_id FROM `pages` JOIN field_team AS field_team ON field_team.pages_id=pages.id AND ((((field_team.data='85366') ) )) LEFT JOIN field_group AS _sort_group_name ON _sort_group_name.pages_id=pages.id LEFT JOIN pages AS _sort_page_group_name ON _sort_group_name.data=_sort_page_group_name.id WHERE (pages.templates_id=43) AND (pages.status<1024) GROUP BY pages.id ORDER BY _sort_page_group_name.name As I said earlier, if I change the syntax to : $allPlayers = $pages->find("template=player, team=$team")->sort("group.name"); The PDO query shows : SELECT pages.id,pages.parent_id,pages.templates_id FROM `pages` JOIN field_team AS field_team ON field_team.pages_id=pages.id AND ((((field_team.data='85366') ) )) WHERE (pages.templates_id=43) AND (pages.status<1024) GROUP BY pages.id and everything works as expected. I've tried my best to be helpful. I hope I've managed to isolate the correct lines and that you can understand something out of it. (and that I'm not responsible for taking this time from you... by responsible, I mean that I did something wrong that caused the problem, which can be a possibility) Sidenote : Trying to isolate the query, I took almost everything out of my page (no more find/get statement except that one, no more head.inc, foot.inc....) and I still have either 41 or 38 (depending on reload) PDO queries on my page). I would imagine that's PW 'doing his job'. Is that so ?

I have the TracyDebugger installed and I'll try to do what you're asking me with pleasure, but I have to stop right now (time to eat at home, daughters hungry... ) I'll do that tomorrow and keep you updated about my progress. And I agree with your advice about installing modules directly from within ProcessWire. It is indeed quite simple

Last thing for the moment while trying to update to PW 3.x : in backend, my modulemanager page tries to read a modulemanager.cache file in files/assets/cache which doesn't exist hence error messages : Warning: fread(): Length parameter must be greater than 0 in /home/celfred/PlanetAlert/site/modules/ModulesManager/ModulesManager.module on line 812 ProcessWire: ModulesManager: cannot read cache file /home/celfred/PlanetAlert/site/assets/cache/ModuleManager.cache (with debug true)

Here's what I did : Restarted mysql to be back with the initial fatal error. Found out that commenting out the line : $allPlayers = $pages->find("template=player, team=$team, sort=group"); got rid of the error Replaced the line with : $allPlayers = $pages->find("template=player, team=$team")->sort("group"); and... It works That's indeed a cleaner solution better than changing my mysql settings. I guess the 'bug' was on my side with the 'sort' command included in my 'find' query. I have to revise the PW doc to see where I saw that syntax or if I made it up without noticing that it never worked (which is possible !). So thanks for pointing this out to me !

EDIT : As a follow-up, still digging on my updating issues : I have found that post : https://stackoverflow.com/questions/23921117/disable-only-full-group-by#36033983 Along with the namespace line on my particular page, it works EDIT #2 : Sorry ! I validated another post instead of editing the first one.

I must admit I'm completely lost on those issues. I have turned off the compiler on my first fatal error that occurred with another module : Pages2Pdf. As I said earlier, I downloaded the last Github version of that module and noticed in the source the FileCompiler=0 line, so I simply did the same in the Service Pages modules Now when trying to get rid of all lines (FileCompiler and namespace) it seems to work as well (?) . Maybe I'm still having a cache somewhere messing up with all my tests, but it seems to work anyway so that's good ! I am now continuing debugging my local site. I have to add the namespace line at some places to get rid of error messages. But all in all, it looks quite correct. One of my pages trigger a 'big' fatal error that is again over my head for the moment : Error: Exception: SQLSTATE[42000]: Syntax error or access violation: 1055 Expression #1 of ORDER BY clause is not in GROUP BY clause and contains nonaggregated column 'planetalert._sort_page_group_name.name' which is not functionally dependent on columns in GROUP BY clause; this is incompatible with sql_mode=only_full_group_by (in /home/celfred/PlanetAlert/wire/core/PageFinder.php line 384) (The page worked fine with PW 2.8). But that's another problem and it's not on a 'major' page Thanks for your concern @adrian

So cool ! I can't believe it was so... simple... I added this line on my templates files, but didn't think it was needed in modules as well. Now, I'll try and face the other issues by myself, but it went a little further. That's cool ! Thanks a lot for your help !

Hello, I'm desperately trying to update my website to PW 3.0.62 and I'm facing issues to to module compatibility. I was stuck with Pages2Pdf which I managed to solve bu updating the module from Github, but now it's the Pages Web Service module... and this time, I don't know wht to do The Module is found there. But it is quite old and I can't find it in the modules catalogue... and my site is making quite a use of it (I can't think of a way to do otherwise, sorry...) After adding the FileCompiler=0 to the module pages, the error I'm stuck with is : Fatal error: Class 'WireData' not found in /home/celfred/PlanetAlert/site/modules/ServicePages/ServicePages.module on line 22 and I have no idea on what to do... I must admit I'm not a programmer but a middle-school teacher... (for your information, here's the site I'm talking about : http://planetalert.tuxfamily.org ) but I'm struggling hard to solve the different issues I have to face and I'm wlling to understand things. I have just spent many hours trying to make 2.8 work on my localhost (and it seems ok ) but I'd like to switch to 3.x to prepare the future If anyone had the will to spend a few minutes to try and help me, I would greatly appreciate. Thanks in advance ! If you need more information to understand my problem, feel free to ask.

I feel a little dumb once again... Your explanation is perfect and I guess I want the 2nd one. I think I was mistaken on that 'not()' filter. I'll have to check everywhere I've used it. The 'further filtered' was the key in your explanation. Thanks !

The $prevEvents->count() is fine. This part works as expected Your post reminded me of trying to cut down through my page to see where the ALL pages request was (ou could be), and reducing things little by little, I saw my checkStreak() function in which I look over the history of each player ti find if the last 10 consecutive events are positive, in which case the player becomes an 'AMbassador' and here was my code : $lastEvents = $player->get("name=history")->find("template=event, sort=-date")->not("task.name=donated|donation|absent, limit=10"); I bet you can see right through it : the 'limit=10' is misplaced and useless where it is !!! The code should be (I think) : $lastEvents = $player->get("name=history")->find("template=event, sort=-date, limit=10")->not("task.name=donated|donation|absent"); Now my 'loaded page' is reduced to about 480. What a gain ! And this checkStreak() function is called each time I save a new event for a player (to check if he needs to become an ambassador...) So I'll keep digging because I can imagine I have the same sort of mistake at other places, but I am already glad of this improvement. So thanks a lot !

Thanks @kongondo for the 1st tip. I get the difference between children('-date')->first() and child('-date'). Although I've just tried to change this line and nothing changed in my Tracy 'Processwire Debug' panel... still many many loaded pages... I'll dig that later when I have more time. Maybe I have the same kind of error somewhere else... About the date>=$prevDay : this solved an issue I had because of the time appended to the date. Actually in my code, I take the last event, get the date, change the time to 0:0:0 and eventually get the >= to find all those events having the same day, but with a superior time. Maybe (I'm sure actually...) there's a better solution. I remember at the time I implemented this, I was surprised not to be able to retrieve the same day events and this was the only workaround I found

Hello, I've been struggling ALL day on my loading time problem and I'm stuck... As I said somewhere else on the Forum, I'm discovering TracyDebugger module to try and help me (and it does help !), but I'm still in a dead-end. Tracy told me one of my page loads about 4,500 pages and I can't find any way to make it less (and let me tell you that my website is far from having so many users/visitors or whatever, it's just a teacher's website to try to work differently in class). I have a feeling I am mis-understanding some concepts here. The page can be seen there : http://planetalert.tuxfamily.org/players/4d (for example, but the problem is the same for all teams in my site). As you can see, there are many information on this page, but not that much either I guess for a computer, but it takes 20 seconds to load ! It feels like my page loads just about EVERYTHING that resides in the backend, but I just can't understand by looking a tmy code. For example, the little green and red circles next to the player's names show the actions that have been saved on the last day (so usually 3-5 actions at most, except for players doing online exercises who may have 10 actions). My query is : $lastEvent = $player->child("name=history")->children("sort=-date")->first(); // Get all events on same day $prevDay = date("m/d/Y", $lastEvent->date); $prevDate = $prevDay.' 0:0:0'; // Select events for the whole day $prevEvents = $player->child("name='history'")->children("date>=$prevDate"); $trend = ''; foreach ($prevEvents as $event) { $HP = $event->task->HP; $title = $event->task->title; if ($HP < 0) { $trendClass = 'negativeTrend'; } else { $trendClass = 'positiveTrend'; } if ($event->summary !== '') { $summary = ' ('.$event->summary.')'; } else { $summary = ''; } $trend .= '<span class="'.$trendClass.'" data-toggle="tooltip" data-html="true" title="'.strftime("%d/%m", $event->date).': '.$title.$summary.'">&nbsp;</span>'; } But in Tracy pages, I see that the COMPLETE history for EACH player is loaded ! I thought my code was limiting it to the events appearing on the same day as the last saved event. Do you see where I am wrong ? What's more : I feel like I must do the same mistake for the players (they are all in althought I'm limiting to the particular team). I must say though, that I'm caching (supposively) ALL players when Admin is logged in so he can have a faster access in the top menu. More? All places and people that players can free are loaded as well. Why is this? I just want to load the places and people that are actually freed by players and show the names (not complete info). I'm not putting an example of the code here because I have a feeling the problem resides in my way of organising the complete thing... and that's what worries me the most... Maybe it's gonna be hard for one of you to understand (and have the time) my awkward way of organizing things (even though I'm trying hard to be well-organized ). So I'd rather point to the Github repository : https://github.com/celfred/PlanetAlertProfile/tree/master/site/templates The file I'm talking about above is team.inc.php. It is included in list-all.php which includes head.inc and foot.inc as well. Things have changed quite a bit on my local site today, but I'm getting in such a mess that I'd rather not commit to remote. I've corrected some things, but what I have just explained above is still there While I'm at it. Other page, other problem detected : my adminTable.php page which you can a look at at : http://planetalert.tuxfamily.org/players/adminTable/4d (I've just noticed that it should be acess restricted, but nevermind for today, don't submit just in case I forgot other things... _ yes, I'm desperate tonight...) and which submits to submitForms.php. Well, I had to put AJAX request for that because it was running into a timeout because of the loading time. It worked (althought I ended up waiting 30 seconds to record my table !) but today, Tracy and debug ON showed me the 'Maximum 100 nested functions limit fatal error' after hours of research, I endend up... writing this post I know my function updateScore() (in my-functions.php) is difficult, but what?? It just updates the scores depending on the task being recorded then checks the impact on other players and if it does have an impact, it triggers itself back to update the score of that other player. But yes, I know it can be some sort of never-ending story, but how can I do differently ? I wonder if it's ok to have wire('pages') in so many of my functions. But again, how can I do in another way? Well, sorry for such a long post, folks, but as you can understand, I'm facing a wall here and I hope someone around here will find some time to try and find his or her way into my code and give me a hint on where I could improve. I have had so many help from this community in the past and it felt such a relief and a revelation (Ended up telling my wife : "Wow, it looks so simple for these guys ! I've spent so many hours on that and... Of course ! Now I get it !! I'm getting better ! No next time !") Well, today is a 'next time'... Thanks if you've read up to this point and THANKS if you go and have a look at my code ! Fred.

Just as a follow-up in case people wonder, here's a copy of the message I sent to @adrian

@adrian Thanks for your reply. My site is at http://planetalert.tuxfamily.org . I have just installed TracyDenugger there. It doesn't appear but from what I've read in the settings, that's normal, everything is stored in log files. Ok, but how do I give you access to useful things (admin access?). Sorry if you feel like I'm messing things up and being a nuisance... Actually, I've just tested 1 thing : forcing the DEV mode so I can see Tracy's bar. I saw the 'Processwire Debug Panel' and clicked on the 'Pages loaded' link, and it went to my 404 page like I said... I turned the setting back to DETECT. Feel free to tell me if you want me to do some things or to private message you some logs or whatever... Edit : PM sent. Thank you so much.

@kongondo : I was definitely thinking of starting a new topic and ask you, guys, about my optimisation issue, but I wanted to find some clues by myself first... and believe me, I am trying to find something (hence my discovering of TracyDebugger ) but at the moment, I stay also curious about why I have to load 5000 pages ! (Except I am the creator of these pages, so this is rather shameful...) So I'll start a new topic soon ! But I still think the expected behavior of my TracyDebugger shouldn't be to redirect me when I click on the 'Pages loaded' link, no? Edit : WHile I'm at it, the 'Show bar in backend' option is NOT ticked and the bar displays in backend anayway And it triggers the 'Fatal Error Maximum function nesting level of '100' reached, aborting!'... Do you think I did something wrong? Should I try and re-install the module?

Hi everyone - Absolute newbie question (I'm discovering the module and actually I'm also discovering PHP debug stuff, so sorry if my question is rather stupid) I've just installed the module and it loads fine on my localhost with the bottom bar. If I look at the 'Processwire Debug Mode' (which I've set ON in my config.php), I see many infos such as 'Pages loaded', 'API Variables', 'Session'... with numbers. I'm very interested since I have a page which loads 4991 pages and so takes quite a long time (and I'm desperately trying to optimize it), BUT, and here's my problem : if I click on the link to see the pages, the menu drops down (cool!) but it triggers a page load which in its turn redirects immediately to my 404 error page... and I lose the current information I am spending quite a while trying to understand how to dive into this PHP debug tool (I have to admit it's for the moment way over my coder's skills), but here, I'm a little stuck on what seems to be... not much. So if anyone can give me hint on that one, I'd appreciate. Thanks ! PS : Maybe I should have opened a new question in the forum rather than following this topic ?

Hello, I'm facing this issue at the moment... A PHP script that seems to take too long and the browser returns a timeout. Flushing works fine on my localhost but doesn't on remote I have no idea what I can do, so if @renobird or anybody else has some advice, I would appreciate Thanks !

Thanks for the replies ! I had tried the OR statement before (like Robin S mentionned), but I had an issue because of those players having 2 places and 1 people and not being concerned (or 1 place and 2 people), but Kixe showed me the way by adding all possibilities in the OR statement. I hadn't thought about this possibility. Anyway, that would be more difficult if I wanted to pick players having more than a cumulated 10 number of free elements. I guess I would have to use my first 'foreach' solution in such a case. Thanks again for your ideas !

Hello, I'm having a little question with this part of code : $rank = $pages->get("template=team, name=$selectedTeam")->rank->name; if ( $rank == '4emes' || $rank == '3emes' ) { $allPlayers = $pages->find("template=player, team.name=$selectedTeam"); $allConcerned = new pageArray(); $notConcerned = new pageArray(); foreach($allPlayers as $p) { // Find players having at least 3 free elements if ($p->places->count()+$p->people->count() >= 3) { $allConcerned->add($p); } else { $notConcerned->add($p); } } $notConcerned = $notConcerned->implode(', ', '{title}'); /* $allConcerned = $pages->find("template=player, team.name=$selectedTeam, (people.count+places.count>=3)"); // Find players having at least 3 places OR 3 people */ /* $notConcerned = $pages->find("template=player, team.name=$selectedTeam, (places.count+people.count<3)")->implode(', ', '{title}'); */ It works fine the way it is, but I just want to make sure that the commented part is not possible with PW (yet?) : having a selector such as people.count+places.count>=3. I feel like it would be more concise and more readbable/maintainable (2 lines instead of 11). Thanks in advance for your insight and advice

Thanks for the answer. HP field can be less than zero. It is an 'Integer' field that holds a negative value when the task is... negative so I'm not sure why it doesn't return the expected result. I like the 'break;' tip. Again, I have a feeling I have missed that 'basic' one in my slow discovery of PHP (and coding)... Sorry about that. Anyway, I'm not at home right now, but I'll do some testing and come back here to edit my post and try and understand something about this nextUntil(). Thanks again ! Edit : After testing further and struggling about half an hour, here's my mistake : prevUntil() was what I was supposed to use (I feel a little bit ashamed...) Yes. I guess my 'sort=-date' in my previous statement was the clue. What I imagined being my first event (hence my nextUntil) was in fact my last event, that's why prevUntil was to be used... Eventually, the most readable code (to me) is your 'foreach' with the 'break' instruction, so I'll go with that Thanks !

Hello, I'm stuck with a nextUntil() use I can't figure out. The following function works if I use the commented part instead of the last 2 lines (with nextUntil). function checkStreak($player) { // Set Ambassador skill if necessary $ok = 1; $streak = 0; /* $lastEvents = $player->get("name=history")->find("template=event, sort=-date")->not("task.name=donated|donation|absent, limit=10"); */ /* foreach ($lastEvents as $e) { */ /* if ($ok == 1) { */ /* if ($e->task->HP >= 0) { */ /* $streak++; */ /* } else { */ /* $ok = 0; */ /* } */ /* } */ /* } */ $lastEvent = $player->get("name=history")->children("template=event, sort=-date")->not("task.name=donated|donation|absent")->first(); $streak = $lastEvent->nextUntil("task.HP<0")->count(); return $streak; } But I know it's ugly and... I sometimes wish I had a tutor to help me improve my coding style In words, what I'm doing is : - Check the last events recorded for a 'player' - I want to check the number of positive events in a row (before a negative occurs) - My function gets the last 10 events (the limit to get a bonus) and checks each one, but stops if it encounters a negative event and return the streak. BUT : looking at the API cheatsheet (thanks @soma), I've noticed this nextUntil() which I would like to use... but my use always returns 0 If someone can understand why and help me improve on that one, I would appreciate. Thanks ! By the way, feel free (if you have a few minutes) to give me advice on how to improve those foreach() I feel like I'm overusing...

Hello, I'm struggling with this : a 504 Gateway Timeout when generating my PDF. I used to be able to generate a particular PDF listing 52 places with a photo (small), a little text and a few pieces of information for each in a table. It worked and I had a PDF of about 350Ko. Now, I'm trying to generate a PDF of about 250Kos and I get the 504 error. I've tried to re-generate my places catalogue and again, I have the 504 error. All my templates work fine (although it takes a little while) to generate smaller PDF files. The page I'm trying to generate is : http://planetalert.tuxfamily.org/players/4e/amandine If someone have advice on how I could find my way out of this, I'd greatly appreciate Thanks !

Thank you so much ! I can't believe I've passed through this one... It caused me a real headache having to manually repair the bad work with my 'delete' function. Now I've learned something again tonight thanks to both of you and I'm still enjoying more and more discovering ProcessWire Thanks again !