# Perspective vs Vision - Trending and other considerations

**URL:** <https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272>\
**Category:** Ignition\
**Created:** [January 14, 2021, 10:52pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272 "2021-01-14T22:52:25Z")\
**Posts on this page:** 12\
**Page:** 1

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 14, 2021, 10:52pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/1 "2021-01-14T22:52:25Z")

</div>

Hi all,

Currently negotiating with a customer who wants to upgrade and standardise their SCADA across several sites, and is looking at Ignition. We’re currently trying to decide whether Vision or Perspective is better for their application.

I’d prefer to go with Perspective, as it’s clearly the way things are going in the future, and it makes mobile implementation easier. However I’m also conscious that the Perspective module is still new-ish, and not as fully featured as the more mature Vision module.

If you were going to take the plunge and go 100% Perspective, what would be your biggest concerns regarding the current limitations of the module? One concern I have is trending - I have developed a very in-depth trend template on the Vision module, which gives huge amounts of visibility/flexibility/control, and the operators that use them rave about them. Am I going to be able to wow people with my trends in Perspective yet?

Any other thoughts?

---

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 17, 2021, 11:22pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/2 "2021-01-17T23:22:17Z")

</div>

Bump - anyone got any insight?

---

<div class="post-metadata">

**Author:** ![jlandwerlen](https://avatars.discourse-cdn.com/v4/letter/j/bc8723/32.png) [@jlandwerlen](https://forum.inductiveautomation.com/u/jlandwerlen)\
**Post date:** [January 17, 2021, 11:57pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/3 "2021-01-17T23:57:22Z")

</div>

My two pennies, graphics should be considered in perspective, as last I checked you still can’t draw a line. There are some symbols, but graphics is still limited. I think users have found ways to overcome, but I would approach cautiously. Also, popups have issues. Not sure when IA will have fixes for popups.

Me personally, I still use Vision almost exclusively for most projects.

You can do Ad Hoc trends in Perspective, I think they are just as good as Vision, others may have more opinions.

---

<div class="post-metadata">

**Author:** ![jlandwerlen](https://avatars.discourse-cdn.com/v4/letter/j/bc8723/32.png) [@jlandwerlen](https://forum.inductiveautomation.com/u/jlandwerlen)\
**Post date:** [January 17, 2021, 11:58pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/4 "2021-01-17T23:58:51Z")

</div>

Also, not sure how multiple monitors work in perspective, I think the newer versions have opened that up, but I honestly don’t know… worth asking though.

---

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 18, 2021, 12:58am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/5 "2021-01-18T00:58:16Z")

</div>

Thanks for the feedback! What are the issues you’ve encountered with popups? I’ve used them OK on a small perspective application - although that was a very simple, one-session iPad controlling a single machine, so it’s not really an exhaustive experiment

---

<div class="post-metadata">

**Author:** ![nminchin](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/nminchin/32/6911_2.png) [@nminchin](https://forum.inductiveautomation.com/u/nminchin)\
**Post date:** [January 18, 2021, 1:22am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/6 "2021-01-18T01:22:05Z")

</div>

> [@ASF](#):
>
> makes mobile implementation easier

Instead of 'easier', I think you mean 'possible' as Vision no longer supports mobile.

I'll say that Perspective still has a way to go before it feels polished. The design experience isn't nearly the same compared to using Vision (performance issues, bugs affecting designing such as missing actions from undo stack, strange keyboard arrow key component movement bugs,) It's certainly come leaps and bounds since the first beta version though.

We've started rolling it out to some customers, as the client runtime is pretty good. Definitely requires more thought though than screens in Vision especially for P&ID mimics with a few gotchas when displaying on different resolutions like font sizes. The load times for P&ID mimics with lots of devices is certainly not as fast as Vision, as the resources ~~are~~ EDIT: **aren't** cached and are loaded every time. You see each device populating on the screen and going through their bindings. A complex mimic might take 2-3s to fully load.

> [@jlandwerlen](#):
>
> multiple monitors work in perspective

Multi-monitor is supported by Workstation where you can configure exactly how and where you want each screen displayed. If using a browser, you can always just open new tabs.

> [@ASF](#):
>
> I have developed a very in-depth trend template on the Vision module

You haven't mentioned the features of your trend page, so it's impossible to say whether or not you'll be able to achieve the same functionality.

---

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 18, 2021, 2:19am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/7 "2021-01-18T02:19:52Z")

</div>

Thanks. I’m definitely familiar with some of the bugs you mention during design - those arrow key moving bugs were the bane of my existence for a while there!

Regarding the display loading time - let’s say I had 20 displays, each with a reasonably simple PID-style mimic, and were switching between them frequently. Would it be slow to load each time? Or once each page loads once, is it cached and loaded quickly on subsequent views? I seem to remember that being the case with the iPad project I did - I mean, I only had one PID-style display, and you’re right, it did take a while to load the first time. But I seem to remember that after the first time, it was pretty quick? If it’s just slow the first time the display is loaded, that’s probably not a big drama as the existing SCADA has the same problem.

Trends - I have a trend template that tries to leverage most of the features of the Vision trend object with runtime bindings. So I have a drop down list for the user to switch between realtime/historical/span modes, I’ve made buttons to automatically scroll half a span/a full span/to start/to end, I’ve given lots of ways to quickly find the data range you want (e.g. enter start/finish date time here, or end time and “previous x hours/minutes” there, etc). Checkboxes to enable/disable the vertical tracker and enable/disable it from automatically following the mouse, easy show/hide tag data so that you can choose between seeing detailed tag information and having more screen real estate for the trend, etc. At design time I’ve made it really easy to create new trends using a “trend” UDT. Probably a few more things I’m forgetting. I mean if I can’t have all of that right off the bat, so be it - it’s just that the last time I looked into real-time trending in perspective (e.g. a live trend of the last 10 minutes), there wasn’t really any such option. There were possibly ways of making something work by brute force, but I’m just hoping that things have developed a bit further since then and that there’s a built-in trending feature that’s at least moderately-well featured.

---

<div class="post-metadata">

**Author:** ![nminchin](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/nminchin/32/6911_2.png) [@nminchin](https://forum.inductiveautomation.com/u/nminchin)\
**Post date:** [January 18, 2021, 2:33am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/8 "2021-01-18T02:33:51Z")

</div>

> [@ASF](#):
>
> let’s say I had 20 displays, each with a reasonably simple PID-style mimic, and were switching between them frequently

I just tested my P&ID mimic nav now directly on the server, and it's ~1.8-2s each time (using a stopwatch 🙂 ) - the page isn't cached, and as I said you can see each component being (quickly) added and bound to tags. I presume this is a caveat of browsers not being able to cache the components. But hopefully something can be done to increase the load speed.

RE the trend, I don't know if you've seen the new Power Chart component, but it addresses the old EasyChart deficiency where you couldn't natively choose tags and change their configuration or the axes configuration. This is all available now as part of the component itself, although you are stuck with their implementation of browsing for tags (e.g. no ability to add the scroll buttons you mentioned, or do any search\* functionality [I believe; I haven't used a PC much before and don't know all the ins and outs yet]). Might have to wait for IA to confirm some of those.

Whoops, just noticed my mistake:

> [@nminchin](#):
>
> as the resources ~~are~~ EDIT: **aren’t** cached and are loaded every time

---

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 18, 2021, 3:27am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/9 "2021-01-18T03:27:17Z")

</div>

Thanks for testing! Do you know if the same (caching) limitations apply when using Perspective Workstation, instead of just a browser? If I were using a browser I’d be largely unsurprised to see that sort of behaviour - typical browser behaviour - but if I were looking at a desktop application I’d expect a better response. I’m optimistic that this might be the case, based on my experience with running Perspective on an iPad. I didn’t specifically test this, and don’t have the machines in front of me to check now - but while I definitely noticed that slow load time, I don’t recall it being annoying. If it was loading slowly every time I looked at it, I feel like I would have been annoyed by it. So I suspect that when using the app, the views may be able to be cached better, and that subsequent views are much faster?

Maybe the developers can chime in on this point?

---

<div class="post-metadata">

**Author:** ![jlandwerlen](https://avatars.discourse-cdn.com/v4/letter/j/bc8723/32.png) [@jlandwerlen](https://forum.inductiveautomation.com/u/jlandwerlen)\
**Post date:** [January 18, 2021, 3:25pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/10 "2021-01-18T15:25:40Z")

</div>

> [@ASF](#):
>
> What are the issues you've encountered with popups?

Dropdowns and almost any entry field if the size of the popup isn't within the viewing area. Plus, there is nothing preventing the user from dragging the popup out of view, then you must refresh the page.

> [@\[Bug-611\]Perspective Dropdown option list collapses if option list scrollbar is clicked](http://forum.inductiveautomation.com/t/bug-611-perspective-dropdown-option-list-collapses-if-option-list-scrollbar-is-clicked/39082):
>
> Perspective 8.0.16 I’m seeming to have an issue with the scroll bar on drop down lists. This is happening in both designer as well as in Chrome. Action in designer: [Designer DDL] Action in Chrome: [Chrome DDL] Basically when there is a list that needs to scroll clicking on the scroll bar or the arrows makes the list disappear. Only way to select stuff is to scroll with the mouse wheel. Any ideas what might be causing this?

> [@Issue with Dropdown & numeric entry field on Perspective Pop Up](http://forum.inductiveautomation.com/t/issue-with-dropdown-numeric-entry-field-on-perspective-pop-up/42151):
>
> I’ve having an issue in 8.0.17 which is similar but not 100% the same as [http://forum.inductiveautomation.com/t/bug-611-perspective-dropdown-option-list-collapses-if-option-list-scrollbar-is-clicked/39082](http://forum.inductiveautomation.com/t/bug-611-perspective-dropdown-option-list-collapses-if-option-list-scrollbar-is-clicked/39082). When I open a pop up (using system.perspective.openPopup) with a drop down or numeric input component I am unable to select anything in either component. If I navigate to the page instead (using system.perspective.navigate), then no issue is seen. Support ticket #5472 created

---

<div class="post-metadata">

**Author:** ![nminchin](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/nminchin/32/6911_2.png) [@nminchin](https://forum.inductiveautomation.com/u/nminchin)\
**Post date:** [January 19, 2021, 3:09pm UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/11 "2021-01-19T15:09:58Z")

</div>

> [@ASF](#):
>
> Thanks for testing!

I actually tried this again comparing Vision to Perspective when connected via 4G (our customer has mobile tablets connected to 4G and VPN'd into their network), and the Perspective pages took ~7s to load fully, while the Vision screens, after initial caching, took ~1s. On first load, Vision took 2.5s. the screens that I tested in both are fairly equal in complexity (roughly 40 valves on the screen with pipe animation). As an operator, 7s would become incredibly frustrating!

Edit: I tested another view that has a number of complex SVGs, and on first load it took a while to load these in, but on second load, it looks like these were cached and it loaded faster, while everything else loaded slowly around the SVGs... Given that, I'm now even more convinced that changing over all of my device templates to be purely SVG will produce faster loading pages... Definitely not work I was hoping to do and certainly far less convenient 😕

Initial load (total time):  
7.5s to display the page (SVGs shown)  
15.5s to fully load the page.

After initial load (total time):  
4.1s to display the page after initial (SVGs shown, after which everything else then loaded)  
10.8s to fully load the page after initial

\>10s is certainly not very responsive!

_**Note: these tests are for a Perspective client running inside of a Vision Web Browser component. I tested running it within Microsoft Edge (chrome or firefox isn't installed..) and it was pretty fast, within 2s. So there must be something up with the Vision web browser**_

---

<div class="post-metadata">

**Author:** ![ASF](https://avatars.discourse-cdn.com/v4/letter/a/c4cdca/32.png) [@ASF](https://forum.inductiveautomation.com/u/ASF)\
**Post date:** [January 22, 2021, 12:08am UTC](https://forum.inductiveautomation.com/t/perspective-vs-vision-trending-and-other-considerations/42272/12 "2021-01-22T00:08:33Z")

</div>

Thanks again for running those tests! Super helpful!

My “complex” view that I referred to in previous posts was almost entirely SVG’s, so your results align with my memory of the view loading slowly the first time, but then more quickly afterward. I guess the upshot is, develop complex graphics in SVG from the get-go, and the issue will be minimised as much as possible.
