Jump to content


Level 2
  • Posts

  • Joined

  • Last visited

About LukeB

Recent Profile Visitors

1,724 profile views

LukeB's Achievements



  1. I have a 2018 iPad Pro and I am also experiencing these slowness issues. A note-taking app should not be so slow on such a powerful device. I do have a number of notes and I excerpt out of books so some of the notes can contain much text, but the computing power inside a 2018 iPad Pro is enormous. Combined with other bugs (old and new), I am seriously considering switching to something else which works across iOS and Windows. As best I can tell, Evernote spends most of its developer time on large businesses, and so their concerns matter much more than ours.
  2. I am having this problem as of updating to iPad iOS 14 and then updating Evernote. I copy out excerpts from books and thus have fairly long notes; since there's no easy way to scroll to the end of a note, I put something like "XXX" at the end of the note, so I can search for it in order to scroll to the bottom of the note. This no longer works, which is rather frustrating! This is also true for shorter notes when I search for content within them.
  3. My note doesn't contain "https://evernote.com"; it contains "https://www.evernote.com" and "evernote:///view". I wrote code to easily see the HTML code for contents on the clipboard; I just examined one of my oldest old green links and it's https://www.evernote.com. So, it appears that DTLow incorrectly diagnosed the problem. As to the sequence of operations you specify, Here's precisely what I did, from left to right: When I did that, the link produced was https://www.evernote.com/shard/s6/…, as I specified in the OP. Can you confirm that (i) you are running Evernote on iOS; (ii) you are using version 8.12? If you meant clicking the icon next to the text "Shareable link off", that doesn't do what I want—it generates a public link. This is undesirable in two ways: (a) it requires an internet connection to generate the link; (b) it makes the link work for anyone who has it, which I do not need and is this an additional possible security hole. However, it does generate a nice pretty green hyperlink with text equal to the note's title.
  4. To reproduce: in Evernote 8.12 for iOS (released 2018-05-18) go to any note tap the person icon with a plus sign tap "More sharing options" tap "Copy internal link" tap "Done" paste from the clipboard into a different note the result is something like https://www.evernote.com/shard/s6/… Note that DTLow reported this, but the problem (on iOS) isn't actually that it is a web link vs. Evernote link (evernote:///view/…); I can click on the ugly link and be taken back to the appropriate note on my iPhone and Windows machine ( The problem is that in iOS, no longer do I get a nice green hyperlink with the title of the note; I now get an ugly blue URL. This is particularly frustrating for me because I link to internal notes rather extensively. It's already bad enough that Evernote on iOS makes it onerous to find another note, copy an internal hyperlink through several taps, navigate back to the original note, find where I was entering text, and paste. But now I'll have to either deal with ugly URLs which don't tell me where they go, or re-edit everything on my PC. To add insult to injury, I ponied up for an iPhone X in part because Evernote functioned so slowly on my iPhone 5S.
  5. This is a ridiculous amount of added margin: Surely Evernote could at least present a checkbox in settings for whether to display that margin. Even if it means that the table config UI is screwed up when the margin is hidden, it'd remove a massive inconvenience for the great number of users who never/rarely use tables. I would rather not be permanently stuck on an older version of Evernote in order to avoid this margin.
  6. Just a warning: last time I checked (sometime in 2016), inter-note links will not be maintained with imports, as the GUIDs for the notes themselves are not exported. Hopefully this has changed, but sadly I doubt it. One solution would be to put a note link to the note itself in the beginning of every note (perhaps via some automation utility) pre-export, doing search/replaces after import. But if you don't have the right skills to do this, it would be quite onerous.
  7. Suppose I copy a note link and paste it into another note. As long as there is some sort of text on the same line before the current cursor location, a space will be added. Here are three examples; the vertical bar represents the cursor location and the second line shows the result of the paste. | Note Link a| a Note Link b | b Note Link Suffice it to say that this is obnoxious behavior. (Especially the third example.) Not only do I link my notes together quite a lot, but I also have an AutoHotkey script which adds red underlining to selected text and can be triggered via whatever hotkey one chooses. This results in a space being prepended to the pasted, underlined text. The pattern is the same as the above; unless I've selected text starting at the beginning of the line, Evernote prepends a space. Now, here's something more obnoxious. Suppose I have the following text: one two three Here's what happens when I red-underline the middle word: one two three Not only is a space prepended, but the highlighting is broken. This is not what I want. I have read elsewhere that some like the fact that pasting prepends a space, but I find it wastes time and makes it much harder to enhance Evernote. I would not be surprised if a great number of people would love the ability to set up AutoHotkey scripts to perform text formatting Evernote doesn't provide UI for, but does support via its HTML rendering engine. Sadly, Evernote's "smart" paste feature—or buggy pasting—makes this an ugly user experience. P.S. Sometimes no space is prepended, if the formatted text were copied from an Evernote. However, then there can be other bugs, which I shall not go into at the time. P.P.S. The same highlight-breaking behavior happens regardless of whether the pasted HTML has either nesting of two <span> elements (one highlight, the other bottom-border), or one combined <span> element.
  8. I don't see this as being off-topic in any way. The resulting impression people have of how important this performance problem is will impact how much attention the performance problem gets. For those actually experiencing this problem, we want more attention. For people like you who aren't experiencing this problem, you perhaps want less attention paid to this issue, because there are bugs and/or new features which would benefit you, competing with this performance problem. First, "professional experience" ≠ "mystic reason". Second, I have presented a very important fact (whether you call that "data" is irrelevant): See: "The slowness problem did not exist until the OS upgrade." This is not the case; before a particular version, Evernote was plenty fast with large notes. The slowdown started around the time that better search was introduced and the display of "Context". Perhaps it was precisely the addition of this feature which caused the problem. This is where I reiterate what I said in the second post in this thread.
  9. You are welcome to deny that long experience with software development gives me expert knowledge in this topic, declaring the matter "opinion". However, this does not make what you say true. There are objective truths in this realm. I gave you an existence proof: prior to a certain version, Evernote performed very well with long notes. We could look at the amount of computation required for "a large, complex, web page" and compare it to 90 KB notes which are mostly text with very little formatting. I did a force-refresh on page one of this thread and Google Chrome's Web Developer Tools noted 1.6 MB transferred. Much less of that was raw text and much more was formatting; it is much more computationally expensive to render formatting than to render text. And once again, you ignore the 'existence proof' that prior to an update, Evernote was lickety-split fast with 90 KB notes. There is a difference between excusing and influencing the priorities. It is the difference between a binary influence and an analogue influence. You are not allowed to dictate, by fiat, what the impact of your words are on the relevant people. You can say whatever you want about what you meant to communicate, but the actual impact—e.g. how big of an issue Evernote thinks this problem is—can indeed be altered by what you write. I am concerned that the net impact, the objectively true impact, is that many folks will consider this a non-issue. Whether this is the case or not is not 'opinion'. Evernote must expend its resources wisely, which means that if a problem is a non-issue for sufficiently many people, that problem will receive correspondingly less effort to resolve. I have. The result was a hacky fix that made the slowdowns spottier, jerkier. That's almost worse, because it'll be lickety split for a while and then all of a sudden slow down, messing up the rhythm of editing the note. The root problem simply was not solved. If slowness, caused by CPU pegging, scales with the length of note, then deductive logic indicates that people who regularly work with notes greater than a certain size will indeed experience precisely what I say they will experience. That you do not experience this indicates that your notes are sufficiently small, and/or you spend sufficiently little time using Evernote in comparison to other applications. In no way do your data constitute a refutation of my point. What does remain to be seen is how many people actually operate in the domain where Evernote will inordinately drain their batteries. My point is that this symptom will be largely hidden from most users who operate in a domain between yours (smaller notes and/or less proportional usage) and mine (large notes, where I'm mostly using Evernote for my session at my computer while on battery power).
  10. If I were to hazard a guess, I'd say Kensington London rejects the reasoning you have expressed here; I do as well. I wrote my first line of code over twenty years ago and am well-aware of the progress in computational power since. Your argument here may work for ten years ago, but it is completely obsolete, today. This is clearly demonstrated by the fact that Evernote Mac did not always have the performance problem identified in this thread. It was introduced. Now, perhaps the performance prior to the offending update was a spurious boon and EN folks never intended that, but I think most people would see this as a post-hoc rationalization. Stating workarounds is fine, but I think the way you're coming off is that you don't ever need this feature, and so the extra clicks and so forth required by others is completely acceptable. Such reasoning relaxes the pressure on EN to actually do something about this problem. At the very least, they could publish some engineering details. I've even offered to drive down to the EN office in the Bay Area and let them install a debugger and/or profiler on my machine to see what's going on. But if the workaround is 'simple', then the urgency on this matter is reduced, making the product experience inferior for a number of people. (But perhaps we're too small a number?) Furthermore, there is a "hidden problem" that I bet many won't recognize: increased battery usage. With long notes the CPU is pegged for long enough for users to notice. With shorter notes, the CPU will still be used inordinately much, but users will be less likely to notice. Instead, they'll just experience that their battery drains more quickly, and perhaps blame their laptop manufacturer. I doubt very many have a CPU trace running in their menu bar. Furthermore, people are still, unfortunately, too used to the poor performance that may have been necessary ten years ago, but is absolutely unnecessary today.
  11. I did not notice a difference when I turned it off (which I did a long time ago). The reason this is probably the case is that Context searching likely depends on the same indexing functionality which gives you better search in Evernote. Alternatively, perhaps turning it off only affects the UI and not underlying computation. There are two different general categories of high CPU usage. One is "one-time", like re-indexing a database—like switching from a version with "dumb" search to one based on fancy indexing. Another is the high CPU usage which accompanies slowness like is well-documented in this thread. The first thread I mentioned doesn't seem like "one-time"-style CPU usage. Anyhow, whether or not the two threads I mentioned are "one-time" or not is somewhat immaterial; the slowness mentioned in this thread will definitely result in excessive battery drain.
  12. This is not the case; before a particular version, Evernote was plenty fast with large notes. The slowdown started around the time that better search was introduced and the display of "Context". Perhaps it was precisely the addition of this feature which caused the problem. Note that what is perceived as slowness on large notes will still take inordinately much processing power on smaller notes, which means that Evernote has a battery usage problem. I quickly searched and found the forum threads High cpu usage and very poor performance (Windows) and Evernote Mac 6.0.5: 100% CPU Utilization (Mac).
  13. This issue is becoming increasingly frustrating. It reliably happens when I am editing large notes (e.g. 79.0 KB, 10,817 words, 67,575 characters, no images). I live in SF proper and I know there is an Evernote office complex nearby along SR 101. Would it be possible for me to come in, have you install a debugger/profiler, and look at the problem in action? Let me reiterate. This slowness is making Evernote unbearable for my use. I really do like what Evernote does, but this slowness is increasingly unacceptable. Note that I am a paying customer.
  14. I wonder if continuously indexing a note while it is being edited, whether for context or search (or perhaps the same indexing for both), is at all responsible for this. Do you suggest anything better than OS X's "Activity Monitor" which I could use to try and see whether the slowness is due to CPU, memory, or disk? I'm pretty sure it's not network. I used to be a Windows guy and perfmon was awesome; I haven't investigated good alternatives on OS X. I use iStat Menus for global indicators, but they've proven insufficient.
  15. I have since noticed that typing (and other mentioned operations) sometimes speeds back up to normal, during e.g. typing up sections from a book. Perhaps a garbage collection takes place? Note that performance degrades again.
  • Create New...