I find it extremely disturbing that you don't just know what I mean. This is a huge change from the previous major release. It's so serious that I'm probably going to have to back to the earlier version of TextMate.

Maybe instead of "load" I should have said "parse"? Try to think of what TM does when it opens a file that it now does differently.

* A typical symptom is that we are at the start of a file and we say command-down-arrow to go to the end of the file and instead we end up somewhere in the late area of the file (not the end) because the whole file has not loaded. After several tries (command-down-arrow again and again) we might get to the end. This was always the case but now it is worse.

* Another typical symptom is that we bring up the global find and click a line to jump to it and we are miles off from it, and have to command-J to reach it. This was always the case, but the difference from the previous version is the actual selected line appears but then jumps away out of view and we have to command-J again, and possibly again, to get the window to steady itself.

* Similarly just on scrolling down page by page, we often jump back up right after the scroll. I take it that this is because more stuff is loading (or parsing or whatever you choose to call it) and TM is revising its mental picture of how long this page is.

The "blank page" phenomenon seems able to arise on any of those. It is as if TM were saying, "sorry, you've reached a point in the document beyond where I've parsed, I've nothing to show you". So the file appears to come to an end in the middle. A few seconds later, the missing material appears.

I enclose a screencast showing the phenomenon. We pick a heading to jump to. Instead of jumping to it, we jump somewhere else, then jump again, and we are still not there. I have to pick the heading _again_ to see it. Everything I do is taking twice as long as it should because I have to do everything twice like this. This is all completely new (to me) in this release; keep in mind that I have not updated since the last major release so I don't know exactly when all this came along.

This is exactly the opposite of how I hoped improvements would go. With my long text (asciidoc) documents, TM is now almost unusable because of how the window contents leap around.

m.

On Sep 17, 2019, at 11:55 PM, Allan Odgaard mailinglist@textmate.org wrote:

On 17 Sep 2019, at 22:12, Matt Neuburg wrote:

I'm having great difficulty persuading the newest TextMate 2 to load the whole of my file when I open it. Actions like scrolling or global find result in long delays or large apparently blank areas. It would be nice if when I say "open this file" TextMate 2 would well and truly just _open_ it. m.

I can assure you that there is no code to only partially load a file.

Is there anything special about these files? For example special encoding, special anatomy like database dumps or XML documents on one single multi-megabyte line? Prose, markup, or code files?

If there are “blank areas” is it something you can screen capture? That might help understand what the underlying issue is.

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

-- matt neuburg, phd = http://www.apeth.net/matt/ pantes anthropoi tou eidenai oregontai phusei Programming iOS 12! http://shop.oreilly.com/product/0636920208662.do iOS 12 Fundamentals! http://shop.oreilly.com/product/0636920208655.do RubyFrontier! http://www.apeth.com/RubyFrontierDocs/default.html

On Wed, Sep 18, 2019 at 10:14 AM Matt Neuburg matt@tidbits.com wrote:

I find it extremely disturbing that you don't just know what I mean. This is a huge change from the previous major release. It's so serious that I'm probably going to have to back to the earlier version of TextMate.

Maybe instead of "load" I should have said "parse"? Try to think of what TM does when it opens a file that it now does differently.

• A typical symptom is that we are at the start of a file and we say command-down-arrow to go to the end of the file and instead we end up somewhere in the late area of the file (not the end) because the whole file has not loaded. After several tries (command-down-arrow again and again) we might get to the end. This was always the case but now it is worse.

• Another typical symptom is that we bring up the global find and click a line to jump to it and we are miles off from it, and have to command-J to reach it. This was always the case, but the difference from the previous version is the actual selected line appears but then jumps away out of view and we have to command-J again, and possibly again, to get the window to steady itself.

• Similarly just on scrolling down page by page, we often jump back up right after the scroll. I take it that this is because more stuff is loading (or parsing or whatever you choose to call it) and TM is revising its mental picture of how long this page is.

The "blank page" phenomenon seems able to arise on any of those. It is as if TM were saying, "sorry, you've reached a point in the document beyond where I've parsed, I've nothing to show you". So the file appears to come to an end in the middle. A few seconds later, the missing material appears.

I enclose a screencast showing the phenomenon. We pick a heading to jump to. Instead of jumping to it, we jump somewhere else, then jump again, and we are still not there. I have to pick the heading _again_ to see it. Everything I do is taking twice as long as it should because I have to do everything twice like this. This is all completely new (to me) in this release; keep in mind that I have not updated since the last major release so I don't know exactly when all this came along.

Thanks for the specific examples. I was able to reproduce the example shown in the screencast with the latest release and in doesn't happen (or at least not as bad) in 2.0-rc10, which I believe was the last release. There were quite a few nightly builds in between, but I will try to determine which release actually introduced this change to narrow the scope.

/Ron

This is exactly the opposite of how I hoped improvements would go. With my long text (asciidoc) documents, TM is now almost unusable because of how the window contents leap around.

m.

...

On Sep 17, 2019, at 11:55 PM, Allan Odgaard mailinglist@textmate.org wrote:

On 17 Sep 2019, at 22:12, Matt Neuburg wrote:

I'm having great difficulty persuading the newest TextMate 2 to load the whole of my file when I open it. Actions like scrolling or global find result in long delays or large apparently blank areas. It would be nice if when I say "open this file" TextMate 2 would well and truly just _open_ it. m.

I can assure you that there is no code to only partially load a file.

Is there anything special about these files? For example special encoding, special anatomy like database dumps or XML documents on one single multi-megabyte line? Prose, markup, or code files?

If there are “blank areas” is it something you can screen capture? That might help understand what the underlying issue is.

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

-- matt neuburg, phd = http://www.apeth.net/matt/ pantes anthropoi tou eidenai oregontai phusei Programming iOS 12! http://shop.oreilly.com/product/0636920208662.do iOS 12 Fundamentals! http://shop.oreilly.com/product/0636920208655.do RubyFrontier! http://www.apeth.com/RubyFrontierDocs/default.html

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

Yes, I was using 2.0-rc10 before (released 2018-06-11).

I did have to say Command-J sometimes after a global find or similar, but this was a well known minor issue, and saying it once was enough, and the window did not also jump around (scroll) at arbitrary times, as when switching to a new tab.

If you could get the new version to behave more like 2.0-rc10, that would be great! Of course I was hoping that the need for Command-J after a search would eventually just go away entirely, but instead the new 2.0 has gone in the opposite direction and made the window jump around all the time, so that I'm constantly wrestling with it by hitting Command-J, often multiple times before the window settles down.

So, because the new 2.0 is completely unusable, I have reverted to 2.0-rc10. I have work to do (writing a 1000-page book); I can't be spending all my time responding to these unexpected multiple jumps in the window.

Please do let me know if there's a new release for me to test! I'll do anything I can to help.

m.

On Sep 18, 2019, at 8:50 AM, Ronald Wampler rdwampler@gmail.com wrote:

On Wed, Sep 18, 2019 at 10:14 AM Matt Neuburg matt@tidbits.com wrote:

...

I find it extremely disturbing that you don't just know what I mean. This is a huge change from the previous major release. It's so serious that I'm probably going to have to back to the earlier version of TextMate.

Maybe instead of "load" I should have said "parse"? Try to think of what TM does when it opens a file that it now does differently.

• A typical symptom is that we are at the start of a file and we say command-down-arrow to go to the end of the file and instead we end up somewhere in the late area of the file (not the end) because the whole file has not loaded. After several tries (command-down-arrow again and again) we might get to the end. This was always the case but now it is worse.

• Another typical symptom is that we bring up the global find and click a line to jump to it and we are miles off from it, and have to command-J to reach it. This was always the case, but the difference from the previous version is the actual selected line appears but then jumps away out of view and we have to command-J again, and possibly again, to get the window to steady itself.

• Similarly just on scrolling down page by page, we often jump back up right after the scroll. I take it that this is because more stuff is loading (or parsing or whatever you choose to call it) and TM is revising its mental picture of how long this page is.

The "blank page" phenomenon seems able to arise on any of those. It is as if TM were saying, "sorry, you've reached a point in the document beyond where I've parsed, I've nothing to show you". So the file appears to come to an end in the middle. A few seconds later, the missing material appears.

I enclose a screencast showing the phenomenon. We pick a heading to jump to. Instead of jumping to it, we jump somewhere else, then jump again, and we are still not there. I have to pick the heading _again_ to see it. Everything I do is taking twice as long as it should because I have to do everything twice like this. This is all completely new (to me) in this release; keep in mind that I have not updated since the last major release so I don't know exactly when all this came along.

Thanks for the specific examples. I was able to reproduce the example shown in the screencast with the latest release and in doesn't happen (or at least not as bad) in 2.0-rc10, which I believe was the last release. There were quite a few nightly builds in between, but I will try to determine which release actually introduced this change to narrow the scope.

/Ron

...

This is exactly the opposite of how I hoped improvements would go. With my long text (asciidoc) documents, TM is now almost unusable because of how the window contents leap around.

m.

...

On Sep 17, 2019, at 11:55 PM, Allan Odgaard mailinglist@textmate.org wrote:

On 17 Sep 2019, at 22:12, Matt Neuburg wrote:

I'm having great difficulty persuading the newest TextMate 2 to load the whole of my file when I open it. Actions like scrolling or global find result in long delays or large apparently blank areas. It would be nice if when I say "open this file" TextMate 2 would well and truly just _open_ it. m.

I can assure you that there is no code to only partially load a file.

Is there anything special about these files? For example special encoding, special anatomy like database dumps or XML documents on one single multi-megabyte line? Prose, markup, or code files?

If there are “blank areas” is it something you can screen capture? That might help understand what the underlying issue is.

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

-- matt neuburg, phd = http://www.apeth.net/matt/ pantes anthropoi tou eidenai oregontai phusei Programming iOS 12! http://shop.oreilly.com/product/0636920208662.do iOS 12 Fundamentals! http://shop.oreilly.com/product/0636920208655.do RubyFrontier! http://www.apeth.com/RubyFrontierDocs/default.html

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

---

TextMate mailing list TextMate@lists.macromates.com https://lists.macromates.com/listinfo/textmate

-- matt neuburg, phd = http://www.apeth.net/matt/ pantes anthropoi tou eidenai oregontai phusei Programming iOS 12! http://shop.oreilly.com/product/0636920208662.do iOS 12 Fundamentals! http://shop.oreilly.com/product/0636920208655.do RubyFrontier! http://www.apeth.com/RubyFrontierDocs/default.html

Hi All,

I’ve reproduced this problem on 2 different macs so far. I use a theme that has a semi transparent background, which no longer works in the 2.0 release after rc31. Is there a known cause?

Sam