View Full Version : Kindle doesn't have enough memory??


wastelander
06-13-2012, 01:36 PM
Hi, I've just loaded up a bunch of books onto my new K5T and opened up a PDF, after 1 or 2 pages it came up saying that the selected page could not be opened and that my kindle doesn't have enough memory?? Im a little confused, my normal books seem ok though, was is just that its pdf? Also im waiting for it to index about 90 items lol, how long should this take and does it index while sleeping?

geekmaster
06-13-2012, 02:27 PM
Hi, I've just loaded up a bunch of books onto my new K5T and opened up a PDF, after 1 or 2 pages it came up saying that the selected page could not be opened and that my kindle doesn't have enough memory?? Im a little confused, my normal books seem ok though, was is just that its pdf? Also im waiting for it to index about 90 items lol, how long should this take and does it index while sleeping?If your book contains a bunch of PICTURES of pages, it can be quite large. What is the file size of the book that has "memory" problems?

HarryT
06-13-2012, 02:34 PM
If it's a complex PDF page, it's entirely possible that the Kindle doesn't have enough memory to render the page. Memory - RAM - not file storage space.

geekmaster
06-13-2012, 03:04 PM
If it's a complex PDF page, it's entirely possible that the Kindle doesn't have enough memory to render the page. Memory - RAM - not file storage space.The file size determines whether it contains text or a lot of big nasty pictures of text, which can make "rendering" (loading and displaying an image of a page) slow or just not work...

wastelander
06-13-2012, 05:00 PM
Thanks guys, the RAM makes sense - the files rang from 10MB-50MB lol:eek:

Thanks again!

Bookatarian
06-13-2012, 11:18 PM
I had that happen to me a couple of times while trying to enlarge a PDF knitting pattern...no pics, only text, though I'm not sure of the file size just at the moment. A restart of the Kindle solved the problem entirely. Hasn't happened again with that or any other document. Just to note, prior to that, I'd been working extensively with my Kindle and multiple PDF's, adding to library, personal docs and more and it had been quite a long time (maybe a month) since I'd actually done a restart.

geekmaster
06-13-2012, 11:20 PM
The free RAM does slowly decrease, eventually needed a restart to clean it up and start with a lot more free RAM.

Ubermensch
06-13-2012, 11:21 PM
If your book contains a bunch of PICTURES of pages, it can be quite large. What is the file size of the book that has "memory" problems?
Don't confuse memory with storage. :P

geekmaster
06-14-2012, 12:09 AM
Don't confuse memory with storage. :PUnless we have an English translation problem, computer "memory" and "storage" are EXACTLY the same thing. The two types of "memory storage" used in the kindles just happen to use different technology: one is faster, and the other is nonvolatile. The latest generations of flash memory are so fast that computers will only have one kind of memory storage in the future, with no artificial separation between fast memory and nonvolatile memory like we use now. They will just have a LOT of fast nonvolatile memory.

Background Information:
Computers have been my hobby, my job, and my LIFE for half a century. I still have some core memory from one of my early computers. I recently ripped apart my last reel-to-reel computer tape drive (like in old Sci-Fi movies) to steal the servo motors for a CNC project. I still have a 12MB hard drive the size of a clothes washing machine. It contains code that I wrote back in the day. :)

I have designed and built computer hardware, and I have written operating system device drivers and mission-critical firmware. I have designed character fonts and rendering engines. I have studied the code inside PDF rendering engines. I am not confused easily by computers. People on the other hand, are a complete mystery to me... :eek:

The point of my post is that PDF files made up of IMAGES instead of OCR text can be very large files, and can consume a lot of RAM for the currently displayed page, and any cached pages used to speed up page changes.

Some people complain that I use too many words, and others incorrectly "fill in the blanks" when I do not use enough words, and then suggest that I do not understand. You cannot make everybody happy... :(

Jessica Lares
06-14-2012, 04:35 AM
Whenever you can do it, I'd honestly suggest converting PDFs to another format. PDFs are great when you want exact placement of things, like a brochure, or huge manual, but not for articles, recipes, photo essays, and readings with images. That's why they developed the K8 format. Richly detailed books without the excess bulk that comes with other formats.

The problem is that the PDFs are meant to scale for printing. Whether it's for a letter sized piece of paper, or it's for a huge poster for a billboard. Open up any manual online and see how far it lets you zoom in until the images and text aren't clear anymore, for example (I just did and it let me zoom to 1,000%+ :eek:). When you embed images, it's just sticking them in without any quality compression. That's where you get those 50MB files for only just a few 20-50 pages.

And no, computer "memory" and "storage" is not the same thing. TECHNICALLY, sure, but let's not get people confused geekmaster. ;)

The Kindle Touch has 256MB of DRAM memory according to the teardowns and specifications of that part. The Kindles before the third/keyboard have 128MB. The Kindle OS probably takes about 20-50 of it on its own, the rest for the reading, browsing, etc "apps" if you call them. It's like when your computer starts and has to load the Windows/Mac OS GUI and you open up Internet Explorer or Safari.

The Kindle storage on the other hand varies from 2-4GB depending on the model. That is where your books and whatever else you keep in there is. However, as you note, when they tell you it's 2GB, and it's only 1.4GB, it's not just because of the fancy number explanations they give you about the bytes and rounding, etc, it's because they've allocated the remaining 600MB for the OS, drivers, firmware updates, and virtual memory (which the Kindles must have very little of). So when you open up that 50MB PDF file, it uses a little bit of your RAM just to open it, and a little bit of virtual memory/that allocated hard drive space to store those images until you either turn off the Kindle completely or it expires (depending on how Amazon handles that anyway). Think of it as your internet browser cache which saves most of the GUI of websites so it can render them faster as you navigate and go back to pages the next day.

So when you get that error that your Kindle doesn't have enough memory to render the rest of the pages, it's because there is not enough space VIRTUALLY to handle the images and embedded text in that PDF. It's not a problem with Amazon bought ebooks because they have guidelines in place that don't let authors use images bigger than 127KB in size.

geekmaster
06-14-2012, 05:12 AM
And no, computer "memory" and "storage" is not the same thing. TECHNICALLY, sure, but let's not get people confused geekmaster. ;)As you say, TECHNICALLY I am correct, so it is YOU (and others) who are confusing people by misusing TECHNICAL terminology in a WRONG and MISLEADING manner.

What you call STORAGE is mmc flash MEMORY. Flash memory is a form of nonvolatile MEMORY. The part of it that you see as a USB drive only LOOKS like a hard drive, but it is really flash MEMORY.

Sure, you can STORE stuff on it, but you also STORE stuff in DRAM, and in fact those contents are nonvolatile as well (even when in screensaver mode), unless your kindle gets restarted.

It is you who are misusing the terminology (and perhaps others here). Yes, in the kindle there are two types of MEMORY just like I said. And the operating system treats them both the same way. What looks like a hard drive is partly in RAM (/tmp), and partly on the mmc flash memory (the root partition, and the exported USB drive partition).

What looks like "RAM" can also be on both devices because linux maps files in using virtual memory. The files are assigned CPU memory addresses. But move back and forth between mmc and RAM depending on CPU load.

I believe that you are misusing technical terminology. Not me.

If you wish to distinguish between the different types of memory storage devices used in the kindle, the CORRECT names are RAM and MMC. They are NOT "memory" and "storage".

Before you falsely accuse me of being wrong again, please look up the definitions of the words. I will even help educate you so you can stop telling people who design this stuff and have used it on a daily basis for decades that THEY are wrong.

Both of the following quoted defining phrases are the first links in the first sentence of each page:

Wikipedia defines MMC (what you call STORAGE) as "flash memory memory card": http://en.wikipedia.org/wiki/MultiMediaCard.

Wikipedia defines RAM (what you call MEMORY) as "computer data storage": http://en.wikipedia.org/wiki/Random-access_memory

You have your terminology the EXACT OPPOSITE of Wikipedia. You will find examples supporting your usage too, which just proves my statement that MEMORY and STORAGE are interchangeable and mean exactly the same thing.

Please stop telling ME that I am wrong until you get your own terminology straight, which is really what "confuses people" (as you said).

P.S. At least HarryT clearly defined his usage of "memory" and "storage" (above) as "memory - RAM" and "file storage space". Those complete phrases are technically accurate, but just using the bare interchangeable words by themselves is not accurate. Accurately using language is critical when you choose to participate in a debate such as this (especially when your argument goes AGAINST the supporting evidence: Wikipedia in this case).

HarryT
06-14-2012, 05:39 AM
Leaving to one side these doubtless interesting discussions of the definitions of different types of storage space, it's worth noting that PDF complexity is not just related to the size of images. A PDF page is rendered by executing drawing instructions stored in the PostScript language, and if a page contains complex diagrams, for example, these instructions can require more memory to execute than the Kindle has available. This can lead to the "out of memory" error that the OP saw.

geekmaster
06-14-2012, 05:50 AM
Leaving to one side these doubtless interesting discussions of the definitions of different types of storage space, it's worth noting that PDF complexity is not just related to the size of images. A PDF page is rendered by executing drawing instructions stored in the PostScript language, and if a page contains complex diagrams, for example, these instructions can require more memory to execute than the Kindle has available. This can lead to the "out of memory" error that the OP saw.There are certainly other forms of complexity in PDF files that can consume memory, but generally, instructions to dynamically build a page are an extremely efficient form of "data aware" compression (the best kind). The real problem is the temporary storage buffers in RAM that these instructions use while building image layers that get displayed. Full-page images of text pages (scanned books that were not OCR'd) use much poorer "generic" compression than dynamically generated images created from PostScript instructions, and are likely to consume more RAM just like they consumed more disk space on the USB drive.

Regarding using up available RAM storage space, OTHER apps that run "all the time" are stored in RAM while they are running, and they will also consume RAM storage space. You can see these with the "top" command from an SSH command prompt (which requires a hack). Kindles also slowly run out of memory and may need rebooting occasionally, to free up some RAM, due to how the single-process Java virtual machine allocates static memory for kindlet apps, which NEVER gets freed until you reboot.

There are reports in other threads that large PDF files that fail to load SUCCESSFULLY load after rebooting.

ThomasC
06-14-2012, 05:59 AM
There are reports in other threads that large PDF files that fail to load SUCCESSFULLY load after rebooting.

Same goes for web-browsing since it clears the memory cache. You can clear the browser cache in settings, but it does not seem to work as well as a reboot. So, in some large memory intensive pdf's it might be worth a try.

Jessica Lares
06-14-2012, 06:11 AM
WHO CARES?! Obviously you do, of course. However, the way you are putting it isn't understandable by the average person who would like to know the difference. It is like typing LOL, BRB, and other shorthand. You have to explain to educate, not to show off you're more intelligent. You also have to understand that just telling people to search for information themselves doesn't do a thing. It is an attitude like that which made me, as well as a lot of other people really hate our teachers in secondary school. You can read a book, but it doesn't mean you'll understand beyond what the text says, except what it does, especially in a dictionary.

Another thing, Wikipedia is user-generated content, anyone can edit it. Anyone who studied in this decade will know that it's not accepted as a single credible source.

geekmaster
06-14-2012, 06:11 AM
Same goes for web-browsing since it clears the memory cache. You can clear the browser cache in settings, but it does not seem to work as well as a reboot. So, in some large memory intensive pdf's it might be worth a try.Actually, the browser cache is stored in the mmc memory, not in RAM memory. Rebooting clears up things that were allocated in RAM memory and not freed, either because the app was still running, or because it was a kindlet and the RAM it used really belongs to the JVM and not the kindlet, and some of it never gets freed until restarting the desktop framework (and rebooting is the easiest way).

Clearing the browser cache would also free some virtual memory address space (cached files mapped into RAM) when those files are deleted, so freeing RAM is just a side-effect of clearing the "on-disk" browser cache.

ThomasC
06-14-2012, 06:27 AM
Oops, I used those confusing technical hacker words again. Darn! I just thought that describing what REALLY happens using scientific facts and accurate technical terminology could help dispell some of the empirical superstitious opinions that seem to be a way of life in some corners of mobileread... Sorry if I confused you with the facts. :eek:

So, in simple non-geek-speak, rebooting can help. However, it will not always help with 'very' complex pdf's or websites. Still, always worth a try imho.

geekmaster
06-14-2012, 06:36 AM
So, in simple non-geek-speak, rebooting can help. However, it will not always help with 'very' complex pdf's or websites. Still, always worth a try imho.Rebooting almost always helps, even on non-kindle devices. All complex systems have a certain amount of acceptable "resource leakage" that makes them get slower until apps start to fail. Rebooting makes it all fresh again. It is just that for kindles, rebooting is not "normal" (but probably should be). At least they provided a menu item to do that, instead of a little hole in the back you would poke a paper clip into to reboot some devices.

ThomasC
06-14-2012, 07:18 AM
Nice simple explanation, thanks.

wastelander
06-14-2012, 08:04 AM
Yep, thank's guys.

geekmaster
06-14-2012, 08:20 AM
WHO CARES?! Obviously you do, of course.

Another thing, Wikipedia is user-generated content, anyone can edit it. Anyone who studied in this decade will know that it's not accepted as a single credible source.I have moderator priveleges at Wikipedia. I can create new threads. There are actually some pretty strict guidelines. Non-accurate content does not last long there. It is MUCH more reliable than it was a few years ago. There are literally thousands of unpublished threads dating back mulitple years waiting for moderator approval. I can approve them. And the content I referenced matches decades of experience in the field that defined these words. Here is another source that calls RAM "storage": http://dictionary.reference.com/browse/non-volatile+storage

When I started using computers, we did not have hard drives. Programs were in nonvolatile core memory, and just stayed there. And it did its job, much like embedded computers in a microwave oven or automobile. You could mount a magnetic tape and change the programming, if you needed to upgrade your program (or if you needed to run a different job, like month-end processing).

Data was stored in "core storage" (also called core memory). Synonyms. You can distinguish internal and external storage, or volatile and nonvolatile, but they are still storage and memory. And future computers are going back to having only one kind of memory storage device.

Saying that memory and storage are not the same thing is like saying that water and hydrogen hydroxide (H2O) are not the same thing. If you do not understand how technical words are used correctly, then please do not mis-correct people who do understand how to use them in their proper context.

ThomasC
06-14-2012, 08:50 AM
Sorry mate but I think you need a Forum-Break.

geekmaster
06-14-2012, 09:21 AM
Sorry mate but I think you need a Forum-Break.This thread asked about "enough memory" which is a TECHNICAL matter. Perhaps this thread should be moved to the Dev Corner forum where people not only care about technical accuracy, but they try to HELP people instead of just guessing and criticizing those who do attempt to be technically accurate.

geoffwood
06-14-2012, 10:56 AM
This thread asked about "enough memory" which is a TECHNICAL matter. Perhaps this thread should be moved to the Dev Corner forum where people not only care about technical accuracy, but they try to HELP people instead of just guessing and criticizing those who do attempt to be technically accurate.

Perhaps if this thread stuck to helping people, instead of mud throwing about terminology, the OP would be able to work with his file by now. Just saying.

geekmaster
06-14-2012, 01:03 PM
The K5 has a LOT less free memory than previous kindle models (even those that have less total memory to begin with). Much more of its memory is used up by the built-in software. Early on when it was first released, several developers (including myself) bricked their kindles just by doing rather simple "ordinary" commands at the command prompt. Even now, developers routinely debrick their kindles. A significant amount of the problem is caused by running out of resources when you have extra programs running along with the kindle framework.

So it is not at all surprising that a large or complex PDF file could cause the K5 to run out of free RAM needed to display it, especially when the kindle has been running a long time since the last restart.

EDIT: On earlier kindle models you could shutdown the framework before running a large custom app, such as a custom PDF reader. On the K5, the system reboots when it notices that the framework (cvm and Xorg) is not in memory. You can pause and resume the framework, but you cannot unload it from memory, as far as I know. Perhaps there is a yet-undiscovered (or unpublished) way to do it. That would let you view large complex PDF files using a third-party PDF viewer program, like you can do now on a K3 or DX.

geekmaster
06-14-2012, 01:19 PM
Perhaps if this thread stuck to helping people, instead of mud throwing about terminology, the OP would be able to work with his file by now. Just saying.That seems to happen every time people accuse me of getting my facts wrong without any supporting evidence other than their own personal opinion, and I feel a need to provide details and references to support my argument. Perceived attacks against my personal credibility have always been a "hot button" of mine. Sorry about that...

Some of my posts also attempted to supply helpful information. I have helped a LOT of people here (and especially in the dev corner forum). That seems to rub some people the wrong way, it seems -- ESPECIALLY when I try to help people who ask technical questions in THIS forum.

QuantumIguana
06-15-2012, 01:34 PM
As far as how the English language is used, yes there is a difference between memory and storage. In a certain context, they are the same. But as most people use the term, they are not. If you do a search of "memory vs. storage" there is quite a lot of disagreement that they are the same thing.

geekmaster
06-15-2012, 02:04 PM
As far as how the English language is used, yes there is a difference between memory and storage. In a certain context [computer hardware], they are the same. But as most [nontechnical] people use the term, they are not. If you do a search of "memory vs. storage" there is quite a lot of disagreement that they are the same thing.They are not always used the same. I did not claim that they were. But they are often used the same. I can use them the same in certain contexts, and be correct. You can use them to mean different things, but unless you add extra keywords to tell them apart, they can be confused that way. Those words are not clear distinguishing features of different kinds of devices. We are both right. But the problem is that certain self-proclaimed "language police" insist that I am wrong, and that my way will confuse people.

The point is, I am saying that "memory" and "storage" CAN be used as synonyms, and people are telling me that they CANNOT be used that way (despite a world full of evidence that supports both my usage and your usage).

Extra descriptive words like I used are needed to avoid confusion. That is how English works. You cannot just leave out the extra words and ASSUME that "memory" and "storage" have unique specific and distinctive meanings when used in a computer hardware context (especially when they are commonly used as synonyms).

Claiming that these words each mean ONLY different things, the way "most" people use them, does not make that the only correct usage, and in fact that way is much more confusing for people who understand the technology with all its history, variations, and overlap, and the evolution of new devices that completely blur these distinctions. Early computers typically used only one kind of "memory storage" device (except when a different "job" needed to be run). Even the CPU registers were stored on drum memory at one time, and core memory was permanent storage (programs rarely reloaded or changed). Future computers are going back to that (RAM and hard drives replaced by large quantities of very fast memory that never forgets, even with the power turned off -- no need for external memory storage devices).

WT Sharpe
06-15-2012, 05:16 PM
To who it may concern:

When placing users on ignore, please don't announce it publicly.

Thank you.

— Moderator

geekmaster
06-15-2012, 06:58 PM
If we look at the definition of "memory" from a non-technical point of view, memory is long-term storage in the human brain of information not being processed now (like hard disk storage). "Storage" refers to how and where information is stored in a memory device, such as in register storage, variable storage, database storage, flat file storage, archival storage, etc. It may also refer to information storage and retrieval hardware, such as core memory storage devices, ROM storage devices, flash memory storage devices, etc.

Because the non-technical term of "memory" generally refers to long-term permanent storage and not information currently being processed in your thoughts, this definition is also commonly used in computer documentation as well. However, the OTHER definition (of memory containing what is currently being processed) is also common, which makes memory understood when referring to any storage device (RAM, disk, or otherwise). Which means they are pretty much interchangeable in technical documentation (as can be seen in wikipedia and other places).

You cannot say that I have them backwards and I will confuse people, because I used it the way non-technical people understand human memory, and the way it is commonly used in technical documentation as well. Saying that general use should be the opposite of the human memory usage is what is confusing.

ThomasC
06-15-2012, 07:54 PM
To who it may concern:

When placing users on ignore, please don't announce it publicly.

Thank you.

Moderator

Sorry about that, I did not know. Won't happen again.

It's a very nice feature though, lol!

geekmaster
06-17-2012, 01:12 AM
The Kindle Touch has 256MB of DRAM memory according to the teardowns and specifications of that part. The Kindles before the third/keyboard have 128MB. The Kindle OS probably takes about 20-50 of it on its own, the rest for the reading, browsing, etc "apps" if you call them.That estimate is not accurate for modern kindles. The kindle OS consists of the linux kernel, and the framework (the GUI) and all supporting device drivers and background processes. On the K5 (touch), this includes a very large Xorg graphics layer. I did very thorough memory testing when the K5 was first released. It uses MOST of the RAM just for itself, leaving little left for apps and books. Your 20-50MB estimate is extremely short of reality, especially on the newer kindles.The Kindle storage on the other hand varies from 2-4GB depending on the model. That is where your books and whatever else you keep in there is. However, as you note, when they tell you it's 2GB, and it's only 1.4GB, it's not just because of the fancy number explanations they give you about the bytes and rounding, etc, it's because they've allocated the remaining 600MB for the OS, drivers, firmware updates, and virtual memory (which the Kindles must have very little of). So when you open up that 50MB PDF file, it uses a little bit of your RAM just to open it, and a little bit of virtual memory/that allocated hard drive space to store those images until you either turn off the Kindle completely or it expires (depending on how Amazon handles that anyway). Think of it as your internet browser cache which saves most of the GUI of websites so it can render them faster as you navigate and go back to pages the next day.I disagree. What you call "storage" is actually the "USB Drive" partition (/dev/mmcblk0p4 on the K4 and K5) of the onboard mmc memory storage device. The kindle does not even contain the hard drive that you mention above. Nor does it contain "virtual memory", which in linux exists in a swap file or swap partition. The kindle has neither of these, because virtual memory swap files are very hard on mmc devices and wear them out prematurely. That means that the very small amount of RAM not already occupied by the OS is all there is.So when you get that error that your Kindle doesn't have enough memory to render the rest of the pages, it's because there is not enough space VIRTUALLY to handle the images and embedded text in that PDF. It's not a problem with Amazon bought ebooks because they have guidelines in place that don't let authors use images bigger than 127KB in size.First, there is no virtual memory in the kindles. Second, the 127KB limit you refer to is the maximum image size allowed in "the Kindle File Format" (.mobi), which has nothing to do with the PDF files being discussed here. It would be helpful and much less confusing if you supplied supporting evidence, references, or links, instead of just claiming that I am "wrong". I would like to examine your evidence so that I can update my own personal knowledge base, if your information is true.

I have read the technical manuals from amazon, and the reference manuals from the chipset manufacturers, and the source code for all of the kindle models that I own (DX,DXG,K3,K4,K5). I have also written a significant base of original code that works on ALL of the eink kindle models, and is being used by other developers as well. I have also discovered jailbreak methods that are in reserve (shared with other developers) that will be published when existing known methods no longer work. I do not see any evidence that your information is better than mine (or even correct). Sorry.

geekmaster
06-17-2012, 01:34 AM
It is an attitude like that which made me, as well as a lot of other people really hate our teachers in secondary school. You can read a book, but it doesn't mean you'll understand beyond what the text says, except what it does, especially in a dictionary."Teacher-hater"? I do not understand. My teachers were there to help me teach myself. They provided me with guidance, and they gave direction to my self-studies. In most cases they appreciated that I finished reading the provided materials in the first two weeks of class and they provided additional reading materials and advanced studies. In the rare case that a teacher did not let me advance ahead of the rest of the class, I had a difficult time with boredom and daydreaming (I had already read the book that the teacher was describing day after day), but I did not hate the teacher just because they did not understand me.

Thankfully, most teachers took the effort to teach me at my own pace (and often gave me extra work and extra priveledges such as a key to the science tool and supply storeroom so I could do my own experiments). I loved working with those teachers best. I did not understand until much later in life that some children did not have these learning opportunities available to them, and I feel sorry for that.

I try to help people young and old learn the things that I know and love whenever I can (despite my occasional attitude that some who do not understand me think of as arrogance, and I am sorry for that misunderstanding).

Anyway, back on topic, an ebook that has a large file size on the kindle USB drive is more likely to have "not enough memory" when trying to render pages to the display, than an ebook that has a smaller file size. That is all I was trying to say, when I was so rudely interrupted by the "experts".

HarryT
06-17-2012, 09:29 AM
Anyway, back on topic, an ebook that has a large file size on the kindle USB drive is more likely to have "not enough memory" when trying to render pages to the display, than an ebook that has a smaller file size.

I'm a little puzzled by this, GM. Surely all that matters is the size of the individual page, not the overall size of the file, is it not? All that's loaded into memory is the page currently being displayed, not the whole book.

geekmaster
06-17-2012, 10:49 AM
I'm a little puzzled by this, GM. Surely all that matters is the size of the individual page, not the overall size of the file, is it not? All that's loaded into memory is the page currently being displayed, not the whole book.The size of the file is indicative of the size of the page. Average page size = file size / number of pages. I originally asked about the file size, to get an idea of how much RAM may be required by that book.

Some books have very large pages, due either to graphic complexity (especially books drawn in a "comic book" style, and books that were not OCR'd where every page is a photo containing text). I have a large collection of PD books going back for decades, and some of them are of the "all images" format.

Memory issues with complex books are especially problematic on newer kindles, that have much larger OS software, and much less for the applications that render PDF files.

HarryT
06-17-2012, 10:52 AM
The size of the file is indicative of the size of the page. Average page size = file size / number of pages.


What I meant was that a large file size can simply mean that the book has a large number of pages, not necessarily that each individual page has a large size. One has to consider both the size of the file AND the number of pages in the book. Simply looking at the file size in isolation is only giving you one of the variables in the equation.

geekmaster
06-17-2012, 11:15 AM
Here is a simpler explanation about how the words "memory" and "storage" are COMMONLY used (in my experience):

As verbs, "memorizing" and "storing" are often used as synonyms. As nouns, "memory" and "storage" are often used as synonyms, indicating a place to store information for later use.

In common use by non-technical people, "memory" is all the stuff you know but are not thinking about now. In common use in computer literature, memory is all the information and data that a computer knows but is not currently processing (stored in CPU registers, RAM, and external archival storage hardware) AND the location used to store this information (including RAM memory, core memory, disk memory, drum memory, etc.), which adds to the confusion. Storage is the process of memorizing this information, and the place that it is stored (also called memory). When using these words it is necessary to include extra words to distinguish them from each other (such as RAM, flash, disk, etc.), because they ARE used in many disparate ways in the trade literature and documentation.

I hope that makes it a little less confusing, and demonstrates that I did initially use these words correctly when asking for information to help determine the cause of the "out of memory" condition when displaying PDF files.

Kindles (especially the newer models) have plenty of flash memory (the USB drive), but not enough RAM memory (especially when rendering large complex PDF files).

pdurrant
06-17-2012, 01:35 PM
The amazon ebook guidelines actually say that all submitted images must be AT LEAST 300 DPI (making them very large), so that amazon can accurately OCR them for searchable text. I do not know where you got your information, but I got mine from this amazon page: http://www.amazon.com/gp/help/customer/display.html?nodeId=13685751
There may well be OTHER amazon guidelines about what ends up INSIDE finished ebooks, but the ebook submission guidlines say what to SUBMIT to amazon.

You have given a link to information about Amazon's Search Inside operation. This is for publishers providing PDFs of their print books so that customers can see samples of the interior pages and, indeed, search inside them.

This has no relation to the guidelines for their Kindle Direct Publishing operation. The limits for images for ebooks can be found here (https://kdp.amazon.com/self-publishing/help?topicId=A1B6GKJ79HC7AN).

geekmaster
06-17-2012, 01:43 PM
You have given a link to information about Amazon's Search Inside operation. This is for publishers providing PDFs of their print books so that customers can see samples of the interior pages and, indeed, search inside them.

This has no relation to the guidelines for their Kindle Direct Publishing operation. The limits for images for ebooks can be found here (https://kdp.amazon.com/self-publishing/help?topicId=A1B6GKJ79HC7AN).Thank you. I missed that when I was searching. Google gives different results for different people. That is why I asked in my previous post.

So it appears that you need to submit ebooks in two formats to amazon, one with large images to be scanned for "search inside" functions, and another with 127KB images in the final ebook format. Is that correct?

pdurrant
06-17-2012, 03:32 PM
So it appears that you need to submit ebooks in two formats to amazon, one with large images to be scanned for "search inside" functions, and another with 127KB images in the final ebook format. Is that correct?

No. Amazon Search inside is a scheme for PAPER books (as mentioned in my previous comment), not ebooks.

If you want to publish ebooks through Amazon, you just need to submit once through the kdp.amazon.com site.

geekmaster
06-17-2012, 04:24 PM
We are getting more reports in the dev corner (including PMs) of K5 (touch) having more cases of PDF files that report "out of memory" errors for some PDF files. Apparently the new 5.1.0 firmware has even LESS available free RAM. This would be less of a problem if there was some virtual memory available, but it is not safe to store virtual memory swap files on flash devices due to premature write wear problems (especially on nearly-full devices which have little space left for write wear-levelling). We can just look forward to this problem getting worse with time, as new firmware versions grow in size.

The new K4 4.1.0 firmware is also larger and uses more RAM.

Unfortunately, you cannot easily increase the RAM, because the CPU is limited to 256MB.

pdurrant
06-17-2012, 04:28 PM
Since the 5.1 and 4.1 firmware are the ones that introduced (I think) support for the KF8 format, I'm not surprised that they take up a bit more RAM.

I suspect that Amazon aren't very worried about PDF support.

But perhaps they'll spend some time on optimisation. 128MB or 256MB is actually quite a lot, if used carefully.

geekmaster
06-18-2012, 07:42 AM
No. Amazon Search inside is a scheme for PAPER books (as mentioned in my previous comment), not ebooks.

If you want to publish ebooks through Amazon, you just need to submit once through the kdp.amazon.com site.The web page at that URL does not restrict the size of images submitted for publishing. It just says that "the Kindle File format" (mobi?) has a limit of 127KB pictures (previous 63KB). We are discussing PDF files here, in which full-page images are poorly compressed and very large (byte-wise), especially when created with a "PDF printer driver". This is why professional-created PDF files often contain vector-artwork (algorithmically rendered from a display list) instead of bitmap photos.

If amazon squeezes a full-page JPEG file down to 127KB to fit it into a .mobi file, it will lose so much detail as to make embedded text unreadable in most cases.

Comparing the maximum image size allowed inside a .mobi file to "out of memory" errors when displaying a PDF file is comparing "apples to oranges", so not relevant to solving this problem.

EDIT: Many of my PDF files contain large schematics (unfortunately, usually bitmaps instead of vector art). Compressing those too much would render the electronic component labels unreadable. They tend to be very large files even when they do not have many pages. I do not have any PDF files that follow the amazon .mobi image size limit guidelines.

UPDATE: Here is a possible SOLUTION to this "PDF out of memory" problem: Many PDF files are designed for high PRINT quality. For onscreen viewing (especially on a Kindle), it would be worthwhile to shrink the PDF file using a "PDF Optimizer". There are many apps that can optimize PDF files, from expensive commercial apps, to free apps such as linux "pdfopt". These PDF optimizers can reduce the amount of disk space needed to hold a PDF file AND can also reduce the amount of RAM needed to display the file.

There are other PDF reader apps that do a better job of memory management and may not have this problem. Unfortunately, you may need to jailbreak your kindle to use them.

WT Sharpe
06-18-2012, 03:43 PM
No. Amazon Search inside is a scheme for PAPER books (as mentioned in my previous comment), not ebooks.

If you want to publish ebooks through Amazon, you just need to submit once through the kdp.amazon.com site.

I'm seeing more and more 'Search inside' functions on Kindle edition pages at Amazon. Whether the function is simply borrowed from the pages for the print editions I can't say.

pdurrant
06-18-2012, 05:21 PM
I'm seeing more and more 'Search inside' functions on Kindle edition pages at Amazon. Whether the function is simply borrowed from the pages for the print editions I can't say.

[OT]
Search Inside for Kindle ebooks just uses the text and formatting of the Kindle ebook. Amazon have access to the full electronic text, after all.

It's only for paper books that publishers need to submit a PDF version. Amazon don't scan paper books to get them in the Search Inside programme.

twowheels
06-19-2012, 10:39 PM
Unless we have an English translation problem, computer "memory" and "storage" are EXACTLY the same thing. The two types of "memory storage" used in the kindles just happen to use different technology: one is faster, and the other is nonvolatile. The latest generations of flash memory are so fast that computers will only have one kind of memory storage in the future, with no artificial separation between fast memory and nonvolatile memory like we use now. They will just have a LOT of fast nonvolatile memory.

True, though there will most likely always remain a separation of transient vs. persistent data. It just so happens that the transient data is currently stored in the faster system RAM and the persistent data is stored on slower magnetic or flash media, but the separation is still important lest you be unable to "reboot" to clear out any state issues, to start over with a clean slate, which is often necessary due to logic bugs in software.

geekmaster
06-21-2012, 08:12 AM
True, though there will most likely always remain a separation of transient vs. persistent data. It just so happens that the transient data is currently stored in the faster system RAM and the persistent data is stored on slower magnetic or flash media, but the separation is still important lest you be unable to "reboot" to clear out any state issues, to start over with a clean slate, which is often necessary due to logic bugs in software.Not always a separation, I think.

When I started, we loaded programs into persistent core memory, and we had to find an "empty spot" (or find a program to "delete" from core memory) to load a new program. We kept a paper chart showing the start and length of all the programs stored in core memory, and to run one, we would set its start address on the front panel switches and press the RUN button.

If you could afford to have a hard disk on your computer, storing programs on it was the same. You needed to know the starting sector and length on disk, and an empty location big enough in core memory, then you could copy bytes between core and disk, or between disk and core, to load or save a program.

Now that flash (and other forms of nonvolatile) memory is getting so fast and cheap (as I mentioned earlier), future computers will have so much persistent fast storage that they will not need a hard drive, and RAM will be obsolete as well.

A computer with terabytes of persistent fast memory will be very nice indeed. And no need to worry about obsolete distinctions between volatile and persistent storage either... :D

booknut
02-02-2013, 03:10 PM
To cut to the chase, I want to put more books on my Kindle Paperwhite, is there a way to reallocate the existing space used to store user content so that there is less for Amazon and more for me?

hunnymonster
02-02-2013, 05:06 PM
To cut to the chase, I want to put more books on my Kindle Paperwhite, is there a way to reallocate the existing space used to store user content so that there is less for Amazon and more for me?

Negative