Predictive tiling

Abstract

This document describes techniques and apparatuses for predictive tiling. These techniques predict tiles of content to pre-render so that an application will be ready to render content quickly in response to a user manipulation. By so doing, these techniques enable fast manipulation of content without unduly impacting drawing resources.

Description

BACKGROUND

Gestures permit users to manipulate content within application interfaces. These gestures permit zooming in or out of a map, panning through a list, and rotating a picture, to name just a few. Conventional techniques for handling gestures, however, often provide a poor user experience or use too many drawing resources.

SUMMARY

This document describes techniques and apparatuses for predictive tiling. These techniques predict tiles of content to pre-render so that an application will be ready to render content quickly in response to a user manipulation. By so doing, these techniques enable fast manipulation of content without unduly impacting drawing resources.

This summary is provided to introduce simplified concepts for predictive tiling that are further described below in the Detailed Description. This summary is not intended to identify essential features of the claimed subject matter, nor is it intended for use in determining the scope of the claimed subject matter. Techniques and/or apparatuses for predictive tiling are also referred to herein separately or in conjunction as the “techniques” as permitted by the context.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments for predictive tiling are described with reference to the following drawings. The same numbers are used throughout the drawings to reference like features and components:

FIG. 1 illustrates an example system in which techniques for predictive tiling can be implemented.

FIG. 2 illustrates an example embodiment of the computing device of FIG. 1.

FIG. 3 illustrates an example embodiment of the remote provider of FIG. 1.

FIG. 4 illustrates an example method for predictive tiling.

FIG. 5 illustrates, for an example image, total content, a viewport, a threshold, and tiles of the total content.

FIG. 6 illustrates a gesture made to an application interface through a touch-screen display and received at viewport.

FIG. 7 illustrates the total content of a news article and viewport and thresholds at a current, intermediate, and final location in the total content.

FIG. 8 illustrates an example method for predictive tiling in part from a perspective of an application responsible for rendering content in an application interface.

FIG. 9 illustrates an example device in which techniques for predictive tiling can be implemented.

DETAILED DESCRIPTION

Overview

Before fully describing these techniques and apparatuses for predictive tiling, this document describes two conventional techniques. The first conventional technique pre-renders all content that is potentially viewable through a manipulated viewport of an application interface. In response to a user manipulation, content is quickly rendered from the pre-rendered content and based on the user's manipulation. A user zooming out to see more of a single picture, for example, may immediately see more of that picture. While this technique often works well for small amounts of potentially viewable content, such as the example single picture, it works poorly or does not work at all in situations where the amount of potentially viewable content is larger than can be easily handled with available drawings resources.

The second conventional technique does not pre-render potentially viewable content. In response to a user's selection to manipulate content, this second technique can result in unrealized areas within the viewport. An unrealized area is part of the viewport that does not immediately render content but instead is blank or shows some sort of graphic but not the desired content. Thus, a user panning through a news article may see a checkerboard graphic until the application's rendering is able to catch up with the user's manipulation. While this technique often avoids using too many drawing resources, it can result in a poor user experience.

The techniques and apparatuses described herein, however, permit applications to quickly render content for a good user experience using fewer drawing resources than pre-rendering the total, potentially viewable content. A user panning through the news article noted above may immediately see new pages of the article, rather than a checkerboard graphic, and without requiring that the user's device use extensive drawing resources. These techniques are often effective in balancing two opposing constraints of many computing devices, namely response speed and drawings resources.

This discussion proceeds to describe an example environment in which the techniques may operate, methods performable by the techniques, and an example apparatus.

Example Environment

FIG. 1 illustrates an example environment 100 in which techniques for predictive tiling can be embodied. Environment 100 includes a computing device 102, remote provider 104, and communication network 106, which enables communication between these entities. Computing device 102 presents an application interface 108 on touch-screen display 110. Application interface 108 includes viewport 112, through which content is displayed to a user and in many case through which a user manipulation is received.

User manipulations may include numerous types of user inputs and input devices, such as using a mouse to click on selectable controls or to make a mouse-based gesture, using one or multi-fingered gestures through a touch-screen display, and using voice activation. User manipulations, whether through gestures or otherwise, may permit zooming in or out (e.g., higher or lower resolution), panning in two dimensions, rotating (in two or three dimensions), transformations, translations, and affine transforms (e.g., a linear transformation and translation together), and others. By way of example, consider user manipulation 114 shown in FIG. 1. User manipulation 114 is a flick-up, single-finger gesture intended to quickly pan through content shown in viewport 112. User manipulation 114 is received at viewport 112 and through touch-screen display 110.

FIG. 2 illustrates an example embodiment of computing device 102 of FIG. 1, which is illustrated with six examples devices: a laptop computer 102-1, a tablet computer 102-2, a smart phone 102-3, a set-top box 102-4, a desktop computer 102-5, and a gaming device 102-6, though other computing devices and systems, such as servers and netbooks, may also be used.

Computing device 102 includes or has access to computer processor(s) 202, computer-readable storage media 204 (media 204), and one or more displays 206, four examples of which are illustrated in FIG. 2. Media 204 includes an operating system 208, predictor 210, and applications 212, each of which is capable of providing a respective application interface 214. In some cases application 212 provides application interface 214 in conjunction with a remote device, such as when the local application is a browser and the remote device includes a network-enabled service provider.

Predictor 210 predicts tiles of content for pre-rendering. Predictor 210, for example, may predict tiles of content for pre-rendering by application 212 thereby enabling application 212 to quickly respond to a user manipulation made through an associated application interface 214 and using fewer drawing resources than some conventional techniques. Furthermore, predictor 210 may do so in real-time and during a current user manipulation.

The techniques are described in the context of square or rectangular tiles of content, though other types and shapes may instead be used. Furthermore, these tiles can be large or small, based on various factors. If very small, the tiles may represent even single-pixel-sized portions of content.

FIG. 3 illustrates an example embodiment of remote provider 104. Remote provider 104 is shown as a singular entity for visual brevity, though multiple providers are contemplated by the techniques. Remote provider 104 includes or has to access to provider processor(s) 302 and provider computer-readable storage media 304 (media 304). Media 304 includes services 306, which interact with users through application interfaces 214 of computing device 102 (e.g., content displayed on display 206 or touch-screen display 110).

Ways in which entities of FIGS. 1-3 act and interact are set forth in greater detail below. The entities illustrated for computing device 102 and/or remote provider 104 can be separate or integrated, such as predictor 210 being integral with or separate from operating system 208, application 212, or service 306.

Example Methods

FIG. 4 depicts a method 400 for predictive tiling. In portions of the following discussion reference may be made to environment 100 of FIG. 1 and as detailed in FIGS. 2-3, reference to which is made for example only.

Block 402 receives a size of a viewport in an application interface through which a portion of total content can be displayed and a size of the total content. These sizes can be actively found by predictor 210 or received from one of applications 212 responsible for rendering content in application interface 214.

By way of example, consider FIG. 5, which illustrates total content 502, viewport 504, threshold 506, and tiles 508 (marked 508-1, 508-2, 508-3, and 508-4). Block 402 receives the sizes of viewport 504 and total content 502. The sizes of viewport 504 and total content 502 can be represented in X-Y coordinates and/or with a number of pixels, or in other formats.

Viewport 504 is one example of a viewport, another of which is shown at 112 in FIG. 1. Both of these viewports display a portion of total content, here viewport 504 displays a portion of total content 502 at a particular time within application interface 214. A user viewing application interface 214, for example, sees two tree trunks within viewport 504 of a larger image have three trees and accompanying foliage.

Block 404 determines, based on the viewport size and the size of the total content, a tile size and a threshold, the threshold including the viewport.

Continuing the ongoing example, predictor 210 determines threshold 506 and tiles 508 based on the sizes of viewport 504 and total content 502. Predictor 210 may also do so based on other information, such a responsiveness of application 212 or a size/availability of drawing resources. The responsiveness of application 212 can be received or determined, as noted elsewhere herein, and be independent or caused by other entities, such as a time to pre-render content along with a time to request and received content from service 306 of FIG. 3 through communication network 106. The size/availability of drawing resources may include a size, allocation, or availability of video memory on a graphics card usable by application 212, for example. This allocation or availability can vary depending on computing device 102's conditions, such as number of other applications demanding drawing resources.

Block 406 determines one or more initial content tiles that include at least a portion of the threshold. Returning to FIG. 5, all of tiles 508 include some portion of threshold 506, though only tile 508-4 includes viewport 504. Here threshold 506 represents a safety buffer or, put another way, a danger zone where application 212 may not be able to immediately display content responsive to a user manipulation. Pre-rendering content within threshold 506, however, often enables application 212 to respond immediately to the manipulation. Thus, a small or initial part of a gesture to display content surrounding viewport 504 within threshold 506 may result in a failure to display that content if content of threshold 506 is not pre-rendered. As noted herein, this can depend on computing device 102, application 212, and current conditions, though the techniques may adjust to these dependencies.

Block 408 indicates, to the application, the one or more initial content tiles effective to enable the application to pre-render the one or more initial content tiles. Block 406 and 408 are directed to initial content tiles pre-rendered to address small manipulations or an initial portion of a larger manipulation. As noted herein, pre-rendering content improves a speed at which the content can be rendered, even for relatively small manipulations. Thus, pre-rendering these tiles of content enables an application to prepare for a small user manipulation or an initial portion of a large user manipulation. Predicting a different location of a threshold or viewport, and thus tiles to pre-render based on this different location, are addressed below.

Block 410 receives information concerning a current user manipulation made to the application interface. This information is useful for determining a location within the total content from which a response to the user manipulation renders content. This location can be represented in various ways, such as singular point within the total content (e.g., a single pixel), a viewport-sized portion within the content, or a threshold-sized portion within the content. As noted below, tiles of content to be pre-rendered are based on this location, with examples described in the content of a threshold-sized portion.

This information can include a velocity, direction, convergence/pinch or divergence/spread rate, and/or inertia of a user manipulation, to name just a few. The example user manipulation 114 of FIG. 1 is a gesture having a velocity, direction, and inertia.

Block 412 predicts, based on the information and a responsiveness of an application responsible for rendering content in the application interface, one or more tiles of content to be pre-rendered. Note that in some cases these predicted tiles are also based on the threshold noted above, though this is not required. Further, these predicted tiles may or may not include the initial content tiles used to prepare for a small or initial portion of a user manipulation noted above. This responsiveness can includes a time needed by application 212 to pre-render content, which can be based on a current or historic behavior of application 212. Thus, predictor 210 may indicate a proportional number of tiles for pre-rendering—a large number if the application is slow and a small number if the application is fast. The responsiveness of a particular application can be received from the application or another entity, or determined by predictor 210, such as based on how often the application makes calls to predictor 210 for tiles.

By way of a second example, consider FIG. 6, which illustrates a gesture 602 made through touch-screen display 110 and received at viewport 604 of application interface 606. As displayed, this gesture is a panning gesture intended to proceed through the news article entitled Weather Satellite May Crash to Earth. This gesture 602's information includes a velocity, inertia, and direction. Consider also FIG. 7, which illustrates total content 702 for the news article partially rendered at viewport 604 in FIG. 6. Viewport 604, in its current location, is also shown in FIG. 7, as well as a current threshold 704 (shown with a dashed-line box).

Because the total content of the news article is no wider than viewport 604, the direction is interpreted as either moving up or down, here moving up to select additional content.

Thus, predictor 210 receives, from the one of applications 212 that is responsible for rendering content in application interface 606, this velocity, inertia, and direction. Predictor 210 then predicts, based on this velocity, inertia, and direction and a responsiveness of application 212, a location within total content for the news article. Either based on a previously determined threshold or a threshold determined at the beginning of receiving the information concerning gesture 602, predictor determines the threshold and location at which a threshold will land in the total content. This threshold is at an intermediate location; another iteration of portions of method 400 can be performed for the same gesture 602 to provide tiles for pre-rendering as the gesture progresses. Intermediate threshold 706 is illustrated in FIG. 7.

Note that current threshold 704 includes a least a portion of tiles 708-1, 708-2, and 708-3 (shown with solid-line rectangles). These tiles can be provided as part of blocks 406 and 408, though that is not required. Assuming that these are provided to application 212, application 212 will have pre-rendered these tiles effective to enable application 212 to quickly respond to gesture 602 in providing content of tile 708-3. The intermediate threshold 706 also includes tile 708-3, but also overlaps with tiles 708-4 and 708-5 (intermediate threshold 706 shown with a dotted-line box).

Block 414 indicates, to the application, the one or more tiles of content effective to enable the application to pre-render the tiles of content in preparation for responding to the current user manipulation. This current user manipulation for which these predicted tiles prepare may or may not include responding to a non-initial portion of the user manipulation, depending on whether or not alternative blocks 406 and 408 have been performed.

Continuing the example illustrated in FIGS. 6 and 7, predictor 210 indicates tiles 708-4 and 708-5 for pre-rendering by application 212 (tile 708-3 was already indicated and pre-rendered). In so doing, application 212 may quickly respond to gesture 602, rendering content at intermediate viewport 710.

As noted in part above, all or portions of method 400 can be repeated for a same user manipulation, if needed. Thus, after rendering content of intermediate viewport 710, application 212 may provide additional information or a new request for a continuation of gesture 602. In response, predictor 210 indicates tile 708-6 based on final a determined, final threshold 712, shown with a dashed-line box in FIG. 7.

By so doing, the techniques are effective to enable an application to repeatedly pre-render, in real-time, until completion of a current user manipulation.

Furthermore, the techniques may cause tiles that are no longer needed to be deleted from computing device 102's drawing resources. Here predictor 210, for example, may cause tiles 708-1 and 708-2 to be deleted after or commensurate with indicating tiles 708-4 and 708-5. Thus, drawing resources are freed up by deleting content of tiles no longer being used (the beginning of the news article when the middle portion is rendered). Further still, when final threshold 712 is determined and final viewport 714 is displaying content at the end of the news article, predictor 210 causes content of tile 708-3 to be deleted. This is but one further way that the techniques enable a good user experience while using fewer drawing resources than some conventional techniques.

Thus, prior repetitions of method 400, which resulted in pre-rendering tiles stored in graphics memory of computing device 102, are deleted from this memory responsive to these tiles no longer being needed.

FIG. 8 depicts a method 800 for predictive tiling in part from a perspective of an application responsible for rendering content in an application interface, such as application 212 of FIG. 2. Method 800 can operate separate from or in conjunction with method 400 or portions thereof.

Block 802 receives, at an application responsible for rendering content for an application interface, an indication of one or more tiles of content to be pre-rendered in preparation for responding to a current user manipulation made to the application interface. As noted above, tiles can be predicted to be useful for responding to the current user manipulation based on information about the current user manipulation and a responsiveness of the application.

Block 804 pre-renders the one or more tiles. This pre-rendering prepares content such that the content can quickly be rendered in a viewport of an application interface. Assume, for example, that a user is viewing, through viewport 112 of FIG. 1, a social networking website. The content rendered into viewport 112 is rendered by a web browser from content provided by service 306 of FIG. 3 and through application interface 108. The user then selects the above-described flick-up, single-finger gesture shown as user manipulation 114 of FIG. 1. The web browser provides information about this gesture to predictor 210. Predictor 210 then provides tiles for pre-rendering, which the web browser receives at block 802. At block 804, the web browser pre-renders the content of these one or more tiles.

Block 806 renders at least a portion of the pre-rendered content of the one or more tiles. This content is rendered in a viewport of the application interface and in response to a current user manipulation. For the ongoing example, in response to the current user manipulation, the web browser quickly renders additional social-networking content.

Method 800 can also be repeated, including for the same user manipulation, effective to enable the application to repeatedly pre-render, in real-time, until completion of the current user manipulation. Whether repeated or not, previously stored tiles (e.g., pre-rendered tiles provided as part of method 400 or a prior iteration of method 800), can be deleted from memory. In one case a tile is deleted when it no longer includes currently rendered content within the application interface or a portion of a threshold that includes the currently rendered content.

Other actions can be performed prior to or following method 800, such as providing a size of a viewport and total content and receiving initial tiles for pre-rendering. These initial tiles enable the application to quickly respond to an initial part of a user manipulation or a small manipulation.

The preceding discussion describes methods relating to predictive tiling. Aspects of these methods may be implemented in hardware (e.g., fixed logic circuitry), firmware, software, manual processing, or any combination thereof. A software implementation represents program code that performs specified tasks when executed by a computer processor. The example methods may be described in the general context of computer-executable instructions, which can include software, applications, routines, programs, objects, components, data structures, procedures, modules, functions, and the like. The program code can be stored in one or more computer-readable memory devices, both local and/or remote to a computer processor. The methods may also be practiced in a distributed computing mode by multiple computing devices. Further, the features described herein are platform-independent and can be implemented on a variety of computing platforms having a variety of processors.

These techniques may be embodied on one or more of the entities shown in environment 100 of FIG. 1 including as detailed in FIG. 2 or 3, and/or example device 900 described below, which may be further divided, combined, and so on. Thus, environment 100 and/or device 900 illustrate some of many possible systems or apparatuses capable of employing the described techniques. The entities of environment 100 and/or device 900 generally represent software, firmware, hardware, whole devices or networks, or a combination thereof. In the case of a software implementation, for instance, the entities (e.g., predictor 210, applications 212, and services 306) represent program code that performs specified tasks when executed on a processor (e.g., processor(s) 202 and/or 302). The program code can be stored in one or more computer-readable memory devices, such as media 204, provider media 304, or computer-readable media 914 of FIG. 9.

Example Device

FIG. 9 illustrates various components of example device 900 that can be implemented as any type of client, server, and/or computing device as described with reference to the previous FIGS. 1-8 to implement techniques for predictive tiling. In embodiments, device 900 can be implemented as one or a combination of a wired and/or wireless device, as a form of television client device (e.g., television set-top box, digital video recorder (DVR), etc.), consumer device, computer device, server device, portable computer device, user device, communication device, video processing and/or rendering device, appliance device, gaming device, electronic device, and/or as another type of device. Device 900 may also be associated with a user (e.g., a person) and/or an entity that operates the device such that a device describes logical devices that include users, software, firmware, and/or a combination of devices.

Device 900 includes communication devices 902 that enable wired and/or wireless communication of device data 904 (e.g., received data, data that is being received, data scheduled for broadcast, data packets of the data, etc.). The device data 904 or other device content can include configuration settings of the device, media content stored on the device, and/or information associated with a user of the device. Device 900 includes one or more data inputs 906 via which any type of data, media content, and/or inputs can be received, such as human utterances, user-selectable inputs, messages, gestures, music, television media content, recorded video content, and any other type of data received from any content and/or data source.

Device 900 also includes communication interfaces 908, which can be implemented as any one or more of a serial and/or parallel interface, a wireless interface, any type of network interface, a modem, and as any other type of communication interface. The communication interfaces 908 provide a connection and/or communication links between device 900 and a communication network by which other electronic, computing, and communication devices communicate data with device 900.

Device 900 includes one or more processors 910 (e.g., any of microprocessors, controllers, and the like), which process various computer-executable instructions to control the operation of device 900 and to enable techniques for predictive tiling. Alternatively or in addition, device 900 can be implemented with any one or combination of hardware, firmware, or fixed logic circuitry that is implemented in connection with processing and control circuits which are generally identified at 912. Although not shown, device 900 can include a system bus or data transfer system that couples the various components within the device. A system bus can include any one or combination of different bus structures, such as a memory bus or memory controller, a peripheral bus, a universal serial bus, and/or a processor or local bus that utilizes any of a variety of bus architectures.

Device 900 also includes computer-readable storage media 914, such as one or more memory devices that enable persistent and/or non-transitory data storage (i.e., in contrast to mere signal transmission), examples of which include random access memory (RAM), non-volatile memory (e.g., any one or more of a read-only memory (ROM), flash memory, EPROM, EEPROM, etc.), and a disk storage device. A disk storage device may be implemented as any type of magnetic or optical storage device, such as a hard disk drive, a recordable and/or rewriteable compact disc (CD), any type of a digital versatile disc (DVD), and the like. Device 900 can also include a mass storage media device 916.

Computer-readable storage media 914 provides data storage mechanisms to store the device data 904, as well as various device applications 918 and any other types of information and/or data related to operational aspects of device 900. For example, an operating system 920 can be maintained as a computer application with the computer-readable storage media 914 and executed on processors 910. The device applications 918 may include a device manager, such as any form of a control application, software application, signal-processing and control module, code that is native to a particular device, a hardware abstraction layer for a particular device, and so on.

The device applications 918 also include any system components, engines, or modules to implement techniques for predictive tiling. In this example, the device applications 918 include predictor 210 and application 212.

CONCLUSION

Although embodiments of techniques and apparatuses for predictive tiling have been described in language specific to features and/or methods, it is to be understood that the subject of the appended claims is not necessarily limited to the specific features or methods described. Rather, the specific features and methods are disclosed as example implementations for predictive tiling.

Claims

1. A computer-implemented method comprising: receiving information concerning a current user manipulation made to an application interface implemented by a computing device;predicting one or more tiles of content to be pre-rendered, the predicting comprising an amount and location of the one or more tiles to be pre-rendered based on: the information to determine the location;an availability of drawing resources on the computing device to use for the pre-rendering of the one or more tiles of content to determine the amount of tiles to be pre-rendered in the location determined from the information such that the amount is larger when the computing device has more resources available and smaller when the computing device has less resources available; anda responsiveness of an application responsible for rendering content in the application interface to further determine the amount of tiles to be pre-rendered in the location determined from the information such that the amount is larger when the application is more responsive and smaller when the computing device is less responsive; andindicating, to the application, the one or more tiles of content based on the prediction effective to enable the application to pre-render the tiles of content in preparation for responding to the current user manipulation.

2. A computer-implemented method as described in claim 1, further comprising repeating the method one or more times until completion of said current user manipulation.

3. A computer-implemented method as described in claim 2, wherein a prior repetition of the method resulted in pre-rendering of one or more prior, pre-rendered tiles, the one or more prior, pre-rendered tiles stored in memory, and further comprising deleting from the memory the one or more prior, pre-rendered tiles responsive to the one or more prior, pre-rendered tiles no longer including currently rendered content within the application interface or a threshold that includes the currently rendered content.

4. A computer-implemented method as described in claim 1, further comprising, prior to predicting the one or more tiles of content to be pre-rendered, determining a size of the tiles of content based on a total content size and a viewport size.

5. A computer-implemented method as described in claim 1, wherein the information includes a convergence/pinch rate or divergence/spread rate of the current user manipulation.

6. A computer-implemented method as described in claim 1, wherein predicting the one or more tiles of content includes determining a location in a total content based on a velocity or an inertia of the user manipulation, the velocity or inertia included within the information, the one or more tiles of content including the location or adjacent to the tile of content that includes the location.

7. A computer-implemented method as described in claim 6, wherein the location is an intermediate location intermediate to a final location at which the current user manipulation completes.

8. A computer-implemented method as described in claim 1, wherein the responsiveness of the application includes a time to pre-render content and predicting an amount of the one or more tiles of content to be pre-rendered is predicted proportional to the time to pre-render content.

9. A computer-implemented method as described in claim 8, further comprising determining the time to pre-render content based on a timing of calls received from the application.

10. A computer-implemented method as described in claim 1, wherein the availability of drawing resources is based on a number of other applications demanding drawing resources concurrent with the application or a quantity of drawing resources being demanded by the other applications.

11. A computer-implemented method as described in claim 1, wherein the availability of drawing resources is based on a size, an allocation, or an availability of video memory.

12. A computer-implemented method comprising: receiving, at an application responsible for rendering content for an application interface implemented by a computing device, an indication of one or more tiles of content to be pre-rendered in preparation for responding to a current user manipulation made to the application interface, the one or more tiles predicted to be useful for responding to the current user manipulation comprising an amount and location of the one or more tiles to be pre-rendered based on information about the current user manipulation to determine the location, an availability of drawing resources on the computing device able to be utilized for the pre-rendering to determine the amount of tiles to be pre-rendered in the location determined from the information such that the amount is larger when the computing device has more resources available and smaller when the computing device has less resources available, and a responsiveness of the application to further determine the amount of tiles to be pre-rendered in the location determined from the information such that the amount is larger when the application is more responsive and smaller when the computing device is less responsive;pre-rendering the one or more tiles based upon the prediction; andrendering, in response to the current user manipulation, at least a portion of the pre-rendered content of the one or more tiles in a viewport of the application interface.

13. A computer-implemented method as described in claim 12, further comprising, prior to receiving the indication, providing the information to a predictor from which the indication of the one or more tiles of content to be pre-rendered is received, the information having a velocity or inertia of the current user manipulation.

14. A computer-implemented method as described in claim 12, wherein the information includes a velocity or inertia of the user manipulation and the responsiveness of the application is based at least in part on a time to pre-render content by the application.

15. A computer-implemented method as described in claim 12, further comprising, prior to receiving the indication of one or more tiles of content to be pre-rendered, providing a size of the viewport and a size of total content to a predictor and receiving an indication of initial tiles for pre-rendering, the initial tiles including content adjacent to content currently presented in the viewport.

16. A computer-implemented method as described in claim 12, further comprising: receiving an indication of previously pre-rendered tiles to delete from memory, the previously pre-rendered tiles not including content currently rendered in the viewport; andresponsive to receiving the indication, deleting the previously pre-rendered tiles from the memory.

17. A computer-implemented method as described in claim 12, wherein the availability of drawing resources is determined based on a number of other applications demanding drawing resources concurrent with the application or a quantity of drawing resources being demanded by the other applications.

18. A computer-implemented method as described in claim 12, wherein the availability of drawing resources is determined based on a size, an allocation, or an availability of video memory on a graphics card usable by the application for rendering content in the application interface.

19. A computer-implemented method as described in claim 12, wherein the responsiveness of the application is based on an historical time needed by the application to pre-render content for display in the application interface or how often the application interface requests content for display from the application.

20. A computing device comprising: one or more computer processors; andone or more computer-readable storage media having instructions stored thereon that, responsive to execution by the one or more computer processors, perform operations comprising: receiving, from an application associated with an application interface implemented on the computing device, a size of a viewport in the application interface through which a portion of total content can be displayed and a size of the total content;determining, based on the viewport size and the size of the total content, a tile size and a threshold, the threshold including the viewport;determining one or more initial content tiles that include at least a portion of the threshold;indicating, to the application, the one or more initial content tiles effective to enable the application to pre-render the one or more initial content tiles to prepare for an initial portion of a user manipulation;receiving information concerning the user manipulation made to the application interface;predicting a location of one or more tiles of content to be pre-rendered based on the information and the threshold, the one or more tiles of content not including the initial content tiles but including at least a portion of a predicted threshold location;predicting an amount of the one or more tiles of content to be pre-rendered in the location based on an availability of drawing resources on the computing device for pre-rendering content including an amount of video memory available for the application and a responsiveness of the application, the responsiveness based on an historical amount of time that the application has taken to pre-render content or an expected current amount of time that the application takes to pre-render content; andindicating, to the application, the one or more tiles of content effective to enable the application to pre-render content of the one or more tiles of content in preparation for responding to a non-initial portion of the user manipulation.