Recommended Posts

- NIGHTLY is now 34.
- Firefox for Android dropped code relevant to Android 2.2.
- Working on Service Workers
- MicroData API is finally dropped from W3C, so I think Firefox might remove it.
- Work on EME and OpenH264 is also in full momentum.
- Recent work on JS Strings in Firefox JS Engine - https://blog.mozilla.org/javascript/2014/07/21/slimmer-and-faster-javascript-strings-in-firefox/ 
 
Few landings:
 
Enhanced Tiles - New Tab Page
https://bugzilla.mozilla.org/show_bug.cgi?id=1036288
 
Firefox Translation
https://bugzilla.mozilla.org/show_bug.cgi?id=1027024
 
[MemShrink] Avoid some allocations - https://bugzilla.mozilla.org/show_bug.cgi?id=1039162
Various cleanups in Ion and Odin Monkey codes - https://bugzilla.mozilla.org/show_bug.cgi?id=1040785
[LANDED & BACKED OUT] Enable OMTC on Linux with Basic Compositor - https://bugzilla.mozilla.org/show_bug.cgi?id=994541
[bACKED OUT] Enable password sync with FxA and master password - https://bugzilla.mozilla.org/show_bug.cgi?id=1013064
Scroll issue through thumb drag on profile manage (with more than 5 profiles) - https://bugzilla.mozilla.org/show_bug.cgi?id=1041207
getBoundingClientRect on range with scaled (transformed) element returns wrong offset - https://bugzilla.mozilla.org/show_bug.cgi?id=863618
[REGRESSION] Opacity does not work with box-shadow, unless the background-color set - https://bugzilla.mozilla.org/show_bug.cgi?id=1003425
OdinMonkey: add async stack-walking support - https://bugzilla.mozilla.org/show_bug.cgi?id=1027885
Enabled GC: Exact Stack Rooting everywhere (B2G as well) - https://bugzilla.mozilla.org/show_bug.cgi?id=753203
Implement support for font-variant-position fallback behavior - https://bugzilla.mozilla.org/show_bug.cgi?id=1024804
Improving USS (Unique Set Size) Gecko Profiler performance on Linux - https://bugzilla.mozilla.org/show_bug.cgi?id=1035396
Redraw issue on tab switching on full page zoomed in - https://bugzilla.mozilla.org/show_bug.cgi?id=1041200
[MEMSHRINK] JS array elements clownshoes [COULD BE BACKED OUT] - https://bugzilla.mozilla.org/show_bug.cgi?id=1039965
[bACKEDOUT] oes_texture_float test failures fix - https://bugzilla.mozilla.org/show_bug.cgi?id=1041830
TextureClient cleanup - https://bugzilla.mozilla.org/show_bug.cgi?id=1040028
Remove Latin1 strings flag - https://bugzilla.mozilla.org/show_bug.cgi?id=1041469
Nightly 33.0a1 crashes on ANGLE_instanced_arrays WebGL demo - https://bugzilla.mozilla.org/show_bug.cgi?id=1041785

Layout performance improvement on nested flexboxes (reflow fix) - https://bugzilla.mozilla.org/show_bug.cgi?id=946167

[LANDED & BACKEDOUT] OMTC flashes in Addon Manager on "Search" text - https://bugzilla.mozilla.org/show_bug.cgi?id=1015718
 
Image decoding (and handling more scaled images) [API refactoring] - https://bugzilla.mozilla.org/show_bug.cgi?id=1031576https://bugzilla.mozilla.org/show_bug.cgi?id=1034209
 
Code cleaups - https://bugzilla.mozilla.org/show_bug.cgi?id=1037100 ; https://bugzilla.mozilla.org/show_bug.cgi?id=1037103 ; https://bugzilla.mozilla.org/show_bug.cgi?id=1037686
 
Add back "min-width:auto" / "min-height:auto" for flex items - https://bugzilla.mozilla.org/show_bug.cgi?id=984711 ; https://bugzilla.mozilla.org/show_bug.cgi?id=1037177 ; https://bugzilla.mozilla.org/show_bug.cgi?id=1015474

  • Like 1

I read about JSON-LD then instantly purged it from my memory, I wonder why :laugh:

I'm not sure why I find RDFa so funny, might be because of the early days where the only way to embed it into HTML involved wrapping it in a comment block, making it completely opaque to browsers. RDFa Lite looks much better though (It's a superset of Microdata and a subset of full RDFa)

Firefox users don't know is that it is possible to configure the browser so that all video resolutions are displayed on YouTube when users connect to the site.

 

  1. Load about:config in the browser's address bar and hit enter.
  2. Confirm that you are careful when the prompt is displayed.
  3. Search for media.mediasource.enabled. The preference is set to false by default.
  4. Double-click the preference name to set it to true and enable it.
  5. You may need to restart the browser before the change takes effect.

 

Note: If you have Adobe Flash installed, you may need to switch to YouTube's HTML5 player. To do so, load https://www.youtube.com/html5 and click on the "Request the HMTL5 player" button.

That gets you WebM playback, but it also doesn't work due to the quality adaption being broken (Or more precisely, Firefox breaks it due to discarding bits of the stream when it should actually be playing them, causing playback to stall)

If you enable MSE and disable auto quality it might work then (And most likely cause CPU usage to skyrocket and stop Firefox from exiting properly due to bugs).

Edit: Also, support for H.264 and AAC playback landed in the latest nightly, not ready for the prime time yet (H.264 playback instantly crashes for me, and AAC works but requires the new MP4 demuxer which also has issues and likes to break) Basically while they're working on making this all work nicely, and for the most part it does work, they're also uncovering new issues on an almost daily basis, so it's still not at a state where they can enable it for testing.

You wanna share your settings maybe?

Well you can go two routes. You can either stick with Windows built in, leave DirectWrite enabled (which is stuck with Windows font rendering regardless), and install something like Anti-Aliasing Tuner, it's on the Firefox extension site, fiddle with the settings until you get it where you like it.  All in all the results are decent once you play with it a bit.

https://addons.mozilla.org/en-US/firefox/addon/anti-aliasing-tuner/

 

Or, you can replace it entirely with something else (MacType, GDI++, etc etc), disable DirectWrite and go for broke. I went with the latter (MacType as a service) as it works with near any Windows desktop application, not just the browser. Lots of presets or you can fiddle with the numbers.

 

For example, Firefox on Windows 7, although I usually use Chrome now:

fonts.png

MacType/GDI++ are going to be slower than DirectWrite and give you lower quality output (GDI style glyph positioning), what you really want is to keep using DirectWrite but disable hinting, which can't be done at the moment unfortunately.

MacType/GDI++ are going to be slower than DirectWrite and give you lower quality output (GDI style glyph positioning), what you really want is to keep using DirectWrite but disable hinting, which can't be done at the moment unfortunately.

Not denying it but never really noticed any performance hits myself although I know there's no font acceleration, "feels" identical and yet looks 1000% better, at least on this particular setup, probably wouldn't say the same on a lower-end setup.  Of course there's mileage, variance and all that, I just prefer it as it covers (almost) all programs running.  And yea, there's a bunch of fiddling you can do with Firefox's antialiasing but no settings for hinting though.. can definitely improve the out-of-the-box settings though, but again personal preference... I know some people who prefer the Windows renderer but *shrug*.  

Windows has like 4 different font renderers built in, so what most people refer to is GDI (Which is really awful in nearly every regard), DirectWrite is much closer to what OS X and FreeType does (Actually exceeds it in most cases), only difference is that OS X ignores hinting, while DirectWrite doesn't (And Freetype can do both) You also have problem that a lot of Windows fonts are actually designed for how GDI renders, so they look odd on other renderers.

It's a rather interesting (yet annoying) subject, it really boils down to 2 things the renderers do differently, sub-pixel positioning and hinting. Freetype is the most flexible as it can do pretty much whatever the host app wants (Which is a problem, as barely any of them implement sub-pixel positioning) and can do anything from no hinting (What OS X does) to full blown hinting like GDI does or even hint the font itself while rendering. DirectWrite is kind of a middle of the road renderer, it does really good sub-pixel positioning (Something like 1/48th of a pixel), but also does partial hinting (It only hints in one direction, because you can't have hinting and sub-pixel positioning) and can even operate without hinting like OS X/Freetype (While GDI would render unhinted characters awfully)

And a rather annoying side effect, is that applications written to use one API depend on how that API renders, so something like MacType hooking GDI rendering need to match how GDI renders characters, it can't extend outside the area GDI would touch as it would give you rendering issues (That was actually a huge problem with the initial releases of GDI++, text would either be bunched up to match GDI, or overflow the bounds and lead to invalidation problems)

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Posts

    • AMD RX 9070 GRE AI, Blender benchmarks vs 9070 XT, 7800XT, Nvidia RTX 5070, 4070 by Sayan Sen Earlier this week, we shared the first part of our review of AMD's new RX 9070 GRE. It was about the gaming performance of the GPU, and we gave it an 8 out of 10. As a follow-up, similar to how we did with the 9070 XT and non-XT, we are doing a dedicated productivity review for the RX 9070 GRE as well, where we compare it against the 9070 XT, 9070, 7800 XT, as well as Nvidia's 5070 and 4070. This will include AI, rendering, compute, and more benchmarks. AI performance, especially, is a very important metric in today's world, and AMD also promised big improvements thanks to its underlying architectural improvements. We will be pitching it against the data we already have for the RX 9070, and RX 9070 XT, but also the Nvidia 5070 FE, MSI GeForce RTX 4070 VENTUS 2X 12G, and Gigabyte Radeon RX 7800 XT GAMING OC 16G as they are in a similar price class, but also because we do not have a comparable 5060 Ti card lying around here that we can compare it against. Before we get underway, this is a collaboration between Sayan Sen and Steven Parker, who lent me his test bed. Also, there was no editorial input from AMD. First up, the specs of the RX 9070, 9070 XT, and 9070 GRE, which were given to us by AMD: Radeon RX 9070 GRE Radeon RX 9070 Radeon RX 9070 XT Boost Clock: Game Clock: up to 2.79GHz up to 2.20GHz up to 2.52GHz up to 2.07GHz up to 2.97GHz up to 2.40GHz Stream Processors 3,072 (48 CU) 3,584 (56 CU) 4,096 (64 CU) Ray Accelerator 48 56 64 AI Accelerator 96 112 128 ROPs 96 128 Texture Mapping Units 192 224 256 Memory 12 GB GDDR6, 18Gbps Clock, 192-bit Bus 432 GB/s 16 GB GDDR6, 20Gbps Clock, 256-bit Bus Effective Memory Bandwidth: 640 GB/s Infinity Cache 48 MB (3rd Gen) 64 MB (3rd Gen) Card Bus PCI-E 5.0 X16 Output 2x HDMI 2.1b 2x DisplayPort 2.1a Power consumption 220W 304W Recommended PSU 650W 750W Slot width 2x 3x Price (SEP) $549 $599 As you can see from the specs above, it is less than the standard RX 9070 in every way that counts, except for slightly higher Boost and Game clock speed. Design Moving on, the RX 9070 GRE we were given is an XFX Swift triple-fan, dual-slot design with two 8-pin connectors. At 30cm (self-measured), it will fit in most systems easily. There is no RGB either. The AMD Radeon RX 9070 GRE by XFX from all angles. Test system Our test system consists of the following: Lian Li O11 Dynamic Mini V2 Flow (Amazon|Newegg) ASUS Z890 ProArt Creator WiFi (Amazon|Newegg) Intel Core Ultra 7 270K Plus (Amazon|Newegg) Thermal Grizzly KryoSheet - 44x37 (Amazon|Newegg) 2x 16GB G.Skill Trident Z5 RGB (7200 MT/s in XMP) (Amazon|Newegg) Sabrent Rocket4 Plus 2TB SSD (Amazon) Windows 11 25H2 (Build 26200.8246) AMD shared a press driver based on the recently released Adrenaline 26.5.2 that we were required to use. We now move on to our benchmarks. First up, we have Geekbench AI running on ONNX. For some reason, the 9070 GRE does exceptionally well here in both half-precision (FP16) and single-precision (FP32). It manages to beat the RTX 5070 and RX 9070 non-XT, and is only behind the 9070 XT. Since Geekbench runs in short bursts instead of continuously hammering the graphics card, it seems the GRE's faster boost clocks are helping here. Next up, we move to the UL Procyon AI test suite, starting with the image generation benchmark. We chose the Stable Diffusion XL FP16 test since it is the most intense workload available on Procyon. The Nvidia cards do very well here, as even the 4070 out-muscles AMD's best fairy easily. The positive thing about the GRE is that it gets quite close to the 9070 non-XT in this test; this indicates that the VRAM does not play a very big role here, as SD XL relies on float16 (FP16). So this is something to keep in mind again. If you wish to work with float32 AI workloads, graphics cards with larger than 12 GB buffers would likely emerge as victors. Regardless, the gains are still massive on AMD's 9000 series compared to the 7000 series. Following image generation, we move to the text generation benchmark. This is one test where the 9070 GRE struggled, quite a lot. It seems that the 12 GB VRAM and lower memory bandwidth of the new Radeon 9070 GRE are hurting it quite a bit; the split is massive, especially in a test like Llama2, which packs 13 billion parameters. As such, in all the tests, the 9070 GRE is the slowest of the lot. Next, we tried Blender, and here the AMD GPUs were beaten by Nvidia. Rendering is something the Green team has always had a lead over the Red side, and it has not changed so far. On the positive side, though, the 9070 GRE shows significantly better results than the 7800 XT, which means AMD is on the right path. Catching up to Nvidia, though, will require a lot more effort. And we hope HIP and ROCm can keep improving. Wrapping up AI testing, we measured OpenCL throughput in the Geekbench compute benchmark. The RX 9070 GRE alongside the 9070 did not fare well here at all, even falling behind the 7800 XT. Interestingly, even the RTX 5070 could not beat the 4070 on OpenCL, so perhaps this suggests that OpenCL optimization may not have been a priority for either AMD or Nvidia in the modern era. Conclusion We reached the end of our productivity performance review of the 9070 GRE, and we have to say it's a mixed bag. Unlike the 9070 and 9070 XT, the GRE excels in some areas while losing ground fairly easily in others. Similar to how it happened in gaming, any time the card's memory subsystem gets hammered, it tends to fall behind the others. This was the case with text generation, wherein we saw the VRAM sometimes hit its maximum available 12 GB of usage with larger model sizes. So what do we make of the RX 9070 as a productivity hardware? It can certainly be used, but you have to know it has its limitations. For those looking for a GPU that can deal with more, AMD recently unveiled the Radeon AI PRO R9700, which is essentially a 32 GB refresh of the 9070 XT with some additional workstation-based optimizations. On a similar note, the new Ryzen AI Halo platform is something you can consider if you want to set up a local AI processing station. Considering everything, we rate AMD's Radeon RX 9070 GRE a 7.5 out of 10 for its productivity performance. Price is less of a factor for those looking at productivity cases compared to those considering the GPU for gaming, and as such, we felt it did quite decently on many occasions and can be handy if you need a 12 GB GPU and, for some reason, don't want to get Nvidia. Purchase links: RX 9070 / XT / GRE (Amazon US) As an Amazon Associate, we earn from qualifying purchases.
    • Does anyone here know if these updates are integrated into the UUP dump isos?
    • Motrix Next 3.9.4 by Razvan Serea Motrix Next is a modern, open-source cross-platform download manager built as the official next-generation successor to the original Motrix project. It has been completely rewritten using Tauri 2, Vue 3, TypeScript, and Rust, while still relying on the powerful Aria2 download engine for high-speed multi-protocol transfers. The app supports HTTP, HTTPS, FTP, BitTorrent, ED2K and magnet links, offering advanced features like multi-connection acceleration, task scheduling, bandwidth control, and batch download management. With a significantly reduced install size (around 20MB), it focuses on being lightweight, fast, and resource-efficient compared to traditional Electron-based download tools. Designed for Windows, macOS, and Linux, Motrix Next delivers a clean, modern UI inspired by Material Design 3 principles, with smooth animations and a minimal workflow. It improves usability through better download organization, system tray integration, and enhanced torrent handling including selective file downloads and tracker management. Motrix Next features: Multi-protocol downloads — HTTP, FTP, BitTorrent, Magnet, .torrent, ED2K, and Metalink tasks BitTorrent — Selective file download, DHT, peer exchange, encryption controls, metadata caching, GeoIP peer flags, and tracker probing Browser extension integration — Embedded Extension API with independent authentication, download confirmation, smart auto-submit, filename hints, referer/cookie forwarding, and real-time controls (Chrome Web Store · Edge Add-ons) Safe filename handling — Content-Disposition, RFC 2047, non-UTF-8, percent-encoded, and extensionless URL resolution with path traversal sanitization Download organization — Favorite and recent folders, optional file-type categorization, stale-record cleanup, and completed history backed by SQLite Concurrent downloads — Independent controls for active tasks, HTTP connections per server, segments per file, and BT peer limits Speed control — Global and per-task upload/download limits with day-of-week and time-of-day scheduling System integration — Tray operation, optional tray speed display, macOS Dock badge/progress, protocol handlers for magnet://, thunder://, and motrixnext:// Lightweight mode — Destroys the WebView on minimize-to-tray while Rust keeps the engine, task monitor, notifications, history, and extension routing alive Notifications and power options — Native task start/complete/failure notifications, keep-awake during downloads, and optional shutdown after completion Network controls — Scoped proxy support for downloads, app updates, and tracker updates, plus system proxy detection Auto-update channels — Stable, Beta, and Latest Across Channels policies with separate download and install phases Diagnostics — Structured logs, exportable diagnostic ZIPs, database integrity checks, automatic DB rebuild, and Linux GPU rendering fallback Personalization — Light/dark/system theme, 10 color schemes, 26 languages, and first-launch system language detection Motrix Next 3.9.4 changelog: Motrix Next 3.9.4 promotes the 3.9.4 beta cycle to stable. This release refreshes bundled engine binaries, improves task detail readability and copy actions, expands link handling for magnet and ED2K workflows, polishes responsive navigation and text wrapping, updates browser extension documentation, and refines network preference controls. New Features Task Detail copy actions — Added copyable values for task metadata and reusable render functions for long text fields. Magnet and ED2K lifecycle support — Added task lifecycle handling for magnet and ED2K links. History cleanup for deleted tasks — Deleted tasks can now remove matching history records. User-Agent management — Added user-agent management and improved related network preference controls. Browser extension documentation — Added the Firefox Add-ons link for the Motrix Next extension. Improvements Engine binaries — Updated bundled binaries for supported architectures. Task Detail readability — Long task names, URLs, tracker values, and copyable metadata now render more clearly. Deletion messaging — Refined localized task deletion text for clarity and consistency. Text wrapping — Improved URI input wrapping and task name multiline display. Navigation layout — Improved sub-navigation responsiveness. Disk allocation default — Changed the default file allocation method to trunc. Proxy controls — Improved proxy button styling in network preferences. Download: Motrix Next 64-bit | ARM64 | macOS ~20.0 MB (Open Source) Links: Website | macOS / Linux | Screenshot Get alerted to all of our Software updates on Twitter at @NeowinSoftware
    • NVIDIA officially supports Ubuntu, as linked above with the GeForce NOW Hands on I did in collaboration with Paul Hill.
    • TO be clear I am not running linux today, however I keep thinking about it. And I want to make sure there are minimal obstacles if I decide to make that switch in the coming months.
  • Recent Achievements

    • Proficient
      Eric Biran went up a rank
      Proficient
    • Dedicated
      Conjor earned a badge
      Dedicated
    • Week One Done
      Windows Guy earned a badge
      Week One Done
    • Dedicated
      Mark Spruce earned a badge
      Dedicated
    • Collaborator
      conkir earned a badge
      Collaborator
  • Popular Contributors

    1. 1
      +primortal
      479
    2. 2
      PsYcHoKiLLa
      244
    3. 3
      Steven P.
      72
    4. 4
      FloatingFatMan
      66
    5. 5
      +Edouard
      66
  • Tell a friend

    Love Neowin? Tell a friend!