Nolite mittere high-tech ante porcos
Following a recent post of mine, I kept reading a number of reviews of recent smartphones, and I’ve been taken aback by the degree of shallowness and sheer stupidity encounter every one in a while.
This is not to say that the general public wouldn’t exhibit a nonchalant idiocy; should I be a maker of computers, laptops, or any other gizmos of interest, I’d have an army of people put to dig on various sites and comment à la “it’s not our product, it’s you, dumb ass!”
For instance, from what I’ve read in the past year or so, there are at least four smartphones for which I’ve read a rather large number of complaints along the line “I cannot open its back cover, what do I do know?”, going up to “I had to send back the smartphone because I could’t fucking open the cover!”. These smartphones were: Lumia 435, Lumia 532, and now Wiko U Feel Lite and Wiko Lenny3.
With regards to Wiko U Feel Lite, on Amazon.fr:
And the obvious:
About Wiko Lenny3, on Otto.de:
And the obvious:
Truth is that some devices are hard to open, especially the first time, should they have a removable battery which is not inside the phone on delivery. And even experienced reviewers (that is, technology bloggers) don’t use their brains when they open the reviewed smartphone: sometimes, all it takes is to use both thumbs as a fulcrum, so that the forefingers form a sort of a lever with the two upper corners or the two bottom corners of a back cover. But people lack even the most elementary sense of mechanics.
Oh, and the aforementioned phones can be opened in 4-5 seconds–video reviews on YouTube are the best proofs, but I won’t post links to them.
I’m now going to show you how unreliable and contradictory can such reviews be. Putting your trust in a technology blogger is pure insanity. For one, generally “experts” and end-users disagree on smartphones (experts value devices that are like a Lamborghini, whereas users care about functionality; experts don’t value that much battery life; and end-users are much more sensitive to the price), but there is more than that. Experts almost expect each and every smartphone to be like a Samsung Galaxy S7 Edge or iPhone 6s Plus, and they penalize anything that’s not at the utmost level of technology and performance. And they can be even dumber than that: a couple of “professional reviewers” have complained about “the strange placement” of the fingerprint sensor in Wiko’s U Feel and U Feel Lite, most likely being accustomed with the back cover placement, but completely forgetting that the iPhone itself has the fingerprint reader on the home button! (E.g. here: “…enrolling and using five different sets of fingerprints is needlessly cumbersome with the sensor’s position.” and “Awkward fingerprint sensor location.”) Yeah, how professional of them…
It’s also amazing how such reviewers seem to have blinkers and never realize that many fabless (fabricationless) OEMs are not OEMs at all, and in some cases they share the same device made by some Chinese manufacturer and almost entirely designed by the latter, with only a few customization for each brand owner. For instance, Wiko Fever is the same as Micromax Canvas 5 E481, and identical to Blu Life ONE X. In the past, Wiko Rainbow was identical to Micromax A-120, etc. This is not a bad thing per se–many such “OEMs”, from the “French” Archos and Wiko to the “Romanian” Allview (whose phones are usually made by Gionee), not forgetting the “American” BLU, the “Spanish” BQ and the fuzzy-located Prestigio (HQ in Cyprus, but belonging to Asbisc from Belarus) have some high-quality products in their line, alongside some crappier devices; but I’d like that some people make an inquiry as to the actual manufacturer of a device, as this might be quite relevant. Also of note, at some point in the past I’ve been able to identify the manufacturer of some Prestigio smartphones based on the CEE certification documents submitted to a German lab–see, it can’t be that difficult, after all.
Getting to some details now–about that “unreliable” thing.
❶ Speaking of that fingerprint sensor, is it a good or a bad one?
It’s a bad one: “The phone also comes with a fingerprint sensor, but not one of the best. Lots of work to be done by Wiko in this department, in my opinion.”
It’s a very good one: “capteur d’empreintes efficace.”; “un excellent capteur d’empreintes digitales.” Same sensor: “capteur d’empreintes efficace.”
Same opinion: “Positiv überrascht hat mich der Fingerabdrucksensor. Dieser ist ziemlich flott und in ca. 9 von 10 Fällen funktioniert er auf Anhieb. Im Übrigen ist dieser immer aktiv auch ohne, dass man die Taste wirklich drücken muss. Ein drauflegen des Fingers reicht vollkommen aus um das Smartphone zu entsperren. Daumen hoch dafür!”
If 9 out of 10 is not good enough, maybe the latest iPhone would be a wiser choice. It costs almost five times as much, but hey, it’s perfect.
❷ How about the display of Wiko Robby? To put it in context, it’s one of the brightest displays around, with 620 cd/m². Most smartphones struggle to reach 500 cd/m², and even recent models can have less than 400 cd/m². Here, Acer Liquid Z630 caps at 340 cd/m² or maybe at 346 cd/m², his larger brother Z630s reaches 367 cd/m², and Samsung Galaxy J5 can’t go beyond 353 cd/m².
Oh, it’s only average in brightness, thank God it was a rainy day: “Die Helligkeit des Gerätes ist durchschnittlich, größere Probleme hatte ich beim Ablesen im Freien jedoch nicht (lag vielleicht auch nur am verregneten Sommer ;-)).”
Change of mood: “Despite having a HD resolution on a 5.5-inch IPS display, the Robby’s display doesn’t disappoint in terms of clarity and brightness, it is also responsive to my touch as I’m able to type really fast on Swiftkey Keyboard, you can also be assured of an immersive video watching experience on the phone with the inclusion of the phone’s loudspeakers.”
NotebookCheck.net: “Brightness: The panel shines with up to 621 cd/m² (614 cd/m² in the practical APL50 test).”
Almost agreeing: “Screen: Las imágenes son nítidas y el nivel de brillo es relativamente alto — sólo cuando estamos a plena luz del sol puede darnos un poco de guerra por culpa de los dichosos reflejos.” Wow, 621 cd/m² is only relativamente alto.
A unique note that I’ve found of interest: “One thing impressive about the phone is the responsiveness of the screen brightness, which is able to change and adapt as quick as the blink of an eye. That however is also the downfall, as it can be quite inaccurate sometimes.”
❸ Same device, assessing the GPS:
Allegedly bad: “The GPS reception was not strong enough for locating the smartphone indoors. It took quite a while outdoors, and then rarely did it find enough satellites to really track us reliably. The maximum accuracy, according to “GPS Test”, was 20 meters (~66 ft). … Robby placed us on the wrong side of the street in the crossing area, and let us “fly” over roofs. We rarely really drove on the path along the riverside, and also flew over the river if we are to believe Wiko’s Robby.”
Or is it? GPS: screenshot shows accuracy of 8 ft (about 2.5 m).
❹ The GPS was a debatable issue for Acer Liquid Z630 and Z630s, with people complaining on Amazon.de that the reported position jumped erratically even when the device was stationary. What is the truth? I simply don’t know.
Apparently, 630s can have a GPS accuracy of 6 ft (2 m). And it can be rather fast at that: “Le GPS ne nous a quant à lui posé aucun problème. GPS Data a ainsi fixé la position du téléphone à froid en 10 secondes en moyenne, ce qui est excellent.”
Or maybe it’s only average: “Acer’s Liquid Z630S utilizes the GPS satellite network for localization. It quickly found satellites with an accuracy of 12 feet (approx. 3.5 meters) outdoors. It was not possible to connect to a satellite indoors. In our GPS test, deviations of approximately 230 meters between the GPS module in the smartphone and the Garmin Edge 500 GPS reference navigation system could be observed. Acer’s handset is on a good level and will be more than enough for geocaching and fitness tracking.”
And it can get worse: “Ich habe auch das Z630. Meine bisherige Erfahrung mit dem GPS: Bei der Navigation (mit here, der Acer Navigations App und Google Maps) läuft es gut, bei here ruckelt der Standortanzeiger etwas, in der Acer App läuft es flüssig. Bei der reinen Positionsanzeige springt der Standort immer etwas hin und her und liegt dann zwischen 20 und 150 Meter vom tatsächlichen Standort entfernt.” (That was vox populi, not a professional reviewer.)
❺ I’ll try now to quickly discuss a broader topic, using Marshmallow with only 1 GB of RAM. And I’ll start by mentioning the £89 Wileyfox Spark, a device on which all the reviewers agree that it’s almost unusable because of its only 1 GB of RAM. (Not that I’d fully trust such reviewers. For a 5″ 720p IPS screen that has 420 cd/m² or 480 cd/m², there are idiots complaining that the screen is awful because it lacks oleophobic coating. But here’s the catch: any oleophobic layer is rendered useless by any screen protector, so whoever plans to add a screen protector shouldn’t care about a smartphone’s own screen coating, but rather look for a quality protector, possibly of tempered glass with oleophobic coating.)
The problem with these reviewers is that practically all of them (at least this, this and this one) have added a SD card used as “adoptable storage,” which is much slower than the internal storage. This is justified by the fact that the nominal 8 GB of storage actually translate to “almost 4 GB,” 3.5 GB or 2.6 GB of usable space, depending on when the reviewer has inspected the issue–right after the first boot, or after all Google apps have been updated (the idiocy with the Google apps is that the updates are in the user storage area, duplicating, not replacing the older copies stored in the firmware, thus wasting more than 1 GB of storage). So the abysmal performance is also due to the apps being read from a low-speed storage. (It also helps to know that even the internal storage is sub-par on this device.)
But there’s one more issue. It’s this mantra of “empty RAM is wasted RAM, because it’s better to have a dormant app in the RAM rather than having it loaded from the eMMC storage, which is slower.” While this is generally true for devices with 2 GB, 3 GB or 4 GB of RAM, this is 100% wrong for 1 GB devices!
With low-RAM devices, having the memory full “because it’s better this way” means that every single time the user wants to load an app, or even when they need to switch back to the launcher and have the home screen shown, Android must trigger the garbage collector and make room for the new app. And this is very slow.
Android is not that dumb. With 2 GB of RAM, one may notice that 1.1 GB are always full, even with no 3rd-party apps installed, but with 1 GB, the system and the mandatory apps (Google Play Services, the keyboard, etc.) can only take 700 MB or less (even less than 500 MB if needed, it depends on the graphics). The automatic memory management can be good… but then the user installs stupid apps.
By “stupid apps” I understand all the apps that load themselves once the system boots up, no matter nobody started them. The load GUI-less, more like a service, and they do so in order to have a quicker start once the user launches them. From about 200 apps that I typically install on my phone (and no, it’s no Marshmallow), at least 40 will end up as “silent services” that eat up my RAM–for a 2 GB device, I usually have no more than 300 MB free, unless I trigger a memory cleanup, and then I only reach 700 MB of free RAM!
Marshmallow has a feature called “Doze” that can “put to sleep” the apps that aren’t actively used for long periods of time, but this is definitely not enough. I don’t even want an app to load itself not even once, as long as I don’t manually launch it. After a reboot I want my memory clean.
Task killers are useless, if not harmless, for they keep killing apps that relaunch themselves in the background. On rooted devices, there are apps (such as SD Maid Pro) that can alter an apps’s autostart receivers so that they don’t start on BOOT_COMPLETED (or on similar intents). This is a tedious undertaking though.
The proper solution seems to be “the Chinese way”: some Chinese manufacturers–from the “truly Chinese” Huawei to the “French” Wiko–have added to recent smartphones a system app that allows the user to block any app from automatically starting, no rooting required! It’s only when an app simply doesn’t start unless told to do so that a user can trust their smartphone to run smoothly!
Unfortunately, this is not a feature of Marshmallow, regardless of its flavors (AOSP, CM13, whatever). Even with 2-3 GB of RAM, such a feature (I don’t recall its name in Huawei phones, but it’s part of Phone Assist or Assistant in Wiko phones) is a must-have!
And this explains why some 1 GB Wiko Marshmallow phones can be much more usable than Wileyfox Spark. Of course, not all the reviewers agree–and, as previously said, even NotebookCheck.net has got an incorect AnTuTu value because they probably didn’t make use of the aforementioned feature–but the overwhelming majority of end-users here in Germany praise the high value-for-the-price and relative fluidity of cheap devices such as Wiko Robby and Wiko Lenny3, despite having only 1 GB or RAM and running Marshmallow!
As for the reviewers, a German one writes about Robby: “Die Performance und Leistung des Wiko Robbys ist milde ausgedrückt leider katastrophal. … Alle Anwendungen laufen extrem langsam, ruckeln, brauchen ewig zum Starten und das Benutzen des Robbys ist streckenweise einfach eine Qual.” However, instead of the normal ~25K score in AnTuTu, he gets 16930, a proof that he too was running AnTuTu with apps that were restarting and reloading themselves in the background! (As a general fucking note, why not running AnTuTu prior to installing any other app?)
Another reviewer believes that “the Robby’s user experience isn’t a satisfactory one … the Robby still delivers a somehow choppy user experience most of the time, I would put the blame on its underpowered MediaTek processor, which is based on a very dated Cortex A7 architecture that shouldn’t exist on smartphones today.” No, honey, the CPU is fine, the RAM is not. There are ways to mitigate that, but you’re too dumb to think of them.
Once again, “normal” German people (that is, not gamers) are very satisfied even with such devices. (That is, those who are able to open the back cover.)
❻ Something about battery life times now. A modest capacity doesn’t always mean what you think it means. Not always.
By NotebookCheck’s professional measurements:
- Acer Liquid Z630, 5.5″ 720p, 4000 mAh battery: WiFi Surfing v1.3: 13h 59min
- Acer Liquid Z630S, 5.5″ 720p, 4000 mAh battery: WiFi Surfing v1.3: 11h 21 min (different CPU, 8-core)
- Wiko Robby, 5.5″ 720p, 2500 mAh battery: WiFi Surfing v1.3: 11h 52min
- Wiko Lenny3, 5.0″ 720p, 2000 mAh battery: Surfing via WLAN v1.3: 9h 9min
Not that catastrophic, right? Sometimes a big 4000 mAh battery doesn’t perform as expected…
- Robby, 5.5″ 720p, 2500 mAh battery: TechTest.org: an hour of YouTube streaming over Wi-Fi (average brightness) takes 12% of the battery; GeekBench3 Battery Test: 10h 13m.
- U Feel, 5.0″ 720p, 2500 mAh battery: TechTest.org: an hour of YouTube streaming over Wi-Fi (average brightness) takes 14% of the battery; GeekBench3 Battery Test: 5h 50m.
- Lenny3, 5.0″ 720p, 2000 mAh battery: TechTest.org: an hour of YouTube streaming over Wi-Fi (average brightness) takes 16% of the battery; GeekBench3 Battery Test: 7h 30m.
There is no benchmark that would match a user’s usage pattern, but I’d rather value the “YouTube Wi-Fi streaming” as relevant for non-gamers. I also value the GeekBench3 battery test, for it stresses the CPU, thus showing how much games or other computational-intensive tasks would deplete the battery. In the above comparison, U Feel (thus also U Feel Lite) disappoints, despite having a bettter battery than Lenny3, most likely because the slightly better MT6735 is more power-hungry than MT6580M.
Knowing that, only a moron could get an incomplete GeekBench3 Battery Test result of 1 minute and 30 seconds of discharging from 72% to 53%!
One more battery assessment for Robby: 7h 21m in PCMark, as compared to Moto G4 Plus (7h 30m) and Samsung Galaxy S7 Edge (7h 39m).
It’s rather difficult to trust a battery benchmark… and I’m afraid it’s even more difficult to trust the (otherwise sympa) guys from Top-for-phone.fr: for all recent Wiko models sporting Marshmallow, they added such a note: “Après une nouvelle session de tests liés à l’autonomie, il apparaît que le Wiko Ufeel Lite est atteint du même problème que l’ensemble des mobiles Wiko sous Android 6.0, à savoir une consommation excessive. Ainsi, les nouveaux tests montrent une consommation de 22-24% de batterie pour 1h de vidéo.” The problem with their video playing eating 20-22% of the battery (as opposed to an expected 14-18%) is the way they tested it: they used VLC to play a movie. What if the problem is with VLC itself, which might run amok under Marshmallow? What if a certain codec (or rather decodec) performs badly under Marshmallow? Why not testing with a YouTube 720p video, which at least doesn’t raise codec issues? Alternatively, why not playing a movie whose codecs are supported by the native Android player?
Trust them, for they don’t know what they’re doing…
❼ I mentioned that a device’s usable storage space can be very different if measured after all the included apps have been updated (and the space decreases specifically after updating the firmware-based apps, such as Google apps; userspace apps are simply replaced by an update), or right after the first boot. When there are non-wanted apps that can be uninstalled, I’d also count the available space only after having them uninstalled–because the space they’re taking is actually available to the user. Here’s how jokingly unreliable are the reviewers with regards to this topic. (Country note: Wiki U Feel, but not U Feel Lite, installs in France, right after the first boot, some 20 “Wiko experience” apps that are not installed in other countries; I disregarded reports for the available space with such apps installed.)
These similar “16 GB” phones with quasi-identical firmware have that much user-available space:
- Lenny3: 11.10 GB or maybe 8.48 GB.
- Robby: 11.01 GB or maybe 8.92 GB.
- U Feel Lite: 10.5 GB.
- U Feel (outside FR): 10.6 GB.
Here NotebookCheck’s reviewers suck.
❽ I’ll end with a strange country policy that Wiko implements for its smartphones. Here you have it, as revealed by: (1) their country-specific websites; (2) the product boxes reviewed in different countries; (3) the products reviewed in France, Germany, Austria, Italy–reviews I’ve read or watched; (4) the products available on different sites, being them Amazon flavors, other online retailers or grandes surfaces; (5) a chat I had with Wiko on Facebook, in which they said they can’t explain why is that so–but it is so!
Suppose you want a Lenny3. In France and Italy, you have the choice of 6 colors, one of which is red (paprika, cinabro, vermillon, whatever; it’s a nice one). In Germany, you cannot have a red phone… but you can opt for a completely black one, whereas outside Germany only a gris sidéral is available, and no black whatsoever! WTF? There are 7 colors, but not even in France one can’t have them all!
Robby is available in France and Italy in 5 colors, different from the 4 colors available in Germany. The website lists them twice, the first time incorrectly (Space Grau, Rose Gold, Limone oder Türkis), then correctly (Schwarz, Space Grau, Rose Gold, Türkis). Once again, black is only available in Germany… but in Germany one cannot have a phone in vert citron! So there are 6 colors in all, 5 of which for France and Italy, 4 of which for Germany.
Going beyond colors, the MSRP (UVP in German; prix public conseillé en français) is identical in France and Germany for models such as Robby (129.99 €) and U Feel (199.99 €; note that U Feel Lite is not available in Germany!), but it’s severely different for Lenny3: 99.99 € in France versus 119.99 € in Germany! CORRECTION: U Feel’s MSRP in Germany is 229.99 €, not 199.99 € like in France. CORRECTION TO THE CORRECTION: As of Sept. 5, 2016, U Feel entered physical stocks of Saturn (even here in Leonberg!) at 199 €! At least, this is the online price, as sometimes they’re a tad bigger in stores.
I’d say that they hate Germany and everyone is free to purchase Wiko phones from Amazon.fr (which ships to Germany for a good price) in case they want a nice vermillon or citron phone, but then… how about the black (100% black, not anthracite or whatnot) that seems specifically made for Germany?!
This is Europe. A continent with completely, totally fucked-up people and companies.
❾ Oh, I forgot to mention that are so many nice phones (and cheap at that!) that I cannot use simply because they only support nano-SIMs. And nobody seems to understand why is this an issue.
In countries like Germany, prepaid SIMs are very popular. And prepaid is never available as nano-SIM, but only as a combo micro/mini-SIM, for a better compatibility. There is a reason most Chinese phones (“Chinaware”–not the top models available in Europe!) are supporting not only micro-SIMs, but even mini-SIMs. It’s called compatibility. Or portability. Not everyone wants to be bound by a contract! (Even if I am so; but I also use a prepaid card for international calls.)
And even if they do prefer contracts, suppose they travel in a different country–bonjour, vacance!–and they they find a local prepaid option that seems convenient for both Internet and international calls–ComputerBILD recently had a feature on that–yet their phone only supports nano-SIMs? Oh my, but nobody told them that prepaid SIMs are usually not available as nano-SIMs! Owners of iPhone, Nexus, Sony Xperia, but also of lots of cheaper devices should rely on the roaming options of their current contract…
That’s why, as long as this doesn’t change, I’ll never purchase a phone that only supports nano-SIMs.
And I’ll keep telling people names, because they’re sooooooo stupid… (here’s a smart guy, but this is rara avis nowadays).
Since in this post I discussed (among others) brightness, relying on NotebookCheck.net for measurements, I’ll comment here on the absurd way brightness is measured by the same guys for AMOLED screens.
Take their review of Samsung Galaxy J3 (2016):
From their review of Samsung Galaxy A5 (2016):
First of all, what the fuck: for TFT or IPS screens, “maximum brightness” really means “maximum” (and the “average” is the average maximum brightness for the entire screen area), yet here they say (J5): Maximum: 306 cd/m² (Average: 298.8 cd/m²), but the “real maximum” is 540 cd/m²; then (A5), Maximum: 397 cd/m² (Average: 379.7 cd/m²), but the “true maximum” is 681 cd/m² . What the fucking fuck? For most phones, enabling the automatic brightness doesn’t give the maximum brightness, but here the contrary is true?! On most phones, the automatic brightness should be disabled and the brightness slider should manually be set to maximum to reach a maximum. Here we have either a “short-term maximum,” or an “automatic-only maximum.” Is Samsung a bunch of nuts, or what’s wrong with these AMOLED screens? Why can’t they reach a stable maximum without such confusing settings? Because the “normal” maximum values of 299-306 cd/m² or 380-397 cd/m² are actually too low for a modern phone. Way too low–most reviews say about a 420 cd/m² phone that “it’s too dark.”