Le football est un sport populaire? C'est surement pour ça que lorsqu'un match à objectif européen a lieu, aucune chaîne française ne le retransmet.

Foot est synonyme de fric, rien d'autre.

Build 4.5.0.4 of xlsgen adds support for custom data mapping when importing data from CSV, HTML, JSON and XML files.

If you know in advance that in column 2 of your file, there's data of the form 152 kms, where kms is short for kilometers, xlsgen will by default import the data as a string and as a result it will not be possible to make numeric calculations on this column without first translating strings into numbers.

Beginning with this build, with the introduction of the ColumnDataMappingFormatByIndex property, you can specify a custom data mapping, before importing the file, which takes care of this and as a result imports the data as numbers with the appropriate number format. The consequence is that any calculation can be triggered right after the import phase process without manual changes.

Here is an example of source code, for an HTML file, but this works equally well for CSV, HTML, JSON and XML files :

xlsgen::IXlsWorkbookPtr wbk = engine->New( L"output.xlsx" );

xlsgen::IXlsWorksheetPtr wksht = wbk->AddWorksheet( L"Sheet1" );

wksht->Import->HTML->Options->ColumnDataMappingFormatByIndex[3] = L"0 \"kms\""; // note the use of double-quotes

wksht->Import->HTML->ImportFile(L"input\\data.html");

wbk->Close();

There is hope that in the course of 2018, pattern recognition, some form of AI, will be added so that xlsgen can figure out those patterns alone, so automatic type inference will be more powerful. It has already been improved in xlsgen 4.5

Windows on ARM (WOA) is dead on arrival (DOA).

What made the success of Windows? Every time a new release came, it came with 100%, or almost 100%, compatibility with every application and driver that existed before. In fact, what every new version of Windows was additions, new capabilities, that developers could embrace, and as a matter of fact take as upgrade incentive for their own business matters.

The release of Windows Vista was the first to cause serious harm to this, and to all software developers out there. That was when Microsoft decided to change the device driver model. This in effect broke every existing device and left users forced to either wait for hardware vendors to ship a compatible driver (never came) or to purchase Windows Vista compatible devices.

That alone almost broke Microsoft. But the install base was so big, everyone took their time to consider Windows Vista, that in fact taking just 3 years before upgrading left enough time for a number of device drivers to come along. In fact, this set of device drivers of the Vista era were shipped as Windows 7 back in 2007.

Everyone then was happy to upgrade and return back to their business.

Then Windows took a big wrong turn with Windows RT, a release that effectively broke application compatibility. No one ever upgraded to Windows RT. Big fail. Fortunately Windows RT was not the only Windows release, Microsoft indeed also shipped Windows 8. That almost failed too, but application compatibility was kept and device drivers were not changed. In fact Windows 8 took a hit only because of the UI changes, Microsoft thought that that would be a useful addition despite the fact that desktops could not take advantage of most of what was new (touch UI).

Windows 10 is released since and you know that it is successful because, again, application compatibility and device driver compatibility. The only reluctance to upgrading is that the system comes with a surveillance tool (that phones home personal data every few minutes), and ad placements in several places. Not enough to convince the majority of Windows users to avoid it at all cost, even though it is known these days that Windows 10 is evil.

Windows 10 S was then released, in 2017 and broke application compatibility. Microsoft indeed chose to only let applications from their own garden be allowed to install, essentially killing vendors out there not willing to submit themselves to uncalled for slavery. We are in 2018 and it is still too soon to tell, but there is good reason to believe that Windows 10 S will fail just like Windows RT. Even though you could find a justification as a software vendor to develop and sell software through Microsoft garden store, it appears that it is a bit too strong a thing to accept for most sensible vendors out there, essentially because Microsoft is willing to act like god : has death and life power on to you, and you lose the ability to control software that your own customers effectively install and use. Your own customers cannot install software from your website anymore.

We learn that in fall 2018, Microsoft will ship Windows RT 2, their second attempt at derailing the x86 era of software, by introducing ARM emulation. Of course, that was a non secret, this comes with compatibility problems for applications, and also zero support for existing x86 device drivers. If you read what was said before, this is effectively twice the reasons enough for Windows RT 2 (also known as WOA 2) to fail in the market. Of course, to be successful, Microsoft would have to stick it to every mouth by essentially forcing OEMs to include it in PC hardware. Never going to happen, because that would incur major risks to all OEMs out there.

J'avais déjà beaucoup d'estime pour Poutine, mais lorsque j'apprends qu'il déclare publiquement ne pas utiliser de smartphone, mon estime augmente encore un peu.

Bravo!

Et à tous ceux qui se disent à quel point c'est rétrograde, qu'ils continuent à donner gratuitement leurs données privées à des dizaines d'entreprises tous les jours.

...un des plus gros exploits du moment vient des états-unis, et BFMTV n'en rendra pas compte!

Il s'agit du l'envoi vers Mars de la plus grosse fusée ayant jamais existé, le tout avec retour en place des futs de carburant, et une voiture tesla perchée sur le satellite. Juste flamboyant.

Mais BFMTV préfère parler de neige en Ile de France, et d'ailleurs pas de la neige ailleurs. Car ailleurs, ça ne compte pas...

Je me permets de mettre un petit condensé, voulant par là suppléer le vide créé par la première chaîne (dés)info de France














Les deux images en bas sont les 2 boosters latéraux. Et en haut à gauche, le booster central qui reviendra sur une plateforme en pleine mer


retour des 2 boosters latéraux avec déploiement des supports d'équilibrage

Toutes ces images sont prises par moi du stream live.



la tesla sur le satellite

Build 4.5.0.3 of xlsgen fixes a problem related to copy/paste. Under a particular scenario, shared formulas could be incorrectly written during a copy/paste, preventing them from appearing in the output Excel file.

Ces temps-ci, lorsqu'on écoute les médias de masse, on ne peut que conclure de ce qu'ils disent comme quoi il y aurait un climat de censure, un voile posé sur la liberté d'expression.

Je crois d'une part que cette censure va de pair avec la société libérale dans laquelle nous vivons. Nous sommes libres, mais uniquement de consommer. Pas de limite pour consommer. Les banques ont même inventé les découverts bancaires pour pouvoir consommer sans limite.

D'autre part, évidemment, si quelqu'un estime qu'il ne peut pas dire ce qu'il pense, la première chose qu'il devrait commencer par dire c'est QUI le lui interdit. QUI fixe le seuil à ne pas dépasser, la ligne rouge à ne pas franchir, la blague qui est drôle de celle qui ne l'est pas?

En France, désigner QUI fixe les règles est impossible. Pour deux raisons, d'une part parce que cela serait vu comme une délation, mais une délation inacceptable, contrairement à toutes celles qui le sont justement officiellement pour dénoncer les "dérapages". D'autre part, parce que l'opinion publique se rendrait rapidement compte que ces gens en question, ceux qui fixent les règles, détiennent un pouvoir que personne ne peut remettre en cause sans craindre pour sa vie (et celle de sa famille), sa carrière, son job. Bref, peu de chances, sauf peut-être de la part d'un mourrant à la rigueur, car lui n'a plus rien à craindre de personne.

...est presque ce qui lui est arrivé de mieux, comme ça moins de gens seront intoxiqués par la propagande véhiculée autour de ces appareils autonomes puants et nocifs pour la santé mentale des gens. Vade retro satanas!

If so, I hope you are one of those whose computer got bricked yesterday by Microsoft's patch. You deserve it. No one with a single brain cell would ever trust Microsoft. Remember, those guys have built a real-time spy surveillance tool in Windows and you can't disable it.

extrait : " Microsoft has reports of customers with some AMD devices getting into an unbootable state after installing recent Windows operating system security updates. After investigating, Microsoft has determined that some AMD chipsets do not conform to the documentation previously provided to Microsoft to develop the Windows operating system mitigations to protect against the chipset vulnerabilities known as Spectre and Meltdown.

– Microsoft spokesperson"

Really Microsoft, you don't test patches before shipping them?

Voici la dernière twitteritude en date, officielle :

extrait : "Blocking a world leader from Twitter or removing their controversial Tweets would hide important information people should be able to see and debate. It would also not silence that leader, but it would certainly hamper necessary discussion around their words and actions."

Twitter annonce que si vous dites de grosses conneries, infondées, diffamatoires, etc., ce n'est pas grave à partir du moment où vous êtes quelqu'un. C'est comme un droit de cuissage, mais pour la liberté d'expression.

Le gueux, lui, devra se contenter de se faire fermer son compte de manière arbitraire.

Décidément, twitter mérite plus que jamais de partir en fumée. Et si vous utilisez activement twitter, en publiant des choses, vous faites partie du problème. Un peu comme le consommateur compulsif est le pire ennemi du citoyen.

I am a bit troubled by the announce that most CPUs out there have a flaw related to kernel mode and priviledge access to memory. The reason why is why should applications store data in memory in the clear? Why shouldn't modern applications store hash and/or encrypted data anytime they are dealing with sensitive data?
So we are supposed to be let known that servers out there host different websites using the same memory base, i.e. shared memory. I think any such website should be shut for endangering their users. Any website worth its salt, pardon the irony (NB : salt is a crypto keyword), should host sensitive data, if in the clear, using non-shared memory, period. And of course, sensitive data should not be stored in the clear, not in the hard drive, not in memory either : it is a matter of good principle.

With those principles implemented, today's cataclismic revelation is just a non-story.

Build 4.5.0.2 of xlsgen adds support for Unicode-based CSV files. Those files are encoded using either Unicode UCS2 (i.e. fix 2-byte characters), or Unicode UTF-8 (i.e. variable length character size, neutrally convertible to Unicode UCS2 and vice versa).

The interesting bit is that it is automatically detected when a CSV file is submitted to xlsgen for opening/importing. So there is no need to know in advance that the CSV file (or buffer) is encoded with Unicode or not, and there is no need to set the corresponding CSV import option (by the way, only the UTF8 option is available in the xlsgen object model, so even if you knew that your CSV file was encoded using Unicode UCS2, there was no mechanism to make xlsgen read it).

This adds up to the arsenal of automatic inference tooling in xlsgen.

Build 4.5.0.1 of xlsgen fixes a problem related to deleting columns (worksheet.Columns("A:F").Delete(); for instance). Under certain circumstances, xlsgen could write back the resulting file with negative column indexes, and Excel would identify this as a file corruption problem.

Je me demande comment il peut se faire que les problèmes à répétition à la SNCF ne soient pas sanctionnés par des licenciements pour faute grave. En effet, d'une part le système tout entier est tellement informatisé que le problème, une fois qu'il apparait, peut être dénoué plus facilement précisément avec l'aide d'algorithmes pensés pour ce faire, des re-routages qui permettent de décongestionner un point bloquant sans que cela n'impacte même minoritairement ce qui fonctionnait jusqu'ici à proximité.

Si cette gestion purement algorithmique ne fonctionne pas et que les usagers se retrouvent dans la panade, c'est la démonstration d'une incompétence crasse, et la direction doit payer pour ces erreurs avec un coup de pied dans le cul. Les algorithmes actuels doivent être mis à la poubelle : ils ont en effet fait preuve de leur inefficacité. Et de nouveaux algorithmes doivent être mis en oeuvre sur le champ, avec une nouvelle équipe de direction.

Google and Apple app stores are known for being wall gardens. But a number of years ago, Apple added constraints to it, as it that wasn't enough, for instance requiring any app update to support the latest SDK available. Well, turns out Google is about to do the same. This is a slippery slope and the consequence for Android users will be the same then Apple users, which is that their device at some time becomes unable to update apps, but for any app that shares its logic on the client side, run apps at all. This makes devices little more than throw away gadgets, which happens to be perfectly in line with what I think of mobile devices.

I don't know if Microsoft store does the same but honestly I don't think this matters because Microsoft has zero traction there.

Deux fois en deux ans que je me fais livrer par Chronopost et deux fois de suite que le colis n'est pas livré et que, lorsque je me connecte sur le site chronopost.fr, ils osent afficher le message suivant :

"Echec de livraison suite à l'absence du destinataire, avis de passage déposé."

Totalement faux. Il n'y a eu aucun passage, et aucun avis de passage déposé dans la boite aux lettres.

Comme je n'ai aucun problème avec les autres livreurs, que cela soit la poste ou des entreprises privées, je sais que c'est chronopost qui est en cause, et qui pratique la fausse livraison de manière habituelle.

J'attire l'attention des lecteurs qu'il est préférable de ne pas utiliser chronopost comme moyen de livraison. Que, de toute façon, la livraison en 24 heures n'est pas respectée. Et que la fausse livraison, à en croire ce que dit google sur le sujet, est une pratique courante de cette entreprise.

Voilà une entreprise à propos de laquelle si j'apprenais qu'elle déposait le bilan, je pourrai me dire que le capitalisme fonctionne bien : les canard boiteux cessent leurs activités. Et ce serait très bien comme ça. Mais il faut croire que le capitalisme est plus une théorie qu'une pratique...

To my amazement, it isn't Firefox only that not automatically deletes cookies on close, despite the option checked in, as a matter of fact Chrome does the same.

If you want to make sure that Chrome actually deletes cookies, you have to manually delete them. I say so, perhaps there is a special setting which allows to do it, but unlike Firefox there is no about:config form.

I'm not invoking a conspiracy theory, but I wonder if, given that Chrome is Google's tool for world domination, perhaps there is a simple explanation for Chrome not removing personal data even when instructed to do so with the corresponding settings.

In theory, in the interest of people, governments out there should sue Google for their active contribution to stealing data they do not own.

With that said, in a world where no government on earth has sued a company such Microsoft despite forcing the use of an operating system designed to steal personal data (in Windows 10, telemetry cannot be disabled), I'm not sure this will happen.

Contrary to one of Firefox's settings "Delete cookies on close", cookies are not deleted. Anyone can understand why Firefox would implement such flaw intentionally, it's because of Firefox's bottom line : they are sponsored by ad-running businesses such as Google, and for this reason Firefox can't, by default, be too destructive about a mechanism for building user profiles such as cookies.

Fortunately, for end users, it is possible to make sure cookies are indeed deleted on close.

Here is how it works :

- in the address bar of Firefox, type about:config and hit enter
- if you are getting a warning, confirm you are willing to do so
- search for setting browser.sessionstore.privacy_level
- by default this setting exists, but it has a value of 0.
- double-click on it, and change the value to 2.
- close the about:config window.

If you update your web browser, chances are the update process will screw up how http and https text hyperlinks work and from then you are unable to click links in applications such as Outlook without getting the following error message :

This operation has been cancelled due to restrictions in effect on this computer. Please contact your system administrator.

or, in French :

Cette opération a été annulée en raison de restrictions en vigueur sur cet ordinateur. Veuillez contacter votre administrateur système.


What happens is two things :

- the internal .htm and .html references are screwed up, and no matter how many Windows settings are supposed to get you back to default Internet Explorer settings, this won't repair those references.

- Firefox, if you happen to want to make it your default web browser, will prompt you to do so on startup, but as you press the corresponding option, won't fix the .htm and .html references and will bork a new one by creating a FirefoxHTML-ECSKJHKJZUIYIUY reference instead of the FirefoxHTML reference.

What you can do, if your intention is to use Internet Explorer whenever a link is clicked in say Outlook :

- hit the Windows+R key on the keyboard
- type "regedit" and hit enter
- Windows will prompt you to confirm. Confirm.
- expand the following registry key : HKEY_CLASSES_ROOT/.htm
- on the right hand side, the first string is called the default string and should contain htmlfile.
- if that isn't the case, right-click on this string, and click Edit/Modify.
- do the same thing with key HKEY_CLASSES_ROOT/.html
- you can now exit the registry editor
- links should work properly

If you would like to use Firefox :

- hit the Windows+R key on the keyboard
- type "regedit" and hit enter
- Windows will prompt you to confirm. Confirm.
- expand the following registry key : HKEY_CLASSES_ROOT/.htm
- on the right hand side, the first string is called the default string and should contain FirefoxHTML.
- if that isn't the case, right-click on this string, and click Edit/Modify.
- do the same thing with key HKEY_CLASSES_ROOT/.html
- lookup HKEY_CLASSES_ROOT/FirefoxHTML
- if it exists, then all is fine
- if it does not exist, but a similarly named key such as FirefoxHTML-EC6489787ZYZ exists, right-click on the key on the left hand side, and choose Rename. Then rename it to FirefoxHTML. In other words, remove in the name the portion of the string starting with the dash.
- you can now exit the registry editor
- links should work properly

The advisory that external researchers have found multiple buffer overflow vulnerabilities in sensitive Intel software ( running priviledged in all recent computers) should either shock anyone who has any remote understanding of IT somehow or simply just find resemblence with a known pattern which comes in 3 letters : NSA, i.e. intentional flaws that are left for US or other foreign sabotage/spy services to exploit.

Recently Microsoft fixed a very old problem in Windows which looks just like this planted NSA vulnerabilities. These are the actual backdoors that everyone, for two decades at least, were speculating about as if this was some kind of a conspiracy theory. Just that these have been true all along and both Microsoft and Intel have failed at us, everyone of us.

Curiously enough, no one is suing those guys. As if sexual harrasment, even in gross terms, was way way more important to sue these days.

Me think this sexual harrasment meme has been planted by powers that be to deceive from the true subjects of discussion.